Page 1

Switch Boards Blur the Line with Routers The magazine of record for the embedded computing industry

November 2007

www.rtcmagazine.com

Server Blades: Many Choices

for the Right Slot

An RTC Group Publication

Wireless Networks: Small Nodes Cover Big Jobs Exclusive! Small Form Factor SIG Debuts


GE Fanuc Intelligent Platforms

Three things you wish you had more of. Time. Space. Money. daqNet gives them to you. In today’s naval computing applications, three things are at a premium. Time. Space. And money. Supposing we told you that there is a data acquisition solution out there that gives you more of all three? daqNet from GE Fanuc Intelligent Platforms is designed for high channel count sonar applications – that’s up to 192 channels of analog I/O or up to 240 channels of digital I/O, with simultaneous sampling up to 625 KHz. Not only is it high performance: it’s highly flexible too, with a choice of up to four I/O modules.

But what’s really important is that it’s a complete, pre-configured, ready to run solution – reducing your integration and testing time, and speeding your time to market – and that it’s small, reducing the amount of space you need to allocate to data acquisition. And it’s also surprisingly affordable. More time, more space, more money to do other things with. Does life get any better?

daqNet Acoustic Data Conversion System

www.gefanucembedded.com

© 2007 GE Fanuc Intelligent Platforms, Inc. All rights reserved.


SERVER BLADES:

Many Choices for the Right Slot

Cover Photo: Continuous Computing’s integrated AdvancedTCA system, with the FlexPacket PP50 packet processing blade, alleviates the need for extensive wiring required with more archaic systems as shown in the background.

12

Server Blade: A Processor with Network and Storage Interfaces

24

MicroTCA Centellis 1000 from Motorola

TABLEOF CONTENTS

46

Advantech’s Versatile and Ultra-Compact Embedded Computers

November 2007

Departments

Technology in Context Server Blades

7 Server Blades: The “Bricks and Mortar” of Telecom 12 Industry Insider 9Latest Developments in the Embedded Why Are There So Many Marketplace AdvancedTCA Blades? 16 Products & Technology 46Newest Embedded Technology Used by Industry Leaders Editorial Yes, Virginia. There is an ATCA

Stuart Jamieson, Emerson Network Power

Todd Etchieson, RadiSys

Solutions Engineering Embedded Switch Boards

Embedded & Switched 22 Ethernet: Boards Form the Path for Industrial Control Siva Raghavareddy, Motorola

for Embedded 28 Networking Switches and Routers: A Primer Nauman Arshad, CWCEC and S. Rajesh Kumar, Aricent

Featured Products

Views and Comment 60News, Embedded Computer Market Continues Suite Offers Choice of Three to Fly High Processor-Based Cards 34 AMC Extreme Engineering Solutions

34

Virtex-5 SXT XMC Card Speeds Integration of Embedded Custom FPGA Computing Curtiss-Wright Controls Embedded Computing

Industry Insight Wireless Infrastructure

Not About Milliamps Alone when Building an Effective Sense 36 It’s and Control Network Niek van Dierdonk, GreenPeak

Industry Watch Small Form Factor SIG

Case for Smaller Form-Factors 42 The in Industrial Systems Robert A. Burckle, Small Form Factor SIG

Digital Subscriptions Avaliable at www.rtcmagazine.com/digital

November 2007




november 2007

Publisher PRESIDENT John Reardon, johnr@r tcgroup.com EDITORIAL DIRECTOR/ASSOCIATE PUBLISHER Warren Andrews, warrena@r tcgroup.com

Editorial EDITOR-IN - CHIEF Tom Williams, tomw@r tcgroup.com MANAGING EDITOR Marina Tringali, marinat@r tcgroup.com COPY EDITOR Rochelle Cohn

Art/Production CREATIVE DIRECTOR Jason Van Dorn, jasonv@r tcgroup.com ART DIRECTOR Kirsten Wyatt, kirstenw@r tcgroup.com GRAPHIC DESIGNER Christopher Saucier, chriss@r tcgroup.com DIRECTOR OF WEB DEVELOPMENT Marke Hallowell, markeh@r tcgroup.com WEB DEVELOPER Brian Hubbell, brianh@r tcgroup.com

Advertising/Web Advertising WESTERN REGIONAL ADVERTISING MANAGER Stacy Gandre, stacyg@r tcgroup.com (949) 226 -2024 WESTERN REGIONAL ADVERTISING MANAGER Lauren Hintze, laurenh@r tcgroup.com (949) 226 -2014 EASTERN REGIONAL ADVERTISING MANAGER Nancy Vanderslice, nancy v@r tcgroup.com (978) 443 -2402 The magazine of record for the embedded computing industry

INSIDE SALES MANAGER Carrie Bowers, carrieb@r tcgroup.com (949) 226 -2029

Billing The magazine of record for the embedded computing industry

Free Online www.rtcmagazine.com Spotlighting the Trends and Breakthroughs in the Design, Development and Technology of Embedded Computers. Search Archived Editions along with the Latest News in the Embedded Community. www.rtcmagazine.com

Mag gie McAuley, mag giem@r tcgroup.com (949) 226 -2024

To Contact RTC magazine: HOME OFFICE The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Phone: (949) 226-2000 Fax: (949) 226-2050, www.rtcgroup.com EASTERN SALES OFFICE The RTC Group, 96 Dudley Road, Sudbury, MA 01776 Phone: (978) 443-2402 Fax: (978) 443-4844 Editorial Office Warren Andrews, Editorial Director/Associate Publisher 39 Southport Cove, Bonita, FL 34134 Phone: (239) 992-4537 Fax: (239) 992-2396 Tom Williams, Editor-in-Chief 245-M Mt. Hermon Rd., PMB#F, Scotts Valley, CA 95066 Phone: (831) 335-1509 Fax: (408) 904-7214

An RTC Group Publication



November 2007

Published by The RTC Group Copyright 2007, The RTC Group. Printed in the United States. All rights reserved. All related graphics are trademarks of The RTC Group. All other brand and product names are the property of their holders.


november 2007

EDITORIAL

Yes, Virginia. There is an ATCA by Tom Williams, Editor-in-Chief

I

s ATCA the conundrum that wouldn’t die? By that I mean less the technical understanding—which is quite complex in itself—but more figuring out how and when it will grow to this projected multi-billion dollar market, who will actually benefit from such growth and what will be needed for all this to come together. It appears that the world of embedded systems is bifurcating into one area dominated by small form-factor boards armed with powerful processors and serial connectivity, and another region of what might be called “the last of the great backplanes”—ATCA and its offshoot, MicroTCA. There has been a lot of controversy over whether or not ATCA is real, and there has also been what I’ve called the “moving hockey stick phenomenon”—the predicted big market upswing to the multiple billions that seemed to be receding ever further into the future, drawing VCs in its wake like rats behind the Pied Piper. There were arguments that really big OEMs had no interest in an open backplane standard, that the spec was too complex and that supporting technologies were not really in place. Many of these reservations were valid. Let’s face it. ATCA is a very involved, complex technological and business issue. But then so is rewiring the planet. It has taken longer than expected for all the elements to come together since the initial announcement of the specification to make a market. The primary factor is that in order to address the demands of IPTV, data, voice and mobile communications over IP, we have to radically raise the idea of what constitutes a “platform.” Basically, a platform represents a basis upon which a customer can begin to provide a unique added value to a customer. There was a time when a platform consisted of a microprocessor and an RTOS, and if you were lucky, a board support package. In the world of digital telecommunications, a platform is a whole different deal, a much more complex mix of hardware and software, compatibility and optimization in a world where timeto-market is literally everything. An ATCA platform consists of hardware—boards in a chassis. But that hardware mix must be compatible within the ATCA spec because there are many

options. Then there is the operating system, which is usually carrier grade Linux, but can consist of Linux and an RTOS. On top of that there is the high-availability middleware, which is increasingly required to support the Service Availability Forum’s interface specifications in order for the customer to reliably develop applications on top of that. Then there is the system management layer, which consists of another mix of hardware and software. In order for the customer to feel that he or she can meet the demanding market window, all these elements must be verified to work together and that the customer can rely on a single vendor for a point of support. The vendor will have to be able to supply reliable proven platforms consisting of chassis, boards, power supplies, cooling, middleware and system management and will have to be able to do this for several mixes of options depending on the target market of the customer. Realizing this, it is no wonder that it has taken far longer than at first was optimistically predicted for this market to really take off. That hockey stick now appears to be swinging up in the 2009 to 2012 time frame. No, we really mean it this time! But it has been a long and involved process and we’re just now getting there. People in this industry are an impatient bunch, so many of the disparaging remarks and articles can be understood. Still, there is a good deal of consolidation going on. Emerson has purchased the Telecommunications unit of Motorola, and RadiSys the telecom board business from Intel. Middleware companies like GoAhead and OpenClovis are forging tight partnerships with hardware vendors. Continuous is launching its FlexTCA line of top-to-bottom integrated platforms with hardware, middleware, protocol and management. And orders are starting to become reality. Once the beachhead is achieved, we can probably expect even more activity that the biggest optimists had predicted, with applications we haven’t even thought of yet. The biggest challenge then will be establishing the bandwidth, not figuring out how to fill it. November 2007




N IO RS L E B A

NE W AV V AI E L

NI LabVIEW.

Limited Only by Your Imagination.

Communicate via multiple protocols including Bluetooth Build and program robots with LEGO® MINDSTORMS® NXT using software powered by NI LabVIEW Graphically program concurrent, real-time applications

Develop your human machine interface (HMI) display

Target 32-bit microprocessors and FPGAs

Independently control multiple servo motors

Real-Time and Embedded PRODUCT PLATFORM

LabVIEW Real-Time Module LabVIEW FPGA Module LabVIEW Microprocessor SDK NI CompactRIO Embedded Hardware Platform

>>

Signal Processing

Industrial Control

When the LEGO Group needed parallel programming and motor control tools intuitive enough for children, it selected graphical software powered by NI LabVIEW. With LabVIEW graphical system design, domain experts can quickly develop complex, embedded real-time systems with FPGAs, DSPs, and microprocessors.

Expand your imagination with technical resources at ni.com/imagine

© 2007 National Instruments Corporation. All rights reserved. CompactRIO, LabVIEW, National Instruments, NI, and ni.com are trademarks of National Instruments. Other product and company names listed are trademarks or trade names of their respective companies. LEGO and MINDSTORMS are trademarks of the LEGO Group used here with special permission. 2007-8936-821-101-D

High-Performance Test

866 337 5041


IndustryInsider november 2007

CP-TA Releases Enhanced AdvancedTCA Interoperability Documents The Communications Platforms Trade Association (CP-TA), an association of communications platform and building block providers, has announced the release of the Interoperability Compliance Document (ICD) 1.1 and the Test Procedure Manual (TPM) 1.1. These updated documents provide enhanced interoperability requirements for thermal, manageability and data transport for PICMG’s AdvancedTCA specification. In addition to providing clarification on requirements, the ICD 1.1 provides new thermal requirements for AdvancedTCA blades, complementing the existing Shelf requirements. The enhanced TPM reflects the changes in the ICD 1.1. The documents reflect several new updates that will help ensure easy and manageable testing. The updated ICD includes new thermal classification of AdvancedTCA blades that correspond with the shelf classifications that are outlined in the ICD 1.0. It also features new airflow measurements at two acoustic noise limits. CP-TA has established automated manageability tests that will Get Connected with technology and be required for CP-TA compliance while the manual tests are highly recommended. Additionally, companies providing solutions now CP-TA has included more test coverage for E-Keying. For dataexploration transport, CP-TA has clarified the test Get Connected is a new resource for further procedures to make itinto easier to set up and configure tests; andgoal signal integrity tests have been products, technologies and companies. Whether your is to research latest datasheet from a company, speakThe directly overhauled to better align withthethe IEEE standards of testing. ICD 1.1 is free for download and with an Application or jump to at a company's technical page, the the TPM 1.1 is available for aEngineer, fee structure www.cp-ta.org/certification. CP-TA members are now goal of Get Connected is to put you in touch with the right resource. working on interoperability requirements and test procedures for AMC and MicroTCA. Whichever level of service you require for whatever type of technology,

Ad Index

Get Connected will help you connect with the companies and products you are searching for.

www.rtcmagazine.com/getconnected

development activities back to Nokia Siemens Networks. They will work together with Nokia Siemens Networks’ program management, product management, architecture, technical support and system level testing to fulfill obligations. Over time, development services will also be offered to other IBM clients as part of the portfolio of the IBM Global Engineering Solutions division, which provides engineering as well as research and development (R&D) services to a variety of industries. The transferring of R&D capabilities to trusted business partners is part of Nokia Siemens Networks’ overall strategy. This step will allow Nokia Siemens Networks to realize greater workforce flexibility and improve cost management, while being able to leverage the expertise of its R&D partners. The agreement was signed on the 28th of September, 2007.

The patch will use a modififour business lines of the Service cation of the ink delivery method, Core and Applications (SCA) Industry Group which involves a superheated vapor Business Unit to a subsidiary of explosion to deliver a drop of ink. IBM. Upon closing of this transCollaborates on Linux While the microneedles do peneaction, expected to take place by Hewlett-Packard has figured Mobile Computing the beginning of December 2007, trate the skin, they do not penetrate out how to adapt its inkjet cartridge Platform Get Connected with technology and companies providing solutions now far enough to affect the nerves. Penup to 235 employees located in technology to be able to deliver  Toisaddress thetherise in conGet Connected is a new resource for further exploration into products, technologies and companies. Whether your goal to research latest etration, however, is an sufficient toEngineer, Munich and will become small doses of medicine in the form sumer for access datasheet from a company, speak directly with Application or jump to aBerlin company's technical page, the goal ofdemand Get Connected is to to putthe you molecules skin barpart of the type IBM organization. of a drug patch. Withinthe exception touch with the right move resource. Whicheverpast level the of service you require for whatever of technology, Internet and advanced applicarieryou that are normally too largeandtoproducts Through the agreement, Nokia of nicotine and atGet least one birthwill help Connected connect with the companies you are searching for. tions on larger display mobile be absorbed. The patch is expected Siemens Networks will leverage control medication that can be abwww.rtcmagazine.com/getconnected devices, ARM and six companies to be about 2.5 centimeters square IBM’s global network and experisorbed through the skin, medical have announced a collaboration and about three millimeters thick. ence, investing more than $6 bilpatches have limited applicability that will result in the development It will also have a battery and a milion in 2006 and with 14 years of because the skin tends to act as a of a Linux-based open source platcroprocessor to control the dosage. patent leadership. barrier. The use of thousands of miform for next-generation mobile Naturally, the technology will need The transferred Research croscopic needles under the control applications. The collaboration, regulatory approval and clinical and Development Center activiof a microprocessor will allow the announced at the fourth annual testing but is expected to be availties include specific parts of R&D delivery of different doses of mulARM Developers’ Conference in able in approximately three years. for Next Generation Voice and tiple drugs from a single patch. Santa Clara, California, builds Multimedia; Media Gateway; Apparently, HP has licensed on the ARM architecture and Mobile Internet Connections; and the technology to Cropson, an its ecosystem of partners to deNokia Siemens Networks Consumer and Business VoIP Irish medical device manufacturliver a standards-based platform and IBM Partner for business lines of the Service Core er. Cropson will manufacture the using Linux. These companies Get Connected with companies and Get Connected and Applications Business Unit devices and use its experience in Transfer of R&D Center all working to accelerate the products featured in this section. companies mentioned in thisare article. as well as with technical support. Enthe medical field to find appropriActivities enablement of truly always on, www.rtcmagazine.com/getconnected www.rtcmagazine.com/getconnected tire teams will be transferred and ate partners with pharmaceutical Nokia Siemens Networks connected mobile computing will continue to deliver necessary companies to deliver appropriate and IBM have signed an agree(CMC) devices. medicines such as those needed ment including the transfer of for treatment of diabetes, arthritis specific parts of Research and Get Connected with companies mentioned in this article. and others. Development Center activities for www.rtcmagazine.com/getconnected Get Connected with companies and products featured in this section.

HP to Morph Inkjet Technology for Medicine Delivery

Products

End of Article

www.rtcmagazine.com/getconnected

November 2007




Industry Insider

Dual-Core Processors on AMC, cPCI and VME

AdvancedMC™ � up to 2.0 GHz Dual-Core Intel® Xeon® processor LV � up to 16 Gbytes DDR2 ECC SDRAM � double-width/full-height � for AdvancedTCA® or MicroTCA™

3U/6U CompactPCI® ®

� choice of Intel Core Duo processor, ® Intel Core™ 2 Duo processor or ® ® dual Dual-Core Intel Xeon processors � up to 8 Gbytes DDR2 ECC SDRAM � extended temperature operation available � rugged versions available ™

The new platform will be released into the open source community and will be naturally adopted by ARM’s large ecosystem of software and silicon vendors, including Marvell, MontaVista, Movial, Mozilla, Samsung and Texas Instruments.  This effort between ARM and its Partners focuses on the development of an open source platform based on Linux, Gnome Mobile and Mozilla Firefox that runs on ARM Partners’ advanced systems on chip (SoCs). The collaboration strengthens ARM’s OS and browser ecosystem, while leveraging the ability of ARM Partners to quickly bring products and technologies with compelling graphics, video, multimedia and Internet capabilities to consumers.   ARM demonstrated multiple Partners’ reference boards running the Linux Mobile Computing Platform at the ARM Developers’ Conference. ARM and its Partners plan to release a full platform in early 2008 and bring devices to the market in early 2009. The platform is subject to open source licensing terms and the collaboration is expected to deliver additional projects over time.

Dialogic Acquires Cantata through Purchase of EAS Group, Inc.

VME/VXS � choice of Intel® Core™ Duo processor, ® Intel Core™ 2 Duo processor or dual Dual-Core Intel® Xeon® processors � up to 8 Gbytes DDR2 ECC SDRAM � extended temperature operation available � rugged versions available

http://www.gocct.com Email: info@gocct.com Tel: (781) 933-5900 All Trademarks acknowledged

10

Untitled-1 1

Sept07/US

November 2007

9/5/07 10:02:17 AM

Dialogic has announced that it has acquired all of the outstanding shares of EAS Group. Subsidiaries of EAS include Cantata Technology, Excel Switching, Brooktrout Technology and SnowShore Networks. The acquisition was a share-based transaction and the former shareholders of EAS are now minority shareholders in Dialogic. Through the acquisition of EAS and its subsidiaries, Dialogic will expand its product portfolio and customer base in the communications technology enabling market segment with the rich portfolio of Cantata communication boards and software. The addition of the SnowShore

IP Media Server, the Excel Converged System Platform, the Excel Multi-Services Platform and Cantata Technology’s IMG 1010 product families to Dialogic’s product portfolio will enable Dialogic to expand its customer base in the service provider market segment. The addition of the Brooktrout TR1000, Brooktrout TR1034, Brooktrout TruFax and Brooktrout SR140 product families will also strengthen Dialogic’s leading position in the enterprise market segment. The new product lines will be added to Dialogic’s already expansive selection of products, which include technology-leading Dialogic Host Media Processing software, Dialogic signaling software and the industry’s broadest selection of media processing boards, SS7 signaling components and robust PBX-to-IP media gateways.

Zigbee Unveils Comprehensive New Features

The ZigBee Alliance, a global ecosystem of companies creating wireless solutions for use in utility, commercial building and consumer applications, has announced it has approved comprehensive new features for the ZigBee specification giving more choices to manufacturers as they design ZigBee products. The ZigBee specification offers all the features released in 2006, and adds new features, giving manufacturers greater flexibility when designing ZigBee-based products. The original set of features published in 2006 is now known as the ZigBee Feature Set. An expanded set of features known as ZigBee PRO maximizes all the capabilities of ZigBee and facilitates ease-of-use and advanced support for larger networks. The newly released Feature Sets are designed to interoperate with each other, ensuring long-term use and stability. Highlights of the expanded ZigBee PRO Feature Set include:


Industry Insider

• Network Scalability – Improved support for larger networks offering more management, flexibility, performance choices • Fragmentation – New ability to divide longer messages and enable interaction with other protocols and systems • Frequency Agility – Networks dynamically change channels should interference occur • Automated Device Address Management – Optimized for large networks with added network management and configuration tools • Group Addressing – Offers additional traffic optimization needed for large networks • Wireless Commissioning – Enhanced with secure wireless commissioning capabilities • Centralized Data Collection – Tuned specifically to optimize information flow in large networks

In addition to its original capabilities, the ZigBee Feature Set also gains optional use of the Frequency Agility and Fragmentation features described above. Access to the updated ZigBee specification is available immediately for all members. Free public availability of the new features is scheduled for early 2008.  ZigBee uses self-organizing and self-healing mesh networking to enable robust communications over the globally available 2.4 GHz frequency, with 868/915 MHz technology available in select countries.  ZigBee networks offer superb interference immunity and capabilities to host thousands of devices. With ultralow-power requirements, ZigBee devices run on regular batteries for years, eliminating the need for wiring to a power source and offering unparalleled maintenance convenience and installation flexibility.

SSI Teams with TI for Low-Power RF and MCU Technology

SSI Embedded Systems has teamed with Texas Instruments’ Low-Power RF and Ultra-LowPower MCU groups to strengthen the relationship between reliable technology and SSI’s embedded software development services. SSI has been developing products that utilize TI semiconductors for many years to help clients more easily match needs with experienced outsource partners. SSI is a member of TI’s low-power RF and MSP430 ultra-low-power microcontroller third-party developer networks. An example of a real-world application that was developed by SSI engineers and is powered by TI products is a wireless sensing device that connects to standard force measurement sensors using a low-cost ZigBee-ready radio for point-to-multipoint

connectivity. This device demonstrates how to use low-power RF technology for a multitude of remote sensing wireless applications. Implementations include environmental, security, or industrial monitoring. Sensor input could be anything from temperature/humidity to pressure to light to liquid levels. SSI is an existing member of TI’s MSP430 third-party network, which offers hardware, software and consultant support for TI’s line of ultra-low-power microcontroller products.

HARTING sets new standards in AdvancedMC™ connector reliability HARTING offers the widest product portfolio of AdvancedMCTM and Power connectors for AdvancedTCA® and MicroTCATM applications.

HARTING now brings the reliability of signal connectors to new heights. Introducing con:card+, a quality seal that identifies press-fit connectors that provides the highest level of mating reliability for AdvancedMCTM modules.

The con:card+ GuideSpring systematically positions the AdvancedMCTM module precisely in the connector, reducing the maximum possible offset between connector contacts and module pads by 60%. This significantly increases the mating reliability of MicroTCATM backplanes and AdvancedTCA® carrier blades. HARTING provides complete design in support, including signal integrity services and 3D modeling.

Learn more about the 5 con:card+ features at www.concardplus.com

Untitled-6 1

8/6/07 November 2007

11

9:54:47 AM


Technology InContext

Server Blades: The “Bricks and Mortar� of Telecomm Now available in proprietary designs or in emerging ATCA/MicroTCA standards, server blades offer power, flexibility, scalability and cost savings for building networking and telecomm systems from enterprise to global reach. by S  tuart Jamieson Emerson Network Power

d

exploration er your goal eak directly al page, the resource. chnology, and products

server blades

S

erver blades are becoming a valu- compute elements can fit into the space able architectural option for a required for a single stand-alone comrange of emerging telecom and puter. Blade server clusters can thus enterprise communication applications. pack more processing performance into Server blades can be implemented in a given space for less cost and lower a variety of sizes and capacities using power demand than other computing proprietary or open-source architec- approaches, attributes particularly valutures and serve in applications ranging able in space-constrained enterprise and from enterprise to global telecommuni- central office applications. cations. Finding the most cost-effective The main differences in server blade approach to a given application requires implementations are in their cost, perforpanies providing solutions now a basic understanding of server blade mance and storage capability, but they ration into products, technologies and companies. Whether your goal is to research the latest function the available implementacan also isbe optimized to support specific lication Engineer, or jump to aand company's technical page, the goal of Get Connected to put you tion technologies. functions. Common applications include ice you require for whatever type of technology, ies and products you are for. basic, a server blade At searching its most database, video, voice over Internet Protoconsists of a processor with networking col (VoIP) and storage, as well as generalcapability and access to high-capacity purpose computing. Server blades can also storage, all in a rack-mounted format be made NEBS-compliant, ready to use in (Figure 1). This format can range in size a wide-area network, and can be designed from a single card to a full shelf chassis, for high-availability applications implebut the functions are essentially iden- menting redundancy and high reliability. tical. Server blades are a compact reIn the most common implemenplacement to stand-alone computers, de- tation, server blades run the Internet signed to share resources such as power Protocol (IP) over 1G or 10G Ethernet supply, cooling and I/O so that multiple as their network connection. Using IP over Ethernet allows them to operate as part of an IP cloud at the center of the Get Connected with companies mentioned in this article. network without incurring the overhead www.rtcmagazine.com/getconnected penalty of protocol conversion. Blades

End of Article

12

November 2007 Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

used as gateways typically have an additional dedicated and secure Ethernet channel that serves as the entrance port to the network. The storage required for a server blade will vary somewhat with the application. In data logging, for instance, data access speeds are not an issue and the blade may use a Fibre Channel link to an external storage rack or use network attached storage. This approach of using external storage allows blades to share the storage array, saving cost, power and rack space. The shared storage approach also benefits applications like call tracking and billing, where data integrity is critical and storage needs to have redundancy. Where access speed is an issue, such as in subscriber database applications, the server blade would require a large, local memory instead. Typically, server blade designs incorporate a generic storage interface so that the same basic blade can be used in any of these configurations.

Blades Require Management

The use of multiple computing elements for a networking application,


Technology InContext

The benefit of proprietary approaches is that the system can be highly customized and optimized for its intended application. The vendor is able to implement unique features and tune system performance by being free to choose protocols and architectures. The proprietary approach also helps eliminate concerns for interoperability of system components and ensures that users know exactly where to turn for obtaining support. One drawback is that the approach tends to lock users into a single-source situation for hardware, although some companies have opened their proprietary standards to increase the number of vendors offering compatible parts. Even if open, however, proprietary standards remain under the control of a single company, limiting development to whatever the original vendor chooses to implement.

AMC

At its most basic, a server blade consists of a processor with network and storage interfaces, and may include onboard storage as well.

Open Standards/Specs Offer Choices

A design approach based on open standards offers users more freedom of choice. Open standards allow a variety of vendors to create compatible

AMC

AMC

280mm

AMC

Figure 1

181.50mm

especially when resources such as storage are being shared, requires some form of system management. Most often, blades are designed to include only generic management interfaces, allowing the system integrator to configure them as needed. Some applications, however, require that the system provide high availability. Server blades intended for such systems need to include the interface and control structures that allow system management software to use the blades in load sharing, hot-swap and failover behaviors. Designers can implement server blades in many different technologies, both proprietary and open-standardsbased. Application needs determine which technology is most cost-effective. Considerations that will impact technology choice include system cost, scalability, performance requirements and physical size restrictions for the intended installation. Three major technologies are the leading implementation approaches for server blades: proprietary approaches, the Advanced Telecommunications Computing Architecture (ATCA) standard, and the MicroTCA standard. Each has its relative strengths. Proprietary approaches typically use a rack-mounted chassis infrastructure to carry the blades. The chassis hold cards of a standard, xU, size, and is designed for NEBS-compliant, highreliability operation. To achieve high reliability, the chassis supports redundant power supplies and cooling fans along with hot-swap capability for cards and components. The chassis backplane connections, management software and network connectivity are often unique to a given vendor instead of being standards-based.

73.5mm 322.25mm

Figure 2

ATCA cards are large enough to carry considerable compute capability and are configurable using Advanced Mezzanine Card modules for I/O or other functions. November 2007

13


Technology InContext

products, allowing the user to pick and choose desired capabilities from a wide selection of alternatives. The standards/ specifications-based approach fosters competition that can result in lower cost and faster technology evolution than proprietary designs. The major drawbacks are the risk of interoperability problems when blending components from different vendors, and a more complex and ambiguous service

and support environment. ATCA and MicroTCA are the two major open specifications for telecommunications systems design, both administered by the PCI Industrial Computer Manufacturers Group (PICMG). They share a number of common elements, including management architecture and some hardware elements, but they target different physical and performance requirements.

The ATCA specification targets the needs of large, high-performance telecommunications systems. It uses 8U cards for its server blades, providing room for implementing powerful compute engines with multiprocessing and massive onboard storage options (Figure 2). The cards are configurable through the use of Advanced Mezzanine Card (AMC) modules to personalize a generic blade with specific networking and I/O alternatives or processor blocks. The chassis backplane supports high-speed serial interconnections and is protocol agnostic, allowing an ATCA system to use whatever connectivity best fits the application. While redundancy and high availability are not requirements in the ATCA standard, they are supported, and ATCA was designed with this capability in mind. ATCA specifications define a standard approach to system monitoring and management: the Integrated Peripheral Management Interface (IPMI). This interface, together with system software and required onboard hardware, gives developers the access and controls needed to initialize and configure boards and AMC modules, query board status, enable or disable a board or module’s backplane connection, and control board and module power. The IPMI management functions thus allow implementation of automatic failover, hot-swap, electronic keying and other high-availability policies. PICMG defined the ATCA specification in 2003 and AMC modules in 2005, and many board vendors have embraced the specifications. As a result, a wide variety of chassis, board and module options are available from a host of suppliers. System developers have a variety of hard drives, processor blades and networking options from which to choose, and new capabilities with greater performance are continually appearing.

MicroTCA Extends Scalability

The MicroTCA specification uses many of the same architectural elements of ATCA, but targets smaller system requirements. To keep things compact,

1 14Untitled-4 November 2007

11/8/07 9:51:08 AM


Technology InContext

Single Tier Shelf

0-#( !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# -#( 0-

Cube Shelf

0-#( !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# -#( 0-

!-#

00-#( !-# !-# !-# !-#

!-# !-#

!-#

!-#

!-#

-#( 00-

0-#( !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# -#( 0-

Back-to-Back Shelf

Two Tier Mixed Width

Two Tier Fixed, Single Width Shelf

The Harsher the Environment, the More You Need MEN Micro! D7 Xeon® Blade Server 6U CompactPCI®

Pico Shelf

!-# !-# !-# !-# !-# !-# -#( 0!-# !-# !-# !-# !-# !-# -#( 0-

0-#( !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# !-# -#( 0-

Figure 3

AMC AMC MCH PM

MicroTCA systems use AMC modules directly in a backplane to provide highly compact system options.

MicroTCA utilizes AMC modules as chassis cards plugged directly into the backplane rather than as mezzanine cards. This reuse expands the market for AMC modules to include both ATCA and MicroTCA implementations, helping further drive down costs and foster innovation. MicroTCA uses the same management architecture and system software as ATCA as well, allowing MicroTCA also to support high-availability applications. Using AMC modules as the chassis card allows construction of MicroTCA systems that pack considerable performance into a small package. The standard allows creation of systems with as few as four cards or as large as a full rack width (Figure 3), giving developers many structural options. Because the AMC modules themselves are large enough for implementation of a full, high-performance compute engine or a slim hard disk drive, MicroTCA is fully capable of creating a moderate-sized blade server. The commonality of hardware and software in ATCA and MicroTCA systems gives system designers an opportunity to offer significant scalability. An application could start out with a MicroTCA system that has only a few server blades. As performance needs

increase, the MicroTCA system can accept additional blades in a chassis or an additional chassis without altering the fundamental architecture or applications software. If requirements increase even further, users can switch to an ATCA chassis while retaining hardware and applications software compatibility with the original MicroTCA design. Similarly, the same application code can run on systems that span a wide performance range, allowing one basic design to serve a host of different performance and installation requirements. Whether standards/specificationbased or proprietary in architecture, the server blade approach is becoming the preferred design style for meeting telecommunication needs. Blades offer system developers a flexible and modular design structure that promises lower cost and higher compute density than stand-alone servers. At the same time, it offers customers a greater range of choices for size and performance levels and scalability to ensure that their systems can keep pace as their needs grow. Emerson Network Power, Embedded Computing Madison, WI. (608) 831-5500. [www.emersonembeddedcomputing.com].

For mission-critical applications, MEN Micro offers the D7 6U CompactPCI Blade Server: ■

■ ■ ■

■ ■ ■ ■

2 dual core Xeon® 1.66 GHz ULV processors (4 cores total) cPCI system slot with PCI 64-bit/ 66MHz or PCI-X 64-bit/133MHz 4 GB DDR2 SDRAM with ECC Non-volatile SRAM and FRAM SATA and PATA support for hard disk, CompactFlash and more I/Os include UARTs, USBs, Ethernet 2 XMC or PMC slots Up to 2 GB Ethernet channels Long-term availability: 5 years+

2007

ICA Technology Showcase

WINNER!

MEN Micro, Inc. 24 North Main Street Ambler, PA 19002 Tel: 215.542.9575 E-mail: sales@menmicro.com

www.menmicro.com November 2007

15


Technology InContext

Why Are There So Many AdvancedTCA Blades? The growing adoption of AdvancedTCA by a variety of equipment segments is fueling a rich mix of COTS and semi-custom solutions. by T odd Etchieson RadiSys

d

exploration er your goal eak directly al page, the resource. chnology, and products

server blaDes

E

www.rtcmagazine.com/getconnected

CPU Blade CPU Blade

November 2007 with companies mentioned in this article. Get Connected

Networking Blades Networking Blades Networking Blades Networking Blades

16

System Manager Base + FC I/F Switch Base + FC I/F Switch

with companies mentioned in this article. www.rtcmagazine.com/getconnected

System Manager Base + Ethernet Switch Base + Ethernet Switch System Manager

Get Connected

System Manager

FC Storage Blade FC Storage Blade

Networking Blades Networking Blades Networking Blades Networking Blades

End of Article

CPU Blade CPU Blade

CPU Blade CPU Blade

System Man Base Interfa Base Interfa System Man

quipment makers, seeking to differentiate their systems, are leverCompute Application aging semi-custom features offered by AdvancedTCA (ATCA) board and system vendors. Although ATCA is an industry standard with requirements around form-factor, thermals, switch Storage Application fabrics and interconnects, there’s still sufficient flexibility for system developers to provide innovative designs panies providing solutions now for target applications. This allows ration into products, technologies and companies. Whether your goal is to research the latest developers to incorporate their proprilication Engineer, or jump to a company's technical page, the goal of Get Connected is to put you etary technology and optimize their ice you require for whatever type of technology, ies and productssystems you are searching for. performance, managearound Networking Application ability and cost. As a result, ATCA is addressing equipment segments beyond telecom and networking, such as industrial and military, with high-availability requirements. The adoption of ATCA platforms is gaining momentum and ramping quickly, with sales expected to grow at Figure 1 AdvancedTCA blades support an 81 percent compound annual growth various applications. rate between 2006-2008. In a recent

study by Venture Development Corporation, sales of merchant AdvancedTCA integrated platforms in 2008 are forecasted to be more than three times that of 2006. “This is consistent with VDC’s view of how new products compete and are adopted into proprietary architectures in cost-competitive markets. At the same time, we forecast strong growth for the larger market in the future,” says Eric Heikkila, director of VDC’s Embedded Hardware and Systems Practice. ATCA blades share a common form-factor, simplifying the task of creating multiple network elements based off a single platform. This common architecture fosters both multiple applications with the same assets (blades) and interoperability among different types of blades used in various applications (Figure 1). It also makes it easier for system vendors to offer comprehensive infrastructure solutions, whereby equipment makers can integrate additional elements from the ecosystem as necessary.


Technology InContext

There are many categories of ATCA blades supporting a range of functionality including compute, storage and networking. Furthermore, board vendors offer a mix of I/O, compute capacity, backplane interfaces and so on, providing equipment makers with a wide choice for components and subsystems (see sidebar “Common ATCA Blade Types” ).

When One Size Doesn’t Fit All

Despite the availability of a wide range of “spec” blades, equipment makers may still need semi-custom blades to reach their cost objectives or meet particular system requirements, such as adding a proprietary interface. In addition to designing a custom ATCA blade or developing a custom module for a carrier blade, board vendors are offering commercial

Essential building blocks … Innovative solutions. From Digital Signal Processing, to I/O interfaces, to IP cores and Software tools … ... at BittWare you’ll find the essential building blocks you need for your signal processing applications. Whether your application needs to thrive under extreme conditions of the battlefield, meet the ever-changing requirements of the communications market, or process highly-precise medical imaging data, BittWare has the capabilities to meet your needs. Reliable cutting-edge products are backed by timely delivery, reduced cost, and a dedicated support team – enabling your innovative solutions.

off-the-shelf (COTS) blades that provide some additional flexibility to address special requirements. Equipment makers migrating to ATCA platforms typically find that a majority of their hardware requirements are met with COTS components. However, there are times when system developers choose to leverage an existing board and software for either differentiation or continuity of legacy functionality. This entails converting the board to an ATCA version, which is a well understood process due to the maturity and extensive use of the ATCA specification. In telecommunications where five nines availability is a critical equipment requirement, some telecommunication equipment manufacturers (TEMs) develop boards in-house to deploy and protect their intellectual property. For example, schemes for keeping CPUs in lockstep so failover is almost instantaneous are typically safeguarded. In these cases, custom blades may be the best choice for maintaining the effectiveness of the proprietary solution and keeping it secure. Carrier blades provide an efficient means to integrate a wide range of standard and custom functionality. Since carrier blades already address system, chassis and manageability requirements, AMC module developers can focus on designing in their special functions. Supporting multiple sites, these blades allow capacity to scale granularly

Zone 3

SAS Disk

COM Express Module

20 Port 10 GE Zone 2

24 Port GE

w w w. B i t t Wa re . c o m

Zone 1

Figure 2

1 18Untitled-1 November 2007

6/19/07 10:50:39 AM

Switch blade with an open module site.


Technology InContext

Figure 3

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

System Manager

Compute Blade

Switch Blade

Switch Blade + System Manager

Switch Blade

Switch Blade + System Manager

System Manager

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Compute Blade

Open site frees up blade slots.

by adding or removing modules as system needs or requirements change. This flexibility makes it easier to tailor systems for large and small customers. Equipment makers can adjust system capacity and capability within the same platform and more quickly address the needs of different market segments and improve their time-to-market. Carrier blades can also be used to combine functionality that may have previously resided on separate blades. A single carrier blade can accommodate modules for both control and data planes resulting in a more consolidated solution. As an alternative to AMC carrier blades, board vendors are adding open module sites to common building blocks like switch and compute blades. For example, the RadiSys Promentum ATCA2210 10 Gigabit switch and control module has a COM Express site that enables equipment makers to add intelligence to the switch by adding a control module or system manager to the baseboard (Figure 2). In addition to providing a higher level of integration, this capability may free up a slot by eliminating the need for a separate blade. In Figure 3, the switch and sys-

tem manager functions are consolidated, usually providing cost reductions and saving a board slot to be used for revenue generation. For highly redundant systems, two slots can be freed up. High-performance compute blades often include an AMC expansion site that can be populated with hard disk drives or Ethernet cards. Adding storage and I/O directly on the compute blade helps reduce cabling and increases system density. In other cases, an adjunct processor can be placed in the expansion slot, providing an offload for functions like security processing. Flexible baseboards may support mezzanine modules other than AMC, which allows developers to forgo implementing some complex features such as hot swap. These modules can also provide TEMs a simpler, faster and cheaper means to integrate their IP without developing a full-blown ATCA blade. TEMS developing next-generation systems may still need to maintain support for legacy signaling software. Instead of designing an ATCA version of older proprietary boards, system developers can use blades that provide highest density media processing and ease the

Common ATCA Blade Types Blades can be categorized as hub boards that provide the central switching resource in a shelf or node boards that communicate to other nodes through backplane interfaces. Here are examples of common ATCA building blocks: •C  arrier blade: A node board containing several (2-8) mezzanine slots, typically Advanced Mezzanine Card (AMC), for integrating a wide range of functionality like processing, storage and I/O functionality. • Compute blade: A node board consisting of a single board computer used for general-purpose processing (a.k.a. CPU/server blade). • DSP Blade: A node board integrating digital signal processors (DSPs) for processing media in applications including voice (VoIP) and video (MPEG). • Line Card: A node board interfacing to subscriber lines supporting services such as POTS, ISDN and DSL. • Networking blade: A node board processing networking I/O—Ethernet, T1/E1 and OC3—often employing network processing units (NPUs). • Storage blade: A node board providing high-capacity storage, normally across storage-specific interfaces such as Fibre Channel. • Switch blade: A hub board generally supporting 12-20 base interfaces, a similar number of fabric interfaces, as well as some auxiliary interfaces. • System Manager Blade: A hub board implementing a management processing hierarchy including Intelligent Platform Management Interface (IPMI) controllers, the shelf management controller and the shelf manager.

November 2007

19


Technology InContext

ETX form factor single board computer NANO-9452

Socket M Intel® Celeron M 1.0GHz CPU

Intel® 852GM + ICH4 chipset / DDR2 memory up to 1GB / 10/100 Ethernet / Dual channel 18-bit LVDS LCD panel / 4 x PCI support / ISA support / 2x IDE / 4 x USB/ 2 x RS-232

Fanless AMD LX-800 Universal Controller

AMD LX-800 500MHz

AMD LX + AMD CS5536 chipset / 8 x COM / Ethernet / 4 x USB / Windows CE 6.0 OPC Modbus support library provide by SDK/ ideal to work with PLC, industrial communication converter/gateway, and factory automation

Intel Core 2 Quad Mini-ITX KINO-9654G4 Mini ITX

migration of DSP software. These blades support debug tools and interfaces to enable DSP management and facilitate the debug of DSP code. For example, such a blade could incorporate Serial Rapid I/ O connectivity that loads code into DSP memory, tracks messages to and from application flash memory of the DSP and offers an alternate path from the powerful local management processor (LMP) to the DSP. Equipment makers differentiating their ATCA systems with semi-custom board techniques need to plan ahead for compliance testing. As for ATCA COTS blades, they’re already compliant with strict regulatory requirements such as the Network Equipment Building System (NEBS). Manufacturers of carrier grade equipment can appreciate the effort that goes into compliance testing, which may include agency testing and NEBS testing. Industry groups, such as the Service Availability Forum (SA Forum) and the Communications Platforms Trade Association (CP-TA), are working together to drive the certification of COTS products for the telecommunications industry. Compliance testing is an expensive and time-consuming process that requires specific facilities and staff. Some board and system vendors, well versed in the complexities and risks involved in NEBS testing, are willing and able to navigate through these challenges for their customers. In some cases, these vendors will test and validate an entire system composed of their platforms combined with customer products and thirdparty elements.

MicroTCA Expands AMC Market Ideal for High performance computing or network appliance

Intel Core 2 Extreme, Core 2 Quad CPU / DDR 2 memory / 4 Gigabit Network / 6x USB / 2xGbE / 4x RS-232 / 1x PCI express expansion / 7.1 Channel Audio

http://www.ieiworld.com

emarketing@usa.ieiworld.com

1-866-276-6754 Tel :1-909-595-2819

20

Untitled-1 1

November 2007

Fax: 1-909-595-2816

10/18/07 10:50:24 AM

Riding the coattails of the ATCA standard, MicroTCA takes AMC cards, originally designed to plug into ATCA carriers, and treats them like blades that slot into modular backplane/chassis systems. By providing a smaller form-factor, MicroTCA can deliver the features of ATCA to equipment at the far reaches of the telecom network. MicroTCA is also starting to get more

visibility from system designers looking for high availability in a smaller footprint, while serving the scalability needs of their network elements, especially in access markets such as wireless base stations. Outside of traditional telecom applications, MicroTCA is generating interest in field-oriented deployments—military, aerospace and government sectors—such as troops on naval ships using Wi-Fi for communications. Interest is also growing in embedded applications valuing high availability, like industrial and process control, medical equipment and instrumentation applications. The growing adoption of ATCA and MicroTCA is fueling the availability of COTS and semi-custom boards from a wide choice of vendors. Whether using COTS or semi-custom boards, developers can shorten product development cycles, while protecting their competitive advantage. Whether equipment makers are building a next-generation system from scratch or trying to add new equipment to an existing solution, flexible board features and technologies are helping to lower their cost and development effort. RadiSys Hillsboro, OR. (503) 615-1100. [www.radisys.com].


G e t o n t h e F a s t Tr a c k Don’t let custom requirements slow you down. Our existing products and custom design expertise help our clients to quickly meet complex design requirements. VadaTech is a leader in AdvancedTCA, AMC, cPCIe, VXS, XMC and custom board-level technology. Dedicated to leading-edge technology for the most demanding applications; we pride ourselves on our extraordinary capability to quickly create designs in response to our customer requirements. Our service to the aerospace and telecommunications markets has resulted in a quickly growing line of AdvancedTCA, AMC, cPCIe, VXS, XMC and custom boards for processing, communication, storage, and platform interoperability.

ATC800

Sprint past the competition with VadaTech as your teammate!

W W W . V A D A T E C H . C O M T E L 1 . 7 0 2 . 8 9 6 . 3 3 3 7


solutions engineering

embedded swtich boards

Ethernet: Embedded & Switched Boards Form the Path for Industrial Control Combining switched Ethernet with MicroTCA offers fast, scalable and standard solutions for industrial automation as well as a clear path into the future. by S  iva Raghavareddy Motorola

O

ver the years, several technologies have continued to promise better performance for enhanced network communication. Given the wide range of available network communication technologies, there has been a great deal of confusion regarding which technologies will prevail, combined with an underlying concern that today’s equipment will quickly become obsolete. Starting as an experimental work led by Robert Metcalfe at Xerox Palo Alto Research Center in the early 1970s, Ethernet has withstood the test of time for the last three and a half decades. During this time, researchers continued to improve Ethernet, accommodating the increase in bandwidth requirements, from 3 Mbit/s to 10 Mbit/s, 1000 Mbit/s, 10 Gbit/s and now looking toward the future for 100 Gbit/s. As more telecom companies offer Ethernet services, the technology is viewed more as a carrier standard and not just a LAN standard. Ethernet dominance is highlighted in a recent report by analyst firm, IDC. According to IDC, Ethernet use is broadening, with demand being driven primarily by overall bandwidth needs, while increasing its presence in various vertical segments. Adding to the complexity of the situation, a general shift toward client-server architecture has significantly increased network traffic. As a result, processes—such as separating smaller LANs by smart switches—must be

22

November 2007

Past

Present

ISDN

Frame Relay

Future Ethernet First Mile

XDSL

WAN

POS/RPR

T1 OC-48 ATM

25

OC-192

OC-3 OC-12 OC-48

10-GE 40/100-GE

100B-VG Token Ring

FDDI

LAN Ethernet 100 kb/s

Figure 1

1BASE5

10BASE-T

1 Mb/s

FE 10 Mb/s

GE 100 Mb/s

1 Gb/s

10 Gb/s

Network Communication Technology Overview. (Courtesy of Ethernet Alliance.)

adjusted to ensure they are sufficient and able to scale to the ever-changing requirements of network communication technology. With various communication technologies available, equipment vendors must decide which one best suits their needs and also allows for future expansion. During this time of continuous change, as shown in Figure 1, Ethernet remains the single technology that has systematically continued to deliver performance with an easy upgrade solution involving minimal equipment changes, thereby enabling end users to maintain state-of-the-art communication equipment at relatively low cost. Ethernet switching has been adopted as the primary communication mechanism

for open architecture systems, beginning with the PICMG 2.16 packet switched backplane to the current AdvancedTCA and MicroTCA architectures. MicroTCA, which uses AMC modules as building blocks, dedicated AMC.2 specification for Ethernet usage. A MicroTCA chassis that supports Ethernet, PCI Express and SATA can be seen in Figure 2. The SCOPE alliance recommends using the first two ports of available 21 ports in an AMC for Ethernet to support control plane requirements of any application. For the remaining ports, depending on the application, OEMs may choose to use Gigabit or 10 Gbit/s Ethernet, or entirely different interconnect technolo-


con:card+

70

the most reliable AdvancedMC connectors

A new standard in reliability con:card+ is a new quality seal that identifies connectors possible offset between connector contacts and module pads providing the highest level of mating reliability for by 60%, thus significantly increasing the mating reliability of con:card+MicroTCA EvenTMunder high and shock and vibration, con:card ® backplanes AdvancedTCA carrier blades. + connectors AdvancedMCTM modules.

will keep the AdvancedMC™ modules in the correct position ™ ™ con:card+ connectors feature the unique GuideSpring By preventing the connector, con:card+ connecmodules. helping tomodule assurefloat the in reliability of your MicroTCA backplane technology. The GuideSpring pushes and holds the tors also enhance the reliability of AdvancedMCTM systems under ® and carriade. AdvancedMCTM module against the connector housing. shock andAdvancedTCA vibrations. ™ module precisely within the slot prior to By systematically positioning the module precisely in the connector,™con:card+ connectors reduce the maximum

system tremendously.

innovation by +

HARTING + ept – at

www.concardplus.com

Learn more about the 5 con:card+ features at www.concardplus.com


SOLUTIONS Engineering

Intel® Pentium® M 745 1.8GHz, 2MBL2, ATX

Figure 2

EPIC

/PM

Based on Intel® Pentium/Celeron® M processor. Intel 855GME / ICH4 chipset with integrated Intel® Extreme Graphics 2 engine 2x32 MB VRAM UMA. 2x DDR-RAM-SODIMM for up to 2GByte.

Fanless Intel® 852GM Celeron® M 600 MHz Micro PC

23

37

S-

EE

Onboard Intel® Celeron® M 600 MHz with 512K L2 Cache CPU Intel® 82852GM Chipset Scalable with Evalue EPIC SBC One SODIMM up to 1 GB DDR SDRAM

IDE Flash Drive Carrier Board with Micro SD Interface

MicroTCA Centellis 1000 from Motorola.

gies like PCI Express, Serial RapidIO or SATA. This shows how important Ethernet is, as compared with other technologies. One of the primary reasons for successful deployment of Ethernet in open architecture systems is the ease in which it allows connection to the outside world. Whatever backplane interconnect technology is used, Ethernet makes it easier to connect to external entities. For example, Serial RapidIO may give the best protocol efficiency when compared to Ethernet; however, this can only be used within the MicroTCA system between the AMCs. In order to connect to systems outside, Ethernet becomes the default choice. With new features such as Power over Ethernet (POE) and Wake-on-LAN (WoL), Ethernet became even more popular. POE allows certain embedded boards to remain in hibernation, until it “wakes up” for a magic packet, allowing the board to consume power only when required. Similarly, ability to transmit power over the Ethernet makes it easy and affordable for a consumer, where he just plugs one cable into the embedded board that can 1973: Experimental Ethernet at 3 Mbits/s

SD

-ID

E-4

1997: Full Duplex Ethernet

1982: Ethernet Bluebook by DEC, Intel & Xerox

0/4

4

support both data and power to operate. With increasing client-server architecture, network convergence with data, Voice over IP (VoIP) and video services, traffic dramatically escalated, adding additional pressure on the existing network infrastructure. In 2000, wireless modems supporting home networks didn’t exist. But today, the majority of devices employ wireless networks, including the latest VoIP phones, which support video features as well. All these devices continue to use Ethernet as the primary communication interface. The biggest drawback of Ethernet when compared to some of the other competing interconnect technologies is the processing overhead that’s required to run the entire TCP/IP software stack in the operating system. For example, unlike TCP/IP, PCI Express and Serial RapidIO do not need a separate software stack to perform the required communication. This requirement removes certain processing bandwidth of the CPU from applications. Thus, a certain percentage of CPU time must be dedicated for the TCP/IP stack, with the remaining time used for the application. As you can see in the Ethernet’s key milestones (Figure 3), the industry is currently in a position to take advantage of 10 Gbit/s speed, with the future looking toward 100 Gbit/s in the next few years. The question remains, can processors take advantage of the high data rates by doing the required processing? As shown in Figure 4, this issue can be resolved by using a separate TCP/IP Offload Engine (TOE) that can process the frames at line rate from the Ethernet and pass the packets directly on to the application. This relieves the OS,

2000: Trunking

2006: 10 Gigabit Ethernet

Available in 40 and 44 pin header configuration Support PIO 0-4 and Ultra DMA 3 mode. Bootable from Transflash/micro SD. Low power consumption.

www.tri-m.com

info@tri-m.com

1.800.665.5600 tel: 604.945.9565

1975: Patent filed by Xerox

1985: First IEEE standard for 10 Mbits/s Ethernet

fax: 604.945.9566

Figure 3

24

Untitled-4 1

November 2007

3/1/07 10:31:23 AM

1995: Fast Ethernet at 100 Mbits/s

Ethernet’s key milestones.

2003: Power over Ethernet 1999: Gigabit Ethernet

20XX: 100 Gigabit Ethernet


What if it rains? Would it matter? Hop on and drive. Up winding mountain roads. Across wide open plains. Darting through city traffic. Feel the engine hum, the sun, wind, dirt, and yes, even the rain. Now consider the fact that a fully functional, ultra-mobile PC powers the all-digital dashboard and controls the Intel® Chopper’s mechanical functions. Are you beginning to get the idea? This is not just a really cool bike – it’s a marvel of Intel® embedded technology in an amazing form factor. What would you do with high-performance technology that rugged, that durable? What kind of form factor would you use? To learn more go to: http://developer.intel.com/design/info/905.htm

Intel Embedded Technology. Igniting Innovation. Intel and the Intel logo are registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. Copyright © 2007 Intel Corporation. All rights reserved.


SOLUTIONS Engineering

PCI/104-Express with CoreTM2 Duo PCI Express

Software

Applications TCP IP

Operating System

Hardware

Applications with two displays video streaming HD-videos

TCP Traditional NIC MAC PHY

Figure 4

TOE Adapter

IP MAC PHY

A TCP/IP Offload Engine (TOE) takes over processing the TCP/IP stack from the main CPU and its operating system thus greatly reducing the overhead in high-speed Ethernet. (Courtesy of Ethernet Alliance.)

MICROSPACE® MSM945 incl. SMX945 _ PCI/104-Express baseboard _ Intel® CoreTM Duo LV L2400 / Intel® Celeron® M ULV423 _ 1.0GHZ to 2x 1.6GHz _ i945GM-PCI Express chipset _ CRT, SDVO, 224MB VRAM _ KB/MS, FD, 1x P-ATA, 2x S-ATA, 2x COM RS232, LPT1, 8x USB 2.0, 1x LAN 10/100 BASE-T, AC97-7.1 HDA _ DDR2-RAM 256 – 2048MB _ IDE _ Watchdog _ Power 5W / typ 10-20W _ Smart cooling concept _ -25°C to +60°C (Option -40°C to +70°C)

DIGITAL-LOGIC AG offers reliable Embedded Computers in PC/104, 3.5”, EPIC, EBX, smartModule, COM Express and other formats. Further information:

www.digitallogic.com

26

Untitled-1 1

November 2007

11/1/07 2:05:30 PM

and hence the CPU, from having to run the TCP/IP stack and handle all the interrupts. With the ability to support high data rates and range of application protocols, Ethernet is rapidly gaining momentum in the industrial automation space. Key attributes of industrial automation are reliability, precision and timing. For Ethernet, this means bounded latency, bounded jitter, guaranteed bandwidth for each traffic group and line-rate capacity depending on the application. With the support of non-blocking full duplex switches, quality of service provision and traffic shaping policies, Ethernet was able to easily meet the requirements of Industrial Automation. Some of the advantages of using Ethernet in an industrial automation environment are remote monitoring, management and control of the equipment. This results in improved efficiency by providing real-time access to the industrial automation equipment, such as sensors and actuators, from virtually anywhere. This also provides an inexpensive and easy way of adding a whole new level of functionality, which can provide additional opportunities to add more intelligence to the entire automation system, maximizing the return on investment. An Ethernet network in an industrial automation environment must provide high availability and reliability for the entire system. This is best solved by providing redundancy at various levels including system, interface, device and network with a keen interest on failover time. Ethernet in Industrial Automation offers: • Real-Time access to accurate information to facilitate fast decision making by management at corporate headquarters

• Quality control with immediate feedback, resulting in better process • Ability to take advantage of existing IT infrastructure for non-time-critical automation monitoring and control • Open Systems approach providing interoperability and support of multiple protocols such as Ethernet/IP, Profinet, etc., over Ethernet • System and/or individual unit scalability that comes with Ethernet as a standard interface • Net result of reduced downtime of any industry or manufacturing floor • Flexibility to incorporate future advances in technology Now that open systems such as MicroTCA are targeting different markets including traditional telecom, rugged military, medical and industrial automation, Ethernet starts playing an important role in connecting various entities. In an industrial automation environment, embedded controllers play a critical role by not only controlling the process, but also by continuously collecting high-speed data, which may be used to alter the process. Here MicroTCA can be used to keep ahead of the curve, due to the existence of high-speed switched fabric interconnects like Ethernet. Any precision equipment manufacturing facility can take advantage of MicroTCA and its high-speed data interconnects to get better mileage. With the help of the PICMG special interest group on Ruggedization of MicroTCA, reliability requirements of industrial automation can easily be addressed. MicroTCA’s hot swap feature can help further by reducing the downtime of the entire plant. Having come a long way, Ethernet continues to grow through research and adoption in various markets with features such as wireless, high speed, scalability, etc. With current work going on for 100 Gbit/s Ethernet, the possibilities are endless to take advantage of the high speed and low cost due to economies of scale. With mixed voice, data and video supported on the same network resulting in a full network convergence, it would be an amazingly connected world. Motorola Embedded Communications Computer Group Tempe, AZ. (800) 759-1107. [www.motorola.com].


Emerson. Consider It Solved is a trademark and Business-Critical Continuity, Emerson Network Power and the Emerson Network Power logo are trademarks and service marks of Emerson Electric Co. ©2007 Emerson Electric Co.

Carrier-grade performance combined with flexibility and configurability makes KAT6200 the right solution The KAT6200 Processor Blade is a high-performance, flexible AdvancedTCA® server node made to: An intelligent FRU for high-availability server with storage and I/O A flexible off-the-shelf hardware and software ecosystem Configures to a wide variety of control and packet processing applications Upgrades securely with redundant firmware and persistent memory for event logging Extend – Fully supports IPMI and AdvancedTCA extensions for standards-based shelf management Order Work Span Last

– – – –

Set your expectations high. Choose the global technology leader in carrier-grade server nodes. Just another reason why Emerson Network Power is the global leader in enabling Business-Critical Continuity ™. ACPower Systems

Embedded Power

Precision Cooling

Connectivity

Integrated Cabinet Solutions

Services

DC Power Systems

Outside Plant

Site Monitoring

Embedded Computing

Power Switching & Controls

Surge & Signal Protection

US Toll Free: 1-800-356-9602 ■ Voice: +1-608-831-5500 www.emersonembeddedcomputing.com


solutions engineering

embedded switch boards

Networking for Embedded Switches and Routers: A Primer From dedicated switches and routers to multi-layer switches, content switches and Web switches, manufacturers are advancing their technologies to keep up with the ever growing demands in traffic and sophistication in today’s networks. by Nauman Arshad, CWCEC and S. Rajesh Kumar, Aricent

28

November 2007

Source Node

Destination Node

Switch or Router

Encapsulated Data

MAC Address: 00-05-5D-4B-C3 IP Address: 10.0.0.1 Application Presentation

Application Presentation

Session Transport Network Data Link Physical

Session Transport Network Data Link Physical

Port 1 2

Figure 1

MAC Address: 00-12-2A-01-D1 IP Address: 10.0.0.2

Layer 2 Switch

Address 00-05-5D-4B-C3 00-12-2A-01-D1

Network Data Link Physical

Layer 3 Router

De-Encapsulated Data

E

thernet is today’s dominant network medium of choice, offering the best mix of simplicity, reliability and cost-effectiveness, across residential, business, enterprise and carrier networks. As more and more devices become networked and as the need to access, disseminate and share information continues to increase, networks and networking protocols have evolved to handle the complexity and the size of the information. To begin, it is important to understand the difference between embedded switches and routers. This requires familiarity with the Open Systems Interconnection (OSI) model. The OSI is a specification published by the International Organization for Standardization (ISO) in 1984 that defines a seven-layered model. which simplifies complex network interactions by breaking them into simple modular elements (Figure 1). In this framework, each OSI layer can only communicate with the layer directly above it, below it, and with its peer layer on another device. Following the typical flow of data between two communicating nodes in a network, a source node sends encapsulated data down the seven-layer OSI stack, across the

Port Address 1 10.0.0.1 2 10.0.0.2

Illustration of seven OSI layers with function and example protocols discussed in this article, including sender, receiver, and where the physical cable, switches and routers fit. It also shows a MAC address and IP address The layers discussed are: • Data Link—Layer 2: VLAN, Link Aggregation, Port Mirroring, Spanning Tree Protocols • Network—Layer 3: IPv4/v6, RIP, OSPF, IP Multicasting • Transport—Layer 4: UDP/TCP • Application—Layer 7: SNMP (?) >


One-Stop Shopping  Bring Us Your List Everything you need for your CompactPCI® Applications Performance Technologies’ CompactPCI® products and solutions are the industry’s most capable and complete open-standards offerings for telecom, aerospace and defense, and commercial applications. We create innovative and robust solutions that enable you to get your products to market faster. Our integrated solutions include: Advanced Managed Platforms™ : These flexible, scalable platforms are ideal for a wide range of compute-intensive applications. Coupled with exceptional support services, we have what you need to quickly and cost-effectively develop and deploy your system. High-Performance Blades: Single and dual-core SBCs, Ethernet switches, intelligent shelf management, voice processors, serial and T1/E1 communications I/O, and RAID storage blades.

Platforms. Blades. Software. Support. Many solutions. One vendor.

NexusWare ® : Our CGL 3.2-registered and POSIX-compliant Linux distribution includes a full operating system, development tools, protocols, HA middleware, and SA Forum-compliant APIs. Best of all, NexusWare's total integration with our blade and platform solutions means you have shorter development cycles. Contact Performance Technologies today and bring us your product development shopping list.

Phone: 585.256.0200

‹

www.pt.com/amp

‹

E-mail: info-request@pt.com


SOLUTIONS Engineering

(a)

(a)

2

tch

i Sw Source Node

STP Blocks VLAN-B Potential loop in a network

h1

tc wi

S

Figure 2

1

3

Destination Node

4

5

Multicast Source

h2

itc

Sw

1

2

Camera 1 Camera 2

multicast

Multicast Table Camera 1: 1, 2, 4 Camera 2: --

Nodes

Multicast Table Camera 1: 1, 2, 4 Camera 2: --

3

4

5

Subscribing Nodes

IP Multicasting (a) shows interested nodes joining a Multicast group for Camera 1; (b) shows Camera 1 multicasting data sent only to the subscribing nodes.

network, and back up the stack to the destination node where it is de-encapsulated. If the two nodes are connected in a network through a switch then the data first flows from the source node to an ingress port on the switch. The switch looks up the destination address embedded in the incoming data frame with an internal switching table. Upon a positive match, the switch forwards the data frame to the appropriate egress port on the switch, which is connected to the destination node. In Figure 1, the address the switch looks up is a unique 48-bit Media Access Control (MAC) address assigned to each node on the network. Switches typically operate at Layer 2 on the OSI stack. Similarly, if the two nodes are connected in a network using a router, rather than the 48-bit MAC address, either a 32bit IPv4 (Internet Protocol version 4) ad-

30

(b)

1

2

4

2

Camera 1 Camera 2

4

join

Figure 3

h3

itc

Sw

STP Blocks VLAN-A

Destination Node

Multicast Source

h1

itc

Sw

2

1

Spanning Tree Protocol (STP) – (a) shows how STP can be used to block the link between Switch 1 and Switch 3 to avoid a loop in the network; (b) shows how VLANs can be used with STP for efficient load balancing across unused links.

(a)

1

tch

i Sw

h3

itc

Sw

Spanning Tree Protocol (STP) Blocks This Redundant Link

Source Node

2

tch

i Sw

November 2007

dress or a 128-bit IPv6 (Internet Protocol version 6) address is used, with a forwarding policy to send the data to the right destination port. Routers typically operate at Layer 3 on the OSI stack. Over the years switches have evolved to where the switching silicon can now switch at higher layers of the OSI stacks (e.g. Layers 3+) making the lines blurrier between switches and routers. These switches are also known as multi-layer switches, content switches or web switches.

LANs and VLANs

A Local Area Network (LAN) is a set of physically connected nodes in a small geographic area (e.g., a network in a home, office, group of buildings, vehicle etc.). A LAN can be divided into multiple “network segments” through the use of switches and routers for security purposes and to im-

prove traffic flow by filtering out packets that are not destined for that segment. A Virtual Local Area Network (VLAN) is the logical grouping of nodes (or switch ports) that behaves as though the nodes were connected on the same LAN segment regardless of their actual physical location. For example, in an office, the Finance Department data may be on one VLAN whereas the Engineering Department data resides on another. On a commercial aircraft, the in-flight entertainment data may be on one VLAN and data communications may be on another. VLANs can be spread across switches. When the number of nodes in a network is large, or spread over an area that cannot be reached by a single switch, multiple switches are used. In such cases, it is quite possible that multiple paths exist between switches. Multiple paths can create loops in the network, which can be harmful because data packets may get duplicated and may potentially circulate on the network indefinitely, severely degrading network performance or at worst bringing down the network altogether.

Spanning Tree Protocols

The spanning tree protocol (STP) is a mechanism that enables switches to detect and prevent loops in a network. In Figure 2a, Switch 3 blocks the port connected to Switch 1 when the protocol detects that there are two paths connecting Switch 3 to Switch 1. This blocking prevents the loop in the network. If the link between Switch 2 and Switch 3 fails, the STP automatically unblocks the link between Switch 1 and Switch 3. In this way, the STP also provides network resiliency. The original STP can require between 30 seconds and multiple minutes to restore alternate paths in a network, which makes it too slow for today’s high-speed and mission-critical networks. The rapid spanning tree protocol (RSTP) is an enhanced variation of STP that offers 10X faster convergence time, enabling the network to recover much more rapidly from changes or failures. RSTP convergence times are typically less than one second for smaller networks. With either RSTP or STP, there are unused links in the network. To better utilize the available capacity, the multiple STP provides a way to prevent loops, cre-


SOLUTIONS Engineering ate resiliency and increase network utilization by load balancing traffic across the alternate paths available. In Figure 2b, Switch 3 blocks the link to Switch 1 for VLAN A while Switch 2 blocks the link to Switch 1 for VLAN B. So, between the two nodes, traffic that is constrained in VLAN A is carried through the links connected to Switch 2 while the traffic constrained within VLAN B is carried over the link between Switch 1 and Switch 3, which provides load balancing and more effectively utilizes network capacity. When multiple switches are used to connect nodes together in a network, the paths between the switches may require higher capacity than any individual link’s speed. Also, in this case, the links between switches become more critical because they are concentrated points of failure. Link aggregation is a technique that helps mitigate the risk of failure by grouping multiple physical links into a logical channel. The resulting logical channel’s capacity is thus the sum of the capacities of the individual physical links. Using link aggregation, if one link in the channel fails, the other links continue to function and traffic is moved to the working links, which provides redundancy and fault tolerance. Other common terms used to refer to an aggregated group are “port channel” or a “trunk.” One advantage of switches is that they enable network operators to monitor specific traffic via a technique called “mirroring.” With mirroring, all traffic that requires monitoring is sent to a specific port on the switch. The network operator can monitor the network by connecting a monitoring node to that specific port to analyze the traffic going through the switch. Mirroring can be based on ports (port mirroring) or on specific flows (flow-based mirroring). Traffic across multiple ports can be monitored simultanesouly using a single port. To ensure data is not lost, filtering is used on the source ports to ensure the data does not exceed the capacity of the port that is doing the monitoring.

IPv4/v6

IP (Internet Protocol) is a network layer protocol. Nodes are assigned IP addresses that are used to communicate with each other. The IP address has a subnet part and a host part. A subnet is a group

of nodes that is logically separated from other groups (much like a VLAN). Nodes within a subnet can directly communicate with each other. Routers are used to interconnect subnets. A popular switching scheme is to map IP subnets to VLANs. When nodes in different subnets need to communicate, they must go through a router. A Layer 3 switch is essentially a router. IPv4, the older version of IP, continues to be widely deployed and used today. IPv4 uses a 32-bit IP address for nodes. With the proliferation of IP-based devices, the 32-bit space was deemed to be insufficient. In response, a new version, IPv6, was developed, which uses a 128-bit addressing scheme. One of the key new features it offers over IPv4 is a much larger address space, enabling a greater proliferation of IP-based devices.

IP Multicasting

Multicast is a very common way to send information like TV or video streaming or public addresses, from one sender to many receivers. In its simplest form, a switch sends a multicast frame to all of the VLAN ports that the multicast came in on. The network operator can set up controls to restrict the multicast to specific ports if required. A more popular approach for multicasting is for nodes to select what multicasts they want to receive. This is accomplished with the Internet Group Management Protocol (IGMP). Nodes that want to receive a multicast for a certain group send IGMP messages to join the group. By analyzing these IGMP join messages the multicast source is able to send the multicast only to the ports for which there are interested receiving nodes (Figure 3). An implementation of a deployable switch that supports IP Multicasting is Curtiss Wright’s SMS-682 SwitchBox II using Aricent’s protocol stacks (Figure 4). With up to 24 ports of Gigabit Ethernet (GbE) and 2 ports of 10 GbE, the SwitchBox II has full support for IP Multicast. In addition, SwitchBox II has support for snooping on the IGMP messages (“IGMP snooping”) sent from the receiver nodes to the sender node, enabling a switch to obtain information about which ports a multicast frame needs to be switched to.

Figure 4

Curtiss-Wright SMS-682 SwitchBox II.

RIP and OSPF

Layer 3 switches are used to interconnect IP subnets. To know which neighboring switch must be accessed to reach a particular subnet, the switches need to be configured with the reachability information. This information is called a “routing table.” The greater the number of switches or subnets the more cumbersome and error prone the configuration becomes. As a remedy, Layer 3 switches use routing protocols to both advertise their own routing tables and learn the routing tables of other switches, which results in greater connectivity. One such routing protocol is Routing Information Protocol (RIP). Using RIP, each switch simply sends updates to inform its neighboring switches which subnets it’s directly attached to. Upon receiving an update, the neighboring switch adds the new data into its own routing table. It then sends its own information, along with the updates, to other neighbors, along with a “hop count.” The hop count is the number of other switches that must be gone through to reach a subnet, and is limited to 16. Reliance on the hop count limits the scalability of RIP. It cannot handle large networks and is slow to react to changes in network topology. An alternative approach, The Open Shortest Path First (OSPF) routing protocol overcomes the limitations of RIP. With OSPF, each switch sends its connectivity information to its neighbors. November 2007

31


SOLUTIONS Engineering

This information is sent in the form of a link state advertisement. Each switch adds the received link state advertisements to its own directly attached link states and sends all of these to its own neighbors. In this manner, every switch in a large network is able to learn the topology of the entire network and calculate the routing table. Because every switch knows the complete network topology, OSPF is extremely scalable

and capable of handling large and complex networks.

TCP and UDP

Every node in the network, although having a single IP address, may support many applications that need to use network facilities. For example, computer users on the network may have several Web browsing sessions, some gaming sessions and an Internet chat session, all

using a single IP address. This is made possible through the use of transport layer protocols such as Transport Control Protocol (TCP) and User Datagram Protocol (UDP), which enable applications to share an IP interface. These transport layer protocols layer logical connections using logical port numbers on top of the same IP address. For example, the Web browser may use TCP port 1 and the gaming application may use TCP port 2, but with the same IP address. The difference between TCP and UDP is that TCP provides a more reliable communication channel by requiring acknowledgements for every transmitted packet. In comparison UDP is more unreliable but much faster.

SNMP

A popular method of managing network elements is through the Simple Network Management Protocol (SNMP). An SNMP manager program installed on a server or a monitoring node sends SNMP requests to the switches that the network operator wishes to manage. The SNMP requests may be messages to change the configuration of the switches or requests for the switches to provide information, such as traffic statistics, existing configuration, etc. The switches to be managed have an SNMP agent program that can process the request from the SNMP Manager. The switches, upon receiving SNMP requests, send back SNMP Responses with the requested information. The SNMP Responses also indicate if the manager’s request was successful or unsuccessful. SNMP has 3 versions—version 1, version 2 and version 3. The latest version, SNMPv3, provides added security to the management scheme to protect the network from intentional attacks by hackers or inadvertent, but still costly damage from inexperienced users. Curtiss-Wright Controls Embedded Computing Ottawa, Ontario. (613) 599-9199. [www.cwcembedded.com]. Aricent Palo Alto, CA. (650) 391-1088. [www.aricent.com].

32

November 2007


LEADERSHIP BUILT ON OVER 40 YEARS OF INNOVATION

Four decades of quality manufacturing, design innovation, and process control has earned us the time-honored trust and confidence of a global network of customers.

DDC IS YOUR DATA NETWORKING SOLUTION • MIL-STD-1553 COTS Boards and Components • Fibre Channel • Gigabit Ethernet • High Speed 1760 • Enhanced Bit Rate 1553 (EBR) • ARINC 429

Toll Free: 1-800-DDC-5757

www.ddc-web.com


FeaturedProducts AMC Suite Offers Choice of Three Processor-Based Cards A trio of new Advanced Mezzanine Cards for use in both AdvancedTCA and MicroTCA systems, are offered as single-width, mid-height AMCs featuring a choice of high-performance dual-core processor. The XPedite5140 hosts Freescale’s MPC8641D, XPedite7040 employs an Intel Core2 Duo, while XPedite8040 targets P.A. Semi’s PWRficient PA6T-1682M. The XPedite5140 is targeted for applications requiring Freescale high-performance Power processors; XPedite5140 uses an MPC8641D running two e600 PowerPC cores at up to 1.5 GHz. The card can be configured with up to 4 Gbytes of DDR2 SDRAM and 4 Gbytes of NAND flash. I/O support includes quad Gigabit Ethernet ports, serial ports and PCI Express. XPedite5140 is an optimal solution for packet processing and general computing applications that demand high performance. The XPedite7040 is targeted for applications requiring x86 compatibility and provides performance and I/O options while minimizing power consumption. The Ultra Low Voltage (ULV) Intel Core Duo processor or Low Voltage (LV) Intel Core2 Duo processor provides XPedite7040 with a high performance-per-watt processing rating. Additionally, the Intel 3100 chipset combines the functionalities of an integrated server-class memory controller (Northbridge) and an I/O controller (Southbridge) into a single device that is specifically intended for power- and board-space-sensitive embedded applications. The XPedite8040, running P.A. Semi’s PA6T-1682M processor, provides AdvancedTCA and MicroTCA system designers the power savings and high performance of a dual 2.0 GHz processor. Available with up to 4 Gbytes of DDR2 SDRAM, quad Gigabit Ethernets ports, dual 10 Gigabit XAUI ports and PCI Express links, the XPedite8040 is unrivaled in features and performance—at half of the AMC specified power budget. Operating system support from Extreme Engineering includes; Linux, WindRiver VxWorks, WindRiver PNE, Green Hills Integrity and QNX Neutrino. XPedite5140, XPedite7040 and XPedite8040 are shipping and available from $2,000 - $4,000 depending on memory configuration, clock speed and volume commitments. Extreme Engineering Solutions, Madison, WI. (608) 833-1155. [www.xes-inc.com].

Virtex-5 SXT XMC Card Speeds Integration of Embedded Custom FPGA Computing

An XMC module (VITA 42) compute engine is based on the Xilinx Virtex-5 FPGA and designed for demanding, high-performance signal and image processing applications including radar, sonar and signal intelligence. The XMC-442 from Curtiss-Wright is the company’s first module to feature the Virtex-5 SXT, the largest, highest-performance DSPfocused FPGA available from Xilinx. With software driver and IP development support from Curtiss-Wright’s Continuum FXtools developer’s kit, the XMC-442 speeds and simplifies the development of custom FPGA designs. The new XMC-442 mezzanine module combines the flexibility of a Xilinx Virtex-5 SXT (SX50T/SX95T) FPGA, the high bandwidth of serial switched fabrics such as PCI Express and Serial RapidIO (SRIO), and rich I/O options. It significantly speeds the integration of high-performance FPGA-based custom calculation blocks onto embedded platforms that feature on-board XMC sites, such as VPX, VME and CompactPCI boards. The XMC-442 works with Curtiss-Wright’s CHAMP-FX2 and other Power Architecture-based signal processing engines to enable system designers to integrate tightly coupled FPGA/ FPGA or FPGA/PowerPC solutions. Curtiss-Wright’s Continuum FXtools developer’s kit provides the software and IP blocks necessary to quickly develop a custom FPGA design and to deploy it on a range of Curtiss-Wright Continuum Software Architecture-compatible base cards. Central to the Continuum FXtools package are the optimized IP blocks and associated drivers, which include memory controllers, I/O controllers, a PCI Express endpoint with multi-channel DMA, a scalable switching infrastructure, and a wide variety of other utility functions. The Continuum FXtools software includes FPGA-specific function libraries for FPGA configuration, DMA control, and other XMC442-specific utility functions (including temperature and power sensors, command bus mappings and diagnostics, etc.). In addition, Continuum FXtools includes a SystemVerilog-based scriptable simulation test bench environment that includes a full set of bus functional models and IP block models. An optional SRIO endpoint block is also available. In addition, Curtiss-Wright’s Continuum IPC communications middleware package has been extended to directly support transfers between the XMC-442’s FPGA and other FPGAs or processor nodes within the system. The XMC-442 is designed to operate in rugged environments and is available in both air- and conduction-cooled formats. Innovative cooling techniques are employed to handle high-performance FPGA implementations. This XMC daughter card is optimized for deployment on a Curtiss-Wright CHAMP-FX2 FPGA VPX engine, but may also be used on CHAMP-AV6, SVME-184, or VPX6-185 single board computers. Curtiss-Wright Controls Embedded Computing, Leesburg, VA. (703) 779-7800. [www.cwcembedded.com].

34

November 2007


INDUSTRY INSIGHT

wireless infrastructure

Minimizing power consumption in wireless networks involves intelligent synchronization of signaling and communication among nodes and awareness of each node’s power status. by N  iek van Dierdonk GreenPeak

d

exploration er your goal eak directly al page, the resource. chnology, and products

It’s Not About Milliamps Alone when Building an Effective Sense and Control Network

A

sports car or a tractor—which one is solved, and in the technology universe, faster? The sports car is the natural that equates to not solved. answer. Wrong! At least if you take Wireless sensor communications and into account the setting of a muddy, freshly low power go hand in hand. In fact, low ploughed field. The sports car won’t move power is just as important as the reliabila meter. The tractor is the faster vehicle. ity of the communications itself. Before The same is true for low-power wire- the advent of wireless sensor communiless sensor operation. Many have been de- cations, low power was synonymous with signed for low power, meaning that they low current consumption. The lower the consume little power when switched on. milliamp figure, the better the device was That is not enough. Wireless sensor ap- at low-power operation. It was all about panies providing solutions now plications require a different way to opti- how many or rather how few milliamps ration into products, technologies and companies. Whether your goal is to research the latest mize for low power. the electronics consumed. To further relication Engineer, or jump to a company's technical page, the goal of Get Connected is to put you Wireless networks and appli- duce power consumption, when the deice you require for whatever type ofsensor technology, are having an enormous impact vice did not need to communicate, it was ies and productscations you are searching for. on our daily lives. Quickly and inevitably, turned off, to be awakened when an alarm wireless technology is making the world situation was raised or a periodic status more comfortable, safe and environmen- update was called for—a technique known tally friendly. Since its inception, wireless as duty cycling. sensor technology has been inexorably Current consumption continues to linked with low-power electronics. Not be very important in wireless sensor netsurprisingly, when a data cable is replaced works. So not surprisingly, state-of-theby wireless communications, one also art wireless sensor communication comwants to eliminate the power cable. Oth- ponents score well on power consumption erwise, the wiring problem is only half and utilization of wake-up/sleep modes for duty cycling. However, power consumption is only part of the solution. Four other Get Connected factors must also be addressed in order to with companies mentioned in this article. www.rtcmagazine.com/getconnected achieve low power in wireless sensor ap-

End of Article

36

November 2007 Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

plications. These are peak current, graceful power failure, low-power mesh routing and sleep current.

Peak Current

The plot in Figure 1 depicts the current consumption in three typical wireless sensor node states for a commonly used wireless sensor platform. In state one, the microprocessor and transceiver are in sleep mode (10 μA). In state two, the microprocessor is switched on while the transceiver is asleep (10 mA). In state three, both the transceiver and the microprocessor are awake (27 mA). These current draws can be sustained with high-power batteries such as alkaline cells, but typically exceed the tight energy budgets available with small batteries or energy harvested sources (see sidebar “Energy Harvesting—Power Is Where You Find It”). These energy sources share an important feature—they have a hard time generating the peak current needed to awaken the electronics, even if they can cope with the average current consumption throughout the wake-up/sleep cycles. A coin cell battery, for example, has a typical maximum out-


INDUSTRY Insight

put power of 15 milliamps, far below the peak value that most wireless communication systems require. When closely examining the power consumption behavior of electronic circuits, it becomes apparent that what initially looks like a flat current curve actually bears more resemblance to a mountain range with peaks and valleys. When certain functional blocks become active, they draw peak current. When two functional blocks switch on simultaneously, the peak amplitude doubles. The secret to reducing the peak power lies in carefully managing the turn-on and turnoff time for key functions so that double peaks can be avoided. Doing so enables such components to run off a coin cell battery without requiring additional energy buffering components.

Graceful Power Failure

When an energy source has dried out, such as a depleted battery or a solar cell at midnight, the electronics cannot communicate and are dead. This is clearly a situation to be avoided whenever possible. But in many applications, this situation can arise and must be taken into account, either as a normal event (solar cell at midnight) or as an exceptional condition (depleted battery). In both cases, the power problem can be dealt with, provided the application is intelligent enough to detect the upcoming problem before the energy source has completely dried out. During this last breath, the device should perform a number of actions to inform its environment of the situation, transmit some critical data and put itself in a state that allows fast recovery when the power is restored. To accommodate failing low-power energy sources such as batteries and

Transceiver On

27 mA 10 mA ~10 µA

Microprocessor On Sleep Current Ch1 20.0mA Ω

Figure 1

A typical wireless sensor platform had three current consumption states, which can seriously compromise battery life.

solar cells, devices must employ a technique known as “graceful power failure.” During normal operation, the devices carefully monitor the state of the power circuits. As they encounter declining power levels, they raise different levels of alarms ranging from early warning to near-death. The alarms are escalated and communicated to other parts of the system, thereby enabling the system to be placed in a state consistent with the alarm condition.

Low-Power Mesh Routing

One of the most dramatic differences between wireless sensor communications technology and other well-known wireless technologies is the ability of sensor nodes to forward messages from other nodes located further down the communications chain. This technique, known as mesh routing or multi-hop networking, provides an effective and reliable means of spanning large infrastructures, beyond

the range of what a single wireless link can do. For a node to forward a message received from another node, it needs to be in an awake and receiving mode when the original wireless message arrives. Unfortunately, the reception mode requires so much power that it can drain batteries in a matter of a few days. As this is too short for real-life applications, the most straightforward solution, as specified by most industry standards, is to limit the multi-hop capability to the nodes that are permanently connected to the main power. In such a framework, low-power devices, which are assumed to be in a power-down mode most of the time, are not capable of retransmitting messages from other devices. These low-power devices, known as end-devices, are located at the end or beginning of the communications chain. This framework, which combines mains-powered mesh routing devices and low-power end-devices, works for some applications. Take, for example, an office lighting application utilizing interconnected wireless luminaires and light switches. The luminaires, which are connected to the main power source, house the mesh routing communication nodes. The switches, which are not mains powered, are a natural place for the end-devices. Many other applications do not fit well in such a framework. Think of gas detection, fire detection, access control, precision farming, battlefield monitoring, perimeter surveillance and warehouse temperature monitoring. In these applications, mains power is not readily available or even present. Running a power cable in these applications would be cost prohibitive, offsetting the benefit of wireless communication. November 2007

37


INDUSTRY Insight

B

C

A and B communication cycle B wakes up

A awake period

B awake period

A and B communicate

Time

C wakes up

B awake period

C awake period

B and C communicate

ep C goes to sle

B wakes up

B and C communication cycle

ep B goes to sle

November 2007

A

ep B goes to sle

38

the electronic circuits—wireless as well as sensor related. Start-ups and large technology companies alike have embraced the first challenge, the energy conversion stage. The second challenge, the energy source monitoring, management and energy transformation, requires careful consideration of how the energy is “leaked” to the environment. A good example is an outdoor solar cell that generates small bits of energy during the day and then usually shuts down completely during the night. The small energy bits need to be collected and transferred to a useful voltage and current level or the electronics. Moreover, the onboard energy management algorithms carefully measure the state of the energy source to distinguish between healthy, intermediate and dying states. The state is an essential piece of information for the application to take appropriate action, such as setting itself in a fail-safe state. A quite different example is a piezoelectric energy har vester. Such a device can be integrated in a light switch to convert the movement of a switch depression into electric energy. This energy is “bursty” in nature. It requires careful management of the total available energy, because the system has to per form the full communication task with this one-off spike of energy, and with no ability to know when the next push of the switch or “energy reload” will take place. By embedding the energy monitoring and management functionality in the wireless sensor itself, the developer is essentially removing the need for a separate functional block that typically includes a microprocessor and analog components. And in cost-sensitive applications, which include most sensor applications, a higher level of system integration has a direct impact on cost. Conversely, integrating these functionalities within the communication system can result in a lower system cost, enabling new applications that otherwise would be beyond the reach of wireless communication technology.

ep A goes to sle

Energy harvesting, ambient energy sourcing, energy scavenging…all these different terms mean very much the same thing: the ability to extract energy from the environment to drive electronic circuits independent of the mains power or a battery. The “raison d’être” of energy harvesting in wireless sensor applications is straightforward: Removing the sensor’s data cable only solves half of the problem; the installer still needs to run a power wire. And half a problem solved is as good as not solving it at all. Furthermore, replacing the mains power with a battery substitutes the wiring cost with a battery replacement problem, which is not an advance at all. Energy from the environment is plentiful. Solar light, vibration, motion, etc., are generally well understood. Common applications include soil acidity monitoring in precision farming using solar cells and motor monitoring in an industrial plant using vibration collectors. Vibration is already used widely in a well-known consumer application: wristwatches that need neither a battery nor rewinding. But also less obvious sources can generate sufficient energy to power a wireless sensor system. The temperature difference between two touching sur faces, such as a pipe carr ying warm liquids and the surrounding atmosphere on its sur face, can be used to drive a pressure sensor and the wireless communication system. Furthermore, the temperature difference between the human body and its environment enables some ver y interesting applications in health monitoring: A temperature difference of 9°F (5°C) generates sufficient power for realworld applications. Another interesting finding is that indoor lighting can actually be used, for example, to power a wireless light switch. The challenge in energy harvesting lies not in identifying the energy source, but first in efficiently converting the energy source into electric energy, and secondly in carefully managing the health of the energy source and transforming the electric energy in a form appropriate for

Node A wants to ass a message to node C

A wakes up

Energy Harvesting—Power Is Where You Find It

Time

Figure 2

Synchronizing “awake” periods among nodes to overlap at specific moments can greatly reduce the power consumed in the sleep state and prolong battery life.

To address this class of applications, which have been found to be more prevalent than mains-powered, multi-hop applications, requires a totally different framework. In this framework, known as low-power multi-hop networking, or lowpower routing, all of the nodes, including the mesh routing nodes, operate in lowpower mode. The key to this approach, which GreenPeak refers to as “synchronized wake-up,” is to coordinate receiving activity in a way that eliminates the need for the mesh routing nodes to continually operate in receive mode, thereby significantly reducing power consumption. Figure 2 depicts how low-power routing works when Node A wants to send a message to Node C, through Node B. All nodes in the picture are low-power nodes, sleeping most of the time. The breakthrough lies in synchronizing the sleep/wake-up cycles of the nodes to each other. Nodes wake up when they expect a message from a neighboring node. This enables the routing nodes to operate in a nearly powerless sleeping state most of the time, thereby achieving ultra-low-power operation. Clearly, more wake-ups will occur than strictly required to carry the data, as neighboring nodes will not always have data to transmit. However,


08

20

...Your Vehicle to Success in the

Fast Paced

Telecom Market

www.mountainviewalliance.org

The Mountain View Alliance puts you in the drivers’ seat! Now you can become part of this fast-paced 2-day conference designed to accelerate the adoption of open-specication based components and platforms. Take your place in the design and evolution of commercial off-the-shelf implementations in the telecommunication and wireless infrastructure. Set your plans to include a diverse and effective conference that brings company’s and individuals together from the entire COTS ecosystem.

Mountain View Alliance Communications Ecosystem Conference March 11 and 12, 2008 South San Francisco Conference Center

www.mvacec.com


DROWNING IN A SEA OF

INFORMATION?

THIS EVENT IS A

REAL-LIFE SAVER

www.rtecc.com


INDUSTRY Insight

the additional power required for periodic wakeups and synchronization is more than offset by the power saved by eliminating the need for continuous receive mode operation.

Sleep Current

Wireless chips are usually specified according to their power consumption in receive and transmit mode. Remember, however, that in order to achieve low power, the devices must be duty cycled, moving between alternate sleep and awake states. The longer the required battery life, the longer the device sleeps between wake-up periods. Unfortunately, electronic circuits never really “sleep.� Although the powered-down circuits don’t do anything meaningful from a functional standpoint, a small leakage current flows through the transistors. This leakage can amount to several tens of microamps. Sleep current is not usually considered an important design factor, but it be-

comes extremely important when designing a circuit that must live for five years or more on a battery, sleeping most of its life. If the design is not optimized for low leakage current, the majority of the power will be spent on sleeping.

Low-Power and Standards: A Natural Fit?

Wireless sensor applications prosper best within a sphere of industry standards, which gives OEMs the freedom to purchase from a larger pool of suppliers. Most importantly, standards allow devices from different vendors to interoperate, a feature that is paramount in applications ranging from building automation to industrial automation. Current standards within the wireless sensor universe include ZigBee, ISA SP 100, Wireless HART and others. As many wireless sensor applications benefit from low-power operation, it is no surprise that wireless sensor communications standards have predefined fea-

tures to enable low-power operation. For other features, it is left to the technology provider to implement the standard in such a way that it achieves low power. Yet other characteristics go beyond the scope of standards. At GreenPeak, we realize that all three sorts of features (standard defined, implementation specific and beyond the standard) are equally important and deserve an equal amount of design attention. That is why we optimize our devices for low power at every design stage. Our sleep current and peak current technology, for example, is implementation-specific, whereas our low-power mesh routing and graceful power failure technology extends beyond the scope of today’s industry standards. GreenPeak Utrecht, The Netherlands. +31 30 262 1157. [www.greenpeak.com].

RTC?FREEADPDF0-

#

-

9

#-

-9

#9

#-9

+

November 2007

41


industry wat c h

small form factor sig

The time has come for a new set of legacy-friendly small formfactor standards that will give mainstream industrial systems a path forward to adopt these new technologies.

nd

er exploration ther your goal speak directly cal page, the ht resource. echnology, s and products

The Case for Smaller Form-Factors in Industrial Systems by Robert A. Burckle, Small Form Factor SIG

O

ver the last couple of decades, computer electronics has spread rapidly beyond the desktop into embedded applications such as medical, industrial, test & measurement, communications, transportation, military, avionics, kiosks and gaming. At several points along the way, new standards for embedded boards have appeared when breakthrough bus and interconnect technologies have emerged. As the performance density of silicon continues to increase, reaching the 65 nm geometry node, the time is ripe for a new wave of small form-factor standards to usher in the latest high-speed serial technologies and mpanies providing solutions now smaller components.

ploration into products, technologies and companies. Whether your goal is to research the latest pplication Engineer, or jump to a company's technical page, the goal of Get Connected is to put you rvice you requireBeyond for whateverthe type ofBoard technology, anies and products youEspecially are searching in for. long-lifecycle applications, demand for small

form-factor boards and components has recently increased dramatically, fueled by the demand for smaller, more portable systems and devices. These burgeoning requirements for new small form-factor systems—such as battery-powered devices and industrial systems—mean that new sets of specifications are needed that take advantage of small-geometry components and take into account several issues beyond the board itself. In addition to the single board computer (SBC), these specifications must also address multiple, competing system-level considerations of smaller systems operating in often harsh environments. These considerations include more efficient thermal dissipation, I/O expansion and reliable Get interconnects. Connected Current advancesmentioned in semiconductor with companies in this article. technology are enabling www.rtcmagazine.com/getconnected either more performance in the same amount of space, or the same performance in a smaller space with the additional benefits of lower

End of Article

Get Connected with companies mentioned in this article. www.rtcmagazine.com/getconnected

n.

42

November 2007

D

P-BASE

DESKTO 20W

-BASED

NOTEBOOK

10W ULTRA MOBILE

2000 0.25µ

Figure 1

2005 0.18µ

0.13µ

90nm

2010 65nm

45nm

Smaller chip geometries are making possible a new class of lower-power, ultra-mobile compute platforms that prioritize longer battery life and smaller footprint over processor performance.

cost and lower power consumption. In many embedded applications an increase in the performance of individual boards or subsystems is simply not needed. Although the high-performance end of all application areas gets the most attention and is well served by multicore CPUs, many embedded systems require only modest processing performance with less power and cost in a smaller space. At this lower end where processor performance is only a few hundred MHz, the production volumes of systems designed for individual application areas may not be high, but collectively they are significant. These application areas include industrial control, data acquisition, automation and process monitoring. Historically, this market segment has relied on either desktop computers with industrial I/O cards attached, or embedded boards


INDUSTRY Watch designed with desktop or laptop processors and chipsets. This segment is at the mercy of desktop-style ICs, since the volumes of these secondary markets usually don’t justify the development costs, including design and mask sets, of unique ASICs.

Ultra Mobility

Once the power dissipation of desktop processors rose past 20 watts and never looked back, the creation of the mobile (notebook/laptop) class of processors saved this lower-end, secondary market segment and many others like it. Low-voltage (LV) and ultra-low-voltage (ULV) mobile processors sport thermal/design/ power (TDP) ratings of 5W to 15W and allow fanless reliability in embedded applications. Over time, though, the once lean-and-mean chips designed for mobile systems have gradually become dual-core 20-30W CPUs and hot, high-performance chipsets. Once again, the industrial market is being saved by actions in the consumer and enterprise spaces, made possible by 90 nm and 65 nm process technologies. The Ultra-Mobile PC runs an entire notebook-class operating system and full-fledged applications, including Web browsers. This emerging category of handheld systems prioritizes longer battery life and smaller footprint over performance (Figure 1). In 2008 and beyond, the smaller, lower-power CPUs and chipsets from this new primary market will lead to a plethora of singlecore, low-power, portable and AC-powered embedded systems in the 500 MHz to 1 GHz range. Here, the modular approach is often used, and the good news is that this trend will continue. In automation and process control, for example, multiple smaller, lower-performance CPU boards or modules are networked together in a distributed system, where not only built-in connectivity but also lower power and cost are vital.

systems, chassis designs must also take these into consideration, including convection-cooling or fan-cooling methods. For example, in upgrading a portable medical system, designers must consider how boards are plugged into each other and how they relate to the display, as well as other considerations of how everything is positioned in three dimensions within the chassis: where connectors, display cables and power supplies will be located, as well as how the SBC interfaces to the display.

New Technologies Enable Smaller Form-Factors

Among the new technologies available to long-lifecycle system and device manufacturers of smaller systems are chips based on 90 nm and 65 nm processes. This silicon includes lower-power and highly integrated x86-based processors—many with self-adjusting speed and power management options or in fanless ULV versions—as well as low-power core logic chipsets and memory that are enabling sub-10W design platforms for computing and communications systems. For example, board designers can now use the 90 nm 500 MHz Eden ULV processor from VIA Technologies, with a maximum operating power of just 1W and idle power as low as 100 mW (Figure 2). Even the storage components of these systems must be smaller and more thermally efficient, whether they utilize solidstate storage (SSD) devices or 2.5-in. and 1.8-in. rotating drives with the space-efficient Serial ATA (SATA) interface. For example, there is a variety of embeddable USB-compatible devices and vibration-tolerant disk drives for rugged applications already available. In addition, higher-density connectors from Samtec and others feature improvements for ruggedness, and high-speed serial interfaces for board-to-board interconnects are available to replace

Legacy-Friendly

When the need to upgrade legacy systems is included as a consideration, the situation becomes even more complex. A system designed for long-lifecycle applications may be upgraded several times throughout its lifetime to increase functionality and/or performance. Especially in industrial applications, there are often tighter space constraints in systems that must be fanless for the purposes of high reliability and maximum uptime, but where low cost and modest performance are also important. New chipsets feature high-speed serial interfaces instead of the low-speed or parallel ones of the past such as ISA, EIDE, COM ports and even PCI, all widely used in industrial applications. Any upgrade to these systems must take into account existing designs and their needs in terms of functionality, interfaces and three-dimensional constraints. A “legacy-friendly” approach focuses on methods that let designers continue to use legacy peripherals and smoothly transition to their replacements over time, rather than forcing multiple, widespread re-designs all at once. Since many embedded systems and subsystems are physically located in space-constrained environments, often within entirely different fittings, the chassis’s form-factor may be dictated by considerations other than the operation of the system itself, such as pre-existing, non-electronic form-factors or the need to fit into an opening where mechanical components reside. Because thermal constraints are a major factor in the design of small embedded

Eden Processor CRT Monitor HDTV output

DDR 400/DDR2 533 Up to two DIMMs

Video Inputs (CCIR656 / TS)

LVDS/DVI panel

CX700M IGP Chipset PATA

8-Channel Audio

VT1708 HD Audio Codec

1 EIDE channel Up to 2 devices

SATA DriveStation

Vinyl HD Audio

4 PCI devices

6X USB USB 2.0

Connectivity

PCI Bus

LPC Keyboard / Mouse

EPROM Serial / IR

VT1211 Parallel LPC Super Floppy I/O

Figure 2

An entire embedded computing and communications system with maximum system power of under 10W can be built using highly integrated CPUs and core logic silicon based on the latest 90 nm process technology, such as VIA Technologies’ 500 MHz VIA Eden ULV processor and the VIA CX700M single-chip system media processor. November 2007

43


INDUSTRY Watch slower, space-consuming parallel interfaces. These newer interfaces include PCI Express, SATA, I²C and Low Pin Count (LPC) Bus, which is the replacement for the ISA bus in desktop PC architecture. Gigabit Ethernet and USB 2.0 are the key subsystem- and system-level interconnects, and are becoming increasingly prevalent on board processors. Newer board and module formats include SBCs, computers-on-module (COMs) and stackables that bring these serial interconnects to carrier boards or I/O expansion modules. Especially as boards and systems get smaller and as the number of small form-factors proliferates, it has become more important to focus on robust, high-speed connector technologies than on merely the details of board sizes. The emphasis is on precisely fitting enclosures with a minimum of cabling, which reduces manufacturing costs, a major area of concern. However, the use of off-the-shelf components to speed time-to-market—another major OEM concern—dictates the creation and consolidation of board outline sizes, since it is not easy to design enclosure mountings for different-sized boards with connectors located in different places.

A New Small Form Factor Consortium

In order to make all of this work well, the industry needs standards. A new standards group focused on small form-factors, named the Small Form Factor Special Interest Group (SFF SIG), was formed in September. This consortium has taken as its charter the development, adoption and promotion of specifications for circuit boards and related technologies that will help electronics equipment manufacturers and integrators reduce the overall size of their next-generation systems. The SFF SIG was formed in order to address the broad scope of new market needs, as well as existing

44Untitled-1

1 November 2007

specifications that are not yet being managed by a trade group. Consolidating suppliers around standards, facilitating crossplatform interoperability and developing common expansion schemes are all goals of the SFF SIG. As history indicates, standards with rich “ecosystems” endure the test of time. The SFF SIG also appreciates the extensive investments in off-the-shelf products made by system OEMs and seeks to preserve those investments. The SIG’s philosophy is thus to embrace the latest technologies, as well as maintain legacy compatibility and enable transition solutions to next-generation interfaces. It seeks to enable practical, mainstream, real-world applications rather than extremely highperformance and high-power dissipation systems. To that end, the SFF SIG has formed three working groups to address product categories that are key to the development of next-generation smaller systems. The SBC Working Group is discussing new small form-factor SBCs. The Modules Working Group is developing a specification for a new, small computeron-module (COM) form-factor. The Stackables Working Group is developing a legacy-friendly stackable interconnect technology that will apply to a number of existing SBCs for smoothly integrating new, high-speed, serial technologies into legacy systems to preserve OEM investments in I/O, cabling and enclosure designs. In addition, the SFF SIG’s Board reviews specifications created outside the organization and submitted by vendors who wish to establish true, open, managed SFF standards. Small Form Factor SIG, Santa Clara, CA. (408) 480-7900. [www.sff-sig.org].

10/24/07 2:23:27 PM


What’s in the Embedded Industry’s Basement? FEATRURED PRODUCTS

Quad port Gigabit Ethernet copper PCI adapter 115 remaining

$299

Make An Offer!

Dual port Ethernet copper PCI adapter 105 remaining

$199

Make An Offer!

Single port Gigabit Ethernet copper PCI adapter 60 remaining

$149

Make An Offer!

Dual PCI-X SCSI Ultra-3 LVD host adapter 65 remaining

Dual PCI SCSI Ultra-3 LVD host adapter 65 remaining

Make An Offer!

Make An Offer!

$249

$249

Purchase excess inventory at clearance prices!

NEWEST

PRODUCTS

19” 8U 1U 1024X768 Rackmount Monitor, 20.5” Deep aRGB, VGA, DVI-D, 15” TFT/LCD S-Video, CVS 1 remaining 1280x1024 $1299 2 remaining Make An Offer!

$2143

Make An Offer!

17” 8U Rackmount Monitor, aRGB, VGA, 1280x1024 2 remaining

$1262

Make An Offer!

1600X1200 8RU 1RU Keyboard 20.1” Panel Shelf with Cherry Mount TFT/LCD PS/2 1 remaining Keyboard/Trackball Combo $2993 2 remaining Make An Offer!

$308

Make An Offer!

Don’t miss out on these great products www.aptyc.com


&TECHNOLOGY

Products

VME64x PowerPC SBC Improves Efficiency Advantech’s New Versatile and UltraCompact Embedded Computers

A series of small, powerful all-inone fanless systems is designed for rugged and space-critical applications in automation, control and wireless gateway applications. A sealed aluminum case provides vibration and water resistance, and also functions as a passive cooling solution. The ARK-1000 compact embedded computers from Advantech feature AMD Geode to Intel Core Duo processors, along with flexible storage and connectivity, from multiple gigabit Ethernet and COM ports to wireless connectivity and high-end graphics subsystems—all with flexible I/O configurations and long-life product support. The ARK-1000 systems can be mounted inside machinery or other equipment. They can also be stand-alone, wall-mounted, DIN-rail-mounted or VESAmounted. The series accepts a wide range of power supplies (DC power in) and comes in a footprint of 7.44” x 1.61” x 5.37”. The rugged cast aluminum case not only provides protection from EMI, shock/vibration, dust, cold and heat, but also offers a passive cooling solution for quiet fanless operation. The ARK-1370, for example, offers four USB 2.0 ports, one LVDS, one LAN and two COM ports and is powered by an AMD Geode processor. Wireless connectivity can be added via the PC card slot for PCMCIA or CardBus cards. It also supports a wide range of input voltages from 9 VDC to 35 VDC. The ARK-1370 is equipped with a solid-state onboard CF card of up to 4 Gbytes, so it easily passes 50 and 5 Grms shock and vibration tests. The ARK1370 is the choice for applications such as signage and HMI, where compact size and low power consumption are important. It also performs well for vehicle applications because of its shock and vibration resistance, as well as its tolerance to a wide range of power inputs.

A new VME64x PowerPC single board computer board incorporates a newer hostbridge controller and DDR2 memory technology to give new life to COTS VME64x single board computers still widely used in real-time/ embedded military and aerospace programs. The Celero CVME-7448ST from Cornet Technology uses the latest host-bridge controller technology, the Tundra Tsi109, which lets the CVME-7448ST offer much higher performance per watt—a key consideration of military embedded applications with tight power budgets, and improved signal integrity providing significant improvement over older generations of VME PowerPC single board computers. Incorporating DDR2 technology and the Tundra controller eliminates memory latency caused by pipeline blockage, improving performance by up to 9% and reducing overall power consumption by up to 6W over boards using older technology.

Advantech, Irvine, CA. (949) 789-7178. [www.advantech.com].

Managed Fiber Ethernet Switch Supports SNMP

Everything over IP (EoIP) means a growing demand for Ethernet network switch gear. Aaxeon Technologies has released its Lanolinx line of Fiber Ethernet Switches. The Lanolinx Fiber Ethernet Switches include 100FX and Gigabit Fiber models, supporting multimode or single-mode fiber. These switches also have up to 24 10/100 Copper ports. In addition, models are available with two fiber ports instead of one. The units support IEEE 802.3/802.3u, store-and-Foreword switching and IP Security. The 10/100, Full/Half Duplex switches have MDI/MDI-X auto-sensing, Rate Limiting, and support both SNMP/Telnet/Console/Web management and port-based VLAN / 802.1 Q Tag VLAN. Offered in a rack-mountable enclosure, the switches provide IEEE 802.1p Class of Service and Port base, Tag base and Type of Service priority method. IGMP with Query mode for multimedia and port mirroring are supported as well. Pricing for the 24-Port SNMP Managed Fiber Switch starts as low as $289. Aaxeon Technologies, Brea, CA. (714) 671-9000. [www.aaxeon.com].

46

November 2007

The CVME-7448ST is targeted at such applications as signal intelligence applications (such as image, radar and sonar processing) that require high performance, low power consumption and superior reliability. Since the board supports both Linux and VxWorks 6.3 operating systems, embedded designers can easily port their existing algorithms and code. The board offers 512 Mbyte DDR2 SDRAM, 256 Mbyte flash, two PMC sites for I/O expansion, two high-speed Gigabit Ethernet ports, two serial ports and two USB 2.0 ports. Price starts at $4,600. An extended temperature version is also available. Cornet Technology Springfield, VA. (703) 658-3400. [www.cornet.com].


5-Port Gbit Ethernet Switch Operates at -10° to 60°C

Full Motion MPEG4 Server Brings Video to the IP Network

A network video server provides gateway functions to convert analog video signals to digital video streams over the IP network. These functions greatly expand the application range of the server into diverse vertical markets. The DVG-1100 from Advantech is a single channel, full motion MPEG4 video server. Powered by a hardware MPEG4 encoder and a RISC processor, the DVG-1100 provides video quality and stable performance. With mature development on the network infrastructure and more devices connected to the IP network, it becomes quite simple to integrate video into networks with management applications. The DVG-1100 supports D1 (720x480 at 30 fps – NTSC or 720x576 at 25 fps – PAL) video compression and streaming. With built-in 2D hardware de-interlacing, video quality is greatly improved while analog signals are instantly converted to digital formats. The audio encoder offers support for applications requiring on-site audio monitoring. In addition to audio and video streaming, the DVG-1100 also incorporates three isolated digital inputs, one relay output and one half-duplex RS-485. All of these signals are transferred through the IP network and can be easily controlled from the backend. The capacity for multi-focal motion detection is built-in, making it possible to set up applications that require sensitive visual sensors for complex monitoring processes. The industrial-grade design of the aluminum enclosure along with the DIN rail kits make installation a quick and easy process. Three different types of software support are available. First, the built-in Web server can be used for simple IE browser access and system configuration. An integral, 16-channel PC recording management software is bundled with the unit to provide a second option for standard video surveillance applications. Lastly, a complete set of driver/ SDK or OCX components with C++ sample code helps the developer successfully integrate the DVG-1100 into their unique applications. Advantech Irvine, CA. (949) 789-7178. [www.advantech.com].

With its latest rugged Ethernet product, the eAutomation Group of Advantech has introduced the EKI-2725 5-port unmanaged industrial Ethernet switch featuring full Gbit capacity for bandwidth-intensive applications. The EKI-2725 is equipped with five 1000Base-T Gbit Ethernet ports and supports 9k Jumbo Frames, which allows larger blocks of data to be sent with each transmission, for increased network throughput and reduced CPU utilization of connected devices. Designed for the harsh environment of industrial applications, the EKI-2725 is packed in a compact metal DIN rail-mount chassis built to withstand shock and vibration. The EKI-2725 features 4,000 VDC Ethernet ESD protection, power line surge (EFT) protection of 3,000 VDC and an operating temperature of -10° to 60°C. The EKI-2725 includes dual 12 to 48 VDC power inputs for redundant power configurations to ensure system uptime. In the event of a problem with external power, a built-in fault relay is available to signal an external programmable logic controller (PLC). The EKI-2725 5-port Gbit Ethernet switch is priced starting at $275. Advantech, eAutomation Group Cincinnati, OH. (800) 205-7940. [www.eAutomationPro.com].

LXI Triggering Box Does Precise Synchronization

LXI, the LAN-based successor to GPIB, offers flexible packaging, high-speed I/O and prolific use of LAN across a broad range of aerospace and military applications. Agilent Technologies has introduced the world’s first LXI trigger box that enables precise synchronization over LAN for LXI Class C and GPIB instruments, elevating their performance to LXI Class B standards. When an LXI Class C or GPIB instrument is connected to the Agilent E5818A LXI trigger box, it gains the timing capabilities of an LXI Class B instrument. Leveraging the IEEE 1588 precision time protocol (PTP) synchronization, the trigger box enables sub-nanosecond time triggering and time stamping of events for the attached instruments. With reliable event-log data, users can trace and troubleshoot faults easily. The Agilent E5818A LXI trigger box is a stand-alone LXI Class B device. It can achieve a synchronization accuracy of up to 13 ns (standard deviation over direct connection) and provide time stamping of up to 5,000 events. Each trigger box provides BNC connectivity to two instruments (any combination of GPIB or LXI Class C). The Agilent E5818A LXI trigger box is available now and is priced at $1,500. Agilent Technologies Santa Clara , CA. (877) 424-4536. [www.agilent.com]. November 2007

47


Products&TECHNOLOGY Dual Dual-Core Intel Xeons Power New VME SBC

A new dual-processor 6U VME64/ VME320 single board computer features two 1.66 GHz Dual-Core Intel Xeon ULV processors, effectively providing four processor cores. The VP 426/23x from Concurrent Technologies provides up to 8 Gbytes of onboard DDR2 ECC dual channel SDRAM and features a variety of I/O interfaces, which can be further expanded via the board’s PMC/XMC site, while still maintaining a single slot solution. The Intel E7520 server class chipset and Intel 6300ESB ICH are used to complement the two Dual-Core Xeon processors to achieve a high-performance, yet low-power, dual processor dualcore architecture. Each Xeon interfaces to the E7520 via a 667 MHz Front Side Bus. The E7520 chipset can access up to 8 Gbytes DDR2-400 ECC dual channel SDRAM at up to 6.4 Gbyte/s —and supports up to 4 Gbytes soldered and up to 4 Gbytes SODIMM, all in a single slot. A PMC/XMC site supports both front and rear I/O; there are two SATA150 interfaces, two graphics interfaces, four Gigabit Ethernet interfaces, making the VP 426/23x suited to intensive data-transfer as well as data-processing applications. The PMC site supports up to 66 MHz PCI operation and also supports an x8 PCI Express XMC interface. The VP 426/23x also supports the VME320 (2eSSt) protocol. For fast transfer of system-wide data, the four channels of Gigabit Ethernet can sustain full-duplex Gigabit data rates as each dual channel Ethernet Controller is connected to its own x4 PCI Express link. For ease of system integration the two rear Ethernet channels can be configured to connect to a VITA 31.1 Gigabit Ethernet backplane or, alternatively, connect to the board’s optional rear transition module. A wide range of onboard I/O is available: the front panel supports dual Gigabit Ethernet, a USB 2.0, RS-232, analog graphics, keyboard and mouse interfaces. As well as the PMC rear I/O and dual Gigabit Ethernet, the rear panel supports three more USB 2.0, two more RS-232, two SATA150, two EIDE, digital graphics, keyboard and mouse interfaces. Other features included are a watchdog timer, LAN boot firmware and options for an onboard 2.5inch SATA150 (or EIDE) hard disk drive or CompactFlash EIDE storage. For embedded systems, an optional fixed USB flash drive is available on the rear transition module. The VP 426/23x family supports many of today’s leading operating systems, including Linux, Windows Server 2003, Windows 2000, Windows XP, Solaris, LynxOS and QNX. Concurrent Technologies, Woburn, MA. (781) 933 5900. [www.gocct.com].

16-Channel Constant-Current LED Driver for Large LED Video Displays

A new family of devices from Catalyst Semiconductor addresses large LED video displays. The new CAT4016 is a 16-channel, constant-current sink LED driver designed for billboard, marquee, instrument and other larger, general-purpose LED displays in both indoor and outdoor applications. The CAT4016 LED channels can operate with an output voltage as low as 0.4V (for 2 mA to 100 mA LED current), to enable more power-efficient designs. The maximum LED current is set by an external resistor. A high-speed, four-wire serial interface with up to 25 MHz clock frequency, controls each individual channel using a shift register and latch configuration. A serial output data pin allows multiple devices to be cascaded and programmed via one serial interface. The CAT4016 also features a thermal shutdown protection function to disable the LED outputs if the die temperature exceeds a set limit. The CAT4016 is available in a wide range of package options including 24-lead SOIC, TSSOP and QSOP. Pricing is $0.54 each in 10,000 piece quantities. Samples are available now. Projected lead-time for production quantities is currently 6 to 8 weeks ARO. Catalyst Semiconductor, Santa Clara, CA. (408) 542-1000. [www.catsemi.com].

48

November 2007

High-Speed Link U Series Modules Offer Compact I/O Options

A series of digital input/output slave modules provides a selection of compact distributed I/O modules with connector options. Each of the three modules in the HSL U series from Adlink Technology features an output current of up to 90 mA/ch and improves on previous designs by encompassing a flat-bottom housing for tight spaces and by offering I/O connector options.

The HSL-DI16DO16-UJ-N/P is an HSL distributed slave module with 16 discrete input channels and 16 discrete output channels. Separate I/O connectors are provided to allow debugging without affecting other I/O channels. The HSL-DI16DO16-US-N/P is also an HSL distributed slave module with 16 discrete input channels and 16 discrete output channels. However, this model offers shrouded connectors featuring a latch-lock to prevent wires from vibrating loose during machine operation. The HSL-DI16-UL is an HSL distributed slave module with 16 discrete input channels and a pulse stretcher function. The pulse stretcher, ranging from 1 to 127 ms, helps latch on to fast-switched input signals while maintaining user-defined value states. This feature prevents the loss of input signals, making this module ideal for high UPH machines utilizing distributed solutions. The HSL U series of modules is priced starting at $140, and are available with discounts in volume. Adlink, Irvine, CA. (970) 377-0385. [www.adlink.com].


Development Environment for Infineon’s New XC2300 Family

Now supporting Infineon’s new scalable XC2300 microcontroller family is an integrated development environment from Hitex, which consists of editor, compiler, debugger and unit-test tool. The heart of this development solution is the TantinoXC JTAG-debugger, which provides complete run-control functionality such as breakpoints, single-step, register and memory view as well as programming of internal and external flash. The XC2300 family provides 32-bit performance and a rich peripheral set.

Untitled-4 1

10/8/07 11:51:05 AM

In particular, the “entry level” TantinoXC JTAG-debugger can be expanded with additional features approaching the functionality of a “traditional” emulator. An optional time-analyzer function specifically designed for customers in the automotive area allows measuring and analyzing of run-time behavior. In addition, a detailed trace and performance analyzer means that specific trace cases can be defined and recorded. The TantinoXC continuously reads out the trace data almost without real-time violation and files it into the trace buffer. Also, Tantino systems that are already in use for XC166 development can be easily adapted to the new XC2300 family by a license upgrade. The TantinoXC is controlled by HiTOP, the universal user interface allowing complete high-level language debugging and the simple integration of additional tools. Attractively priced bundles consisting of debugger and compiler are available directly from Hitex. Hitex Development Tool Irvine, CA. (949) 863-0320. [www.hitex.com].

Untitled-6 1

November 2007

49

10/8/07 1:16:08 PM


Products&TECHNOLOGY CompactPCI Chassis with Built-in System Monitoring

A new Type 39 CompactPCI enclosure includes built-in system monitoring. With inherent redundancy, hot-swap ability and monitoring, the Type 39c chassis from Elma Electronic offers a powerful solution for reliability. The new chassis features a 3U 4HP configurable system monitor with Ethernet, RS-232, Web and command line interface (CLI) capabilities. The unit monitors up to 8 voltages, 12 fan-fail signals, 8 temperature sensors, and controls the speed of the fans depending on the temperature buildup. The chassis comes with a CompactPCI or PICMG 2.16-compliant backplane in up to 8 slots. The backplanes have three pluggable 47-pin connectors for hot swapping power supplies for a 500W, N+1 solution. With a sheet metal design and full plug ability, the Type 39c chassis family offers ease of manufacturing and saves costs. The chassis are available in 1U-4U heights in horizontal-mounting orientations. Compliant to the latest PICMG specifications and IEEE 1101.10/.11, the enclosures feature side-to-side 200 CFM (cubic feet per minute) and 300 LFM (linear feet per minute) cooling, 300 mm depths and rear I/O options. Elma specializes in quick and affordable customization of their chassis offerings. Elma’s Type 39c chassis family starts under $700 depending on configurations and options, with a 4-6 week lead time.

3U VPX Card Sports Dual Core PowerPC 8641D

A 3U VPX single board computer is the first 3U VPX single board computer to feature Freescale’s dual core PowerPC 8641D and to provide support for a PMC/XMC site. With up to 1 Gbyte of SDRAM, the SBC310 from GE Fanuc Embedded Systems is the tenth member of the GE Fanuc Embedded Systems family of VPX products. The SBC310 is optionally available in any of five ruggedization levels, making it suitable for deployment in the harshest environments.

Elma Electronic, Fremont, CA (510) 656-3400. [www.elma.com].

Camera Link Mezzanine Board for StreamStor Amazon Digital Data Recorders

A Camera Link mezzanine board for the StreamStor Amazon highspeed recording system from Conduant provides connectivity for 1-2 cameras at data rates up to 500 Megabytes per second. Combined with the direct to disk capability of the StreamStor Amazon recording system, the CL-500 provides high speed and long duration recording capability required when shooting high resolution and/or high frame rate digital video. The high-speed capability of the StreamStor system is especially important when the application requires the use of uncompressed video. The Camera Link mezzanine board provides support for nearly any Camera Link full, medium or base configuration. Featuring the fastest available Channel Link components and programmable logic, the CL-500 provides a flexible and highly customizable platform for digital video recording applications. When combined with Conduant’s StreamStor Amazon SATA Disk Controllers, the system can reliably record high frame rate and/or high resolution video for many hours. The StreamStor CL-500 Camera Link board is priced at under $4,000 and is an add-on to Conduant’s Amazon SATA Disk Controller. Conduant, Longmont, CO. (303) 485-2721. [www.conduant.com].

50

November 2007

The SBC310 features two 4-lane PCI Express ports across the backplane, providing high-speed data communication to other system elements. It also provides a number of high-speed interfaces for off-board communication, including two 10/100/1000 Base-T Gbit Ethernet ports, two USB 2.0 ports and a Serial ATA disk interface. In addition, there is support for legacy interfaces, including two RS232 ports and four general-purpose I/O lines. The SBC310 utilizes state-of-the-art cooling technology to allow a PMC or XMC to be fitted without compromising processor node performance. GE Fanuc Embedded Systems Charlottesville, VA. (800) 368-2738. [www.gefanucembedded.com].


Configurable Industrial PC Offers DIN Rail or Wall Mounting

Supplied in a robust CE enclosure that is designed for DIN rail or wall mounting, a new embedded computer features fanless operation, simplicity and flexibility of configuration. The RD-103 from Diamond Point can integrate a range of processors from the cost-effective 800 MHz Pentium Celeron up to the 1.8 GHz Pentium M with 2 Mbytes of level 2 cache memory. Intel Core 2 Duo support will be available shortly. The RD-103 supports up to 1 Gbyte of DDR-SDRAM via a SODIMM socket and integrates an Intel 82855GME 400 MHz FSB chipset and an Intel Extreme Graphics 2 video controller with up to 32 Mbyte UMA Video RAM providing WUXGA/ QXGA (2048 x 1536) resolution with 2D and 3D acceleration, and a CRT port. Available with a range of 1 Gbyte to 64 Gbyte 2.5-inch industrialsolidstate drives or 120 Gbyte rotating hard disk memory as standard, the system may be supplied with higher capacity solid-state drives to special order. A CompactFlash socket is also incorporated. The RD-103 integrates a host of interfaces, including four serial ports, optionally configured as either all RS232 or two RS422/two RS232, three USB 2.0 EHCI ports, one 10/100 BaseT Ethernet port, one Enhanced parallel port and a PS/2 keyboard/mouse port. A further USB port has front panel access. A GPS input is provided for applications requiring either accurate positioning or timing functions. Expansion slots for three PC/104 or PC/104+ I/O cards are incorporated and a wide range of optional third-party approved I/O cards are also available upon request, including ProfiBus, CAN Bus, BitBus, 8-port serial RS232/RS422/ RS485, digital and analog, single channel dual redundant 1553, 8-channel ARINC 429 and quad frame grabber. The unit offers a 0° to 60°C operating temperature range, a 5% to 95% (non-condensing) operating humidity range and is CE/EMI tested. A DIP brazed chassis with Mil38999 connectors is optionally available for military/aerospace applications. All system components are manufactured lead-free and are RoHS compliant. The RD-103 comes pre-loaded with either Windows XP or Linux operating systems and can be specially configured for Windows XP Embedded, LynxOS and VX-Works. Diamond Point International, Rochester, Kent, UK. +42 01634 300900. [www.dpie.com].

Gigabit Ethernet Switch/ Router Climbs Aboard 6U VPX

VPX, the fabric-based next-gen VME form-factor, is gaining more and more momentum every month. A variety of SBC products have emerged, and now special function boards like Ethernet switch boards are adding to the VPX ecosystem. For its part, Curtiss-Wright Controls Embedded Computing has announced the first high-density 6U VPX Gbit Ethernet multilayer switch/router board designed for rugged embedded aerospace and defense applications. The new VPX6-684 FireBlade II, available with 12, 20 or 24 Gbit Ethernet ports and up to 4x10 Gbit Ethernet ports, is ideal for system integrators architecting secure high-performance IPv4/v6 Intra-Platform Networks (IPNs). The board, which operates as either a fully managed or an unmanaged switch/router, provides significant performance and configuration advantages to developers building Layer 2 or Layer 2/3+ networks. Additional feature enhancements include support for routing up to 4x10 Gbit Ethernet to the FireBlade’s P1 connector, and support for copper interfaces to the backplane for all of the board’s 12, 20 or 24 Gbit Ethernet ports. Curtiss-Wright Controls Embedded Computing, Leesburg, VA. (703) 779-7800. [www.cwcembedded.com].

PCI Express Board Serves up 16 RS-232 Ports

A new 16-port RS-232 serial I/O adapter is designed for the PCI Express bus design. The COMM+16 PCIe from Sealevel Systems uses 16C854 UARTs with 128-byte FIFOs—16 times larger than other boards. Each serial port provides a maximum data rate of 460.8 Kbits/s, and the COMM+16.PCIe has an operating range from 0° to +70°C. Extended temperature versions operating from -40° to +85°C are available. All Sealevel I/O products have a lifetime warranty. The board is suitable for connecting to PLCs, bar code readers, scales and other data acquisition/control devices using the included DB25M fan-out cable (DB-9M fan-out cable available as an option). The product includes SeaCOMsoftware for Windows 98/ME/NT/2000/XP/ Vista and Linux operating systems. As an added value, customers also receive WinSSD, a full-featured application for testing and diagnostics including BERT (Bit Error Rate Testing), throughput monitoring, loopback tests and test pattern message transmissions. The COMM+16.PCIe standard price is $679 and product is available for shipping. Sealevel Systems, Liberty, SC. (864) 843-4343. [www.sealevel.com]. November 2007

51


Products&TECHNOLOGY Dual Channel Digitizer Features 95 dB SFDR and Virtex-4 FPGA A platform that enables application-specific I/O functions minus the expense of custom hardware development features a high-performance front-end tightly coupled to a Xilinx Virtex-4 FPGA. The FPGA communicates with the host processor through a dedicated PCI/PCI-X bridge, leaving the majority of logic uncommitted. Simple interfaces to the I/O, SRAM and local bus are easily integrated with user application logic. The Channel Accelerator Plus 16/130 dual channel receiver from Red Rapids is based on the Linear Technology LTC2208 16-bit A/D converter. The sample clock is supplied by either the onboard frequency synthesizer or an external source. The frequency synthesizer is phase locked to a local 10 MHz TCXO’ an external reference can be provided to achieve system-wide phase coherence. The analog inputs can be either AC- or DC-coupled to the A/D converters. The AC-coupled configuration supports direct IF sampling (bandpass sampling) beyond the first Nyquist zone. The FPGA can be selected from the Virtex-4 high-performance logic (LX) or signal processing (SX) platforms. A variety of size and speed grade options are offered to further optimize the price/performance ratio over a wide range of applications. The SX-55 device includes 512 DSP slices for math-intensive applications. The FPGA is connected to four 16-bit QDR SRAMs for high-speed local data storage. The QDR SRAM provides separate read and write ports to achieve a combined 8 Gbyte/s data transfer rate between the FPGA and memory. This memory can also be used as a high-speed snapshot recorder to store segments of data without interruption from PCI bus traffic. A dedicated trigger input is provided to synchronize data collection to an external timing strobe. An additional Micro-D connector supplies eighteen bidirectional general-purpose I/O links for digital control signals. A DMA FPGA core provided with the product manages data transfers between the Channel Accelerator and host memory. The DMA engine allows the receiver to automatically initiate a PCI/PCI-X burst transaction when data is available. DMA chaining and scatter-gather techniques are supported by both the hardware and software to optimize data transfer efficiency. Channel Accelerator is available in PMC, PCI and cPCI form-factors. Single unit pricing starts at $4,990. Red Rapids, Richardson, TX. (972) 671-9570. [www.redrapids.com].

Embedded Networking Gateway Supports Ethernet and Wi-Fi

A pluggable extension for the DIL/ NetPC ADNP/9200 computer from SSV Embedded Systems integrates this compact computer module into IEEE 802.11 WiFi networks. The E2W/ESL1 expansion is attached directly to the DIL/NetPC, so the ADNP/9200 provides an ideal solution for companies looking to network-enable a variety of machines and appliances with Ethernet-LAN and wireless. The E2W/ESL1 expansion works in a transparent bridge mode and supports the Infrastructure or Ad-hoc mode according to IEEE 802.11b (11 Mbits/s) and IEEE 802.11g (54 Mbits/s). The ADNP/9200 has two completely independent 10/100 Mbit/s LAN interfaces with separate MAC and PHY-units. Both interfaces are controlled by a 180 MHz ARM9 processor. Memory consists of 32 Mbyte flash and 64 Mbyte SDRAM. The flash includes Embedded Linux with the appropriate drivers, a complete TCP/IP protocol stack and different servers. In addition to the two Ethernet interfaces, there are four UARTs, three USB ports (2 x host, 1 x device), one 16-bit bus interface and numerous GPIOs available. With the ability to add software components and to execute them under the embedded Linux of the ADNP/9200, this platform is ideal to act as a wireless gateway for industrial Ethernet networks. The E2W/ESL1 is an attractive solution for applications where both wired network connection as well as device mobility is important. With the DIL/NetPC-specific long-time availability in mind, the ADNP/9200 in combination with the E2W/ESL1 expansion lets OEMs easily add wireless connectivity to their products. SSV Embedded Systems, Hanover, Germany. +49 (511) 40 00 045. [www.ssv-embedded.de].

52

November 2007

PIC24 Microcontrollers Get ThreadX RTOS Support

Demand is on the rise for sophisticated multithreaded software on ultra small form-factor platforms. Taking aim at that arena, Express Logic has made available Express Logic’s new ThreadX Microcontroller Edition RTOS for the 16-bit PIC24 microcontrollers (MCUs) and dsPIC digital signal controllers (DSCs) from Microchip Technology. Until now, ThreadX has been available only for 32-bit processors and DSPs. ThreadX/MCU is a fully upward-compatible, reduced-size, fine-tuned 16-bit edition of Express Logic’s popular 32-bit ThreadX RTOS and is available exclusively for Microchip’s 16-bit PIC24 MCUs and dsPIC30/33 DSCs. ThreadX/MCU provides full preemptive scheduling, interrupt management, message passing, thread synchronization, resource locking, event management and timer control; and supports up to 10 threads, queues, timers, mutexes, event flag groups, block pools and byte pools. ThreadX/MCU is designed to work with Microchip’s powerful and affordable MPLAB development tools. ThreadX/MCU, NetX/MCU, FileX/MCU and PEGX/MCU are available immediately in full production licenses beginning at $5,990. Demo licenses are available free of charge. Express Logic, San Diego, CA. (858) 613-6640.[www.expresslogic.com].


Expandable Rugged Box Can Take the Heat

DASYLab Graphical Daq Package now Supports Serial and ARINC-429 Boards

United Electronic Industries (UEI) is pleased to announce its DASYLab, an extremely easy-to-use, yet powerful data acquisition and control package developed by measX GmbH & Co of Germany. DASYLab now supports the DNA-SL-501 serial (RS-232/422/485) I/O board and DNA-ARINC-5xx series ARINC-429 interfaces from United Electronic Industries (UEI). The new serial and ARINC-429 support, combined with UEI’s existing support for its analog, digital and counter/ timer interfaces, offers scientists and engineers wishing to take advantage of the ability to not only monitor and control analog and digital signals, but also to interface directly with avionics equipment over the ARINC-429 bus. The addition of serial I/O support allows the user to interface to a wide assortment of instrumentation supporting RS-232, 422 or 485 as well as control or monitor a host of other computer peripherals. The DNA-SL-501 is a 4-channel fully isolated RS232/422/485 interface for UEI’s PowerDNA Cube series. It offers sample data transfer rates up to 1 Mbit/s. The DNA-429512 and DNA-429-566 are 12-channel ARINC-429 interface boards with 12 Rx channels or 6 TX / 6 RX channels respectively. ARINC-429 is the primary avionics bus on virtually all currently flying commercial and general aviation aircraft. PowerDNA (Distributed Networked Automation)—is a compact (4 x 4 x 4” or 4 x 4 x 5.8”) Ethernet-based data acquisition and control device. Each PowerDNA Cube consists of a core module (that holds the processor and network interface) along with three or six open I/O slots or layers. Users configure their system by selecting appropriate I/O boards from UEI’s extensive offering of analog input (with resolutions of 16, 18 or 24 bits, inputs to 100V or 0-20 mA), analog output (to 40V), digital input/output, counter/timer, quadrature encoder, CAN-bus interface, multi-port RS-232/42/485, and power conversion. Pricing for the DNA-SL-501 starts at $650; for the DNA-ARINC-5xx, $3000 and for DASYLab, $499.

A high-performance mobile server arrives as the latest member of Octagon Systems’ Core Systems line of rugged systems with expandable I/O and fanless operation. The RMB-S is a “no compromise” design that optimizes the electrical, thermal and mechanical components for maximum reliability. The basic unit includes the processing power, mobile power supply, memory, connector card and I/O for most applications. Standard I/O includes dual Ethernet, quad USB 2.0, dual serial, CRT & LCD video and digital I/O. The RMB-S is fully functional out of the box, and additional I/O, such as GPS, analog, radio or Wi-Fi, can be readily added via PC/104 and PC/104-Plus modules. An option panel can be easily removed and punched for custom annunciators, connectors and controls. Heat from the system is channeled directly to the case to help prevent internal hot spots. The RMB-S mobile server operates in ambient temperatures from -40° to 70°C, depending upon the processor speed, user options and mass storage devices. A MIL-810F version offers a case with military-grade connectors and gasket sealing to provide dust-resistant, waterproof protection in outdoor environments. Octagon Systems, Westminster, CO. (303) 430-1500. [www.octagonsystems.com].

800-1600 MHz VCO Aims at SDR

A voltage-controlled oscillator (VCO) operates from 800 MHz to 1600 MHz with a control voltage range of 0.5V~19V. The CVCO55CW-0800-1600 from Crystek features a typical phase noise of -100 dBc/Hz @ 10 KHz offset and has excellent linearity. The model CVCO55CW-0800-1600 is packaged in the industry standard 0.5in. x 0.5-in. SMD package. Input voltage is 11.5V, with a max current consumption of 30 mA. Pulling and Pushing are minimized to 5.00 MHz and 1.00 MHz/V, respectively. Second harmonic suppression is -10 dBc typical. The CVCO55CW-0800-1600 is suitable for use in applications such as digital radio equipment, fixed wireless access, satellite communications systems and base stations. Pricing starts at $10.39 each in volume. For additional pricing details, contact Crystek Corporation. Crystek, Ft. Myers, FL. (239) 561-3311. [www.crystek.com].

United Electronic Industries, Walpole, MA. (508) 921-4557. [www.ueidaq.com].

November 2007

53


Products&TECHNOLOGY PXI Card Packs in Sixteen 24-Bit Data Acq Channels

It is becoming easier to build complex, synchronized systems for high-channel-count applications thanks to a new crop of integrated PXI systems from National Instruments. National’s PXI-4498 and PXI-4496 dynamic signal acquisition (DSA) modules offer 16 simultaneous 24-bit analog inputs per module and IEPE constant current signal conditioning for precision measurements with microphones and accelerometers in high-channel-count systems, such as noise mapping, beam forming applications and structural vibration. The NI PXI-4498 features four analog input voltage ranges, while the NI PXI-4496 has two analog input voltage ranges. With 16 simultaneous channels in a single PXI module, these are the most dense DSA modules from NI. Engineers can use the modules with an 18-slot PXI chassis to hold 272 synchronized channels in a single chassis and more than 13,000 synchronized channels in a distributed PXI system. The modules feature a maximum sampling rate of 204.8 ksamples/s per channel and include TEDS smart sensor support for error-free setup. Pricing for the PXI-4498 and PXI-4496 DSA modules starts at $7,999.

Industrial Ethernet Module for PROFINET, EtherCAT, EtherNet/ IP und Powerlink

An Industrial Ethernet module lets devices be quickly and cost-efficiently adapted to various Industrial Ethernet technologies. The Powerlink Controlled Node from IXXAT now supports PROFINET RT IO-Device, EtherCAT Slave and EtherNet/IP Adapter. The module is based on an FPGA from Altera and in addition to an integrated 32-bit CPU and an Ethernet Controller, it also comprises the PHYs and the RJ45 jacks. Depending on the Industrial Ethernet protocol being implemented, the Ethernet interface is either single routed (PROFINET, EtherNet/IP) or double routed (Powerlink, EtherCAT).

National Instruments, Austin, TX. (512) 683-0100. [www.ni.com].

Slim Version of MicroTCA Portable Chassis

A new 4U wide MicroTCA portable tower from Elma Electronic is a slimmer and more compact version than the 6U wide unit announced by the company early this year. The 4U Type 32M MicroTCA Portable Tower features a Star backplane with up to 6 AMCs. The unit also features 1 MicroTCA Carrier Hub (MCH) slot and a Power Module slot. The backplane has a JSM (J-Tag Switch Module) slot, used for diagnostics. Ideal as a development chassis, the unit facilitates either single or double width format modules in the same backplane. The Type 32M features advanced EMC shielding, scratch-resistant vinyl clad aluminum covers and power components. Cooling is achieved with 2 x 90 cfm fans and the chassis has 5 temperature sensors throughout the unit. Elma has performed thermal simulations to ensure the optimal performance. The wider 6U version Type 32M has a Dual Star topology with 2 MCH and 2 Power Modules. Elma also offers Subrack MicroTCA enclosures in 4U-8U heights and a 1U Pico-style MicroBox. The Type 32M MicroTCA Tower price is priced under $2,300, depending on volume and options. The lead time is 6-8 weeks. Elma Electronic, Fremont, CA (510) 656-3400. [www.elma.com].

Using the disclosed programming interface to the Host-CPU (Host-API in C source code), the vendor can connect his device easily using SPI or a standardized ÂľC-Interface. The Host-API encapsulates the functions of the industrial Ethernet protocol running on the module thus the application can be developed independent of the underlying network. Additionally protocol-specific features can also be used by the application. With the optional available baseboard the module can directly be connected to processor modules from Phytec (phyCORE-XC161und phyCORE-LPC2294) or Spectrum Digital (eZdsp F2812), thereby making this solution very suitable for rapid prototyping. The module can also be produced with customerspecific form-factors. IXXAT also offers the integration of customer-specific hardware and software interfaces. In addition to the direct usage of the module, IXXAT also offers a license for the design-in to the specific formfactor of the customer device. IXXAT, Bedford, NH. (603) 471-0800. [www.ixxat.xom].

54

November 2007


Core2 Duo Rides VME Dual PMC SBC 2U Rackmount Dual Intel Xeon Network Appliance

Available to OEMs as an SBC or in a 2U enclosed cabinet, a high-performance 2U network appliance features Intel Dual Xeon processors and the Intel R E7520 chipset providing FSB 800 MHz, plus Intel’s Extended Memory 64-bit Technology (EM64T). The PL-01041, and its control board with configurable Ethernet module design, the MB-09043, have been announced by WIN Enterprises. The MB-09043 board supports system memory up to 8 Gbytes DDRII. PCI Express yields a total system bandwidth of up to 32 Gbits/s. MB-09043 offers eight Gigabit Ethernet LAN ports with optional Ethernet bypass module on four ports. The design utilizes the Intel 82571EB Ethernet controller to enable flexible Plug-and-Serve configuring of the system’s I/O. The system can be configured to support eight Copper or four GbE SFP and four GbE Copper with optional bypass function on 4 to 8 ports. An optional rechargeable daughter card for bypass SFP Ethernet module supports SFP bypass for up to 3 hours. For system login with monitoring and building control proxy, the daughter card has two SATA connectors and a CompactFlash type II socket to enable rapid software loading. Other features include two serial ports, two USB 2.0 ports, one parallel port, one E-IDE connector, three digital I/O, a watchdog timer and hardware system monitoring capability. The control board and appliance also support two 64-bit PMC slots and one Mini PCI slot for additional expansion.

The PL-01041 appliance contains the MB-09043 and supports two removable 3.5” SATA HDDs that can be inserted/removed easily from the front panel. The front panel also has one USB 2.0 port, RJ-45 console port, a 20 x 2 character LCM, a four buttons keypad, a LED indicator for power, HDD, SFP link and bypass function for local system management. The PL-01041 appliance is priced at $2,460, and the MB-09043 SBC, also available separately, is priced at $497 in OEM quantities. WIN Enterprises, North Andover, MA. (978) 688-2000. [www.win-ent.com].

Concurrent Technologies has released their latest single-slot VME dual-core processor, dual PMC, single board computer. The VP 417/03x uses the latest mobile dual-core processors from the Intel embedded roadmap, the 1.5 GHz or the 2.16 GHz Intel Core 2 Duo processor. The VP 417/03x family is I/O-compatible with the popular and longstanding VP 31x/02x family and gives an even greater improvement in measured performance/watt. The board is ideal for low power intensive processing applications where the dual processor cores can access up to 4 Gbytes on board DDR2 ECC SDRAM at up to 6.4 Gbytes/s. With two 66 MHz PMC sites, suitable for a wide range of PMC modules, the board also includes extensive I/O functionality and can support VITA 31.1. Commercial and extended temperature versions are now available, and ruggedized, conduction-cooled or air-cooled versions will be available shortly. VITA 31.1 Gigabit Ethernet on a VME64x backplane enables a tried and tested method of implementing a LAN-based multiprocessor architecture by leveraging readily available Ethernet hardware, TCP/IP software, clustering and other network management tools. Concurrent Technologies, Woburn, MA. (781) 933-5900. www.gocct.com.

PCI Express 8-Channel 25 Msample/s 12-bit Oscilloscope / Digitizer Card

With eight 25 Msample/s 12-bit ADCs for simultaneous sampling without time-skew errors and up to 2 Gsamples onboard memory, a new PCI Express oscilloscope/digitizer card supports streaming to a host PC to the maximum transfer rate of the PCI Express x1 bus. In the UF2e-3132 from Strategic Test, each channel has a software-programmable amplifier allowing input ranges of ±50 mV to ±10 V. The UF2e-3132 offers a wide range of triggering possibilities, including level, window, pulsewidth, re-arm and double triggers, or an external TTL trigger input. Trigger conditions can be set on multiple cards or systems combined with AND/OR logic, a feature often used in production test. SDKs for Microsoft Windows Vista, XP64, XP and Linux (RedHat, Fedora, SuSe, Sarge) and the SBench 5.3 oscilloscope program are supplied with the card. SDKs for LabVIEW, MATLAB; Agilent-VEE, DASYLab and LabWindows/CVI are available as options. The effective Number of Bits is 10.1 bits; the signal-to-noise ratio is >62.5 dB and the total harmonic distortion is < -61.0 dB and -3dB bandwidth > 12.5 MHz. In addition, the UF2e-3132 scope card can be customized with hardware options: Onboard memory is available for 32, 64, 128, 256, 512, 1024 or 2048 Msamples; multiple recording via memory segmentation and gated sampling using an output clock controlled by external TTL signal and use of BaseXIO adds 8 asynchronous digital I/O lines. Prices start at $10,390 with discounts for volume. Strategic Test Corporation, Woburn, MA. (617) 621-0080. [www.strategic-test.com]. November 2007

55


Products&TECHNOLOGY Mini-ITX Motherboard Supports Core2 Duo and Graphics

An Intel 945GM-based Mini-ITX motherboard supports advanced dual core processor technology. The AIMB-253L from Advantech includes the integrated Intel GMA 950 graphics controller and features 256 Mbytes of shared video memory along with DirectX 9 3D hardware acceleration. For multiple displays or future graphics upgrades, a PCI Express x16 slot supports higher performance graphics cards that permit convenient connection to various display devices like CRTs and LCDs. The AIMB-253L offers a host of integrated I/O interfaces that speed up installation and deployment. Providing dual LAN connectivity, the PCI Express-based Intel 82573 Gigabit Ethernet controller delivers up to 1000 Mbits/s of bandwidth. The AIMB-253L also has two Serial ATA II 300 Mbyte/s connectors. Graphics and I/O expansion are supported by the PCI Express x16 and 32-bit PCI slots. In addition, there are four USB 2.0 ports; three RS-232 and two RS-422/485 ports; one parallel port; LVDS and VGA ports and one PS/2 port. Integrated graphics are provided by the Intel GMA 950 graphics core, which supports up to 256 Mbytes of video memory for greater resolution, more colors and more realistic game play. The AIMB-253L also has Intel 7.1 channel high definition audio via an S/PDIF output. It is suitable for customers who need dual display and high-quality graphics support. The low power and rich I/O design reduces the total cost of ownership (TCO) for end consumers, and makes the AIMB-253L the most cost-effective choice. The AIMB-253L supports Microsoft Windows XP, 2000, XPE and CE operating systems. The AIMB-253L is RoHS compliant. Advantech, Irvine, CA. (949) 789-7178. [www.advantech.com].

32 Gbyte 2.5” SATA SSD Features a Read; Write: 60 Mbyte/s; 48 Mbyte/s

A new, non-volatile 32 Gbyte SATA solid-state disk (SSD) drive offers performance that is almost twice as fast as a conventional hard disk drive. Additionally, a 64 Gbyte version of this 2.5” SATA drive is slated to be available in late November. The Ridata SSD SATA 2.5-inch drive from Advanced Media reduces boot up and seek times to seconds compared to conventional hard disk drives. Users can install new software programs in seconds rather than minutes. The data transfer rates are 60 Mbytes/s for maximum sequential read and 48 Mbytes/s for maximum sequential write. It requires a five-volt power supply and because of its low power consumption will dramatically extend user time on a laptop or other mobile computing application. SSD drives are inherently resistant to vibration, shock and temperature extremes. The RoHS-compliant drive is very reliable with more than 4,000,000 MTBF. Data integrity is further supported by dynamic wearing leveling and a built-in ECC that corrects up to 3 random 12-bit symbols (Reed Soloman-4) of error per 512-byte sector. With a write endurance cycle (P/E) of more than 2,000,000 times, these drives are built to safely write and store critical data. With no moving parts, the drive meets bus interface industry standard Serial ATA (SATA), so there are no host compatibility or upgrading issues. The drive produces virtually no noise, no heat generation and no vibration, while offering reliable temperature (0° to 70°C) and humidity resistance that ensures long-lasting data storage. It is engineered with semiconductor components and uses fast Single Level Cell (SLC) NAND flash memory. The 2.5” 32 Gbyte drive weighs 64.5 grams with dimensions of 101.85 x 69.85 x 9 mm. Advanced Media, Diamond Bar, CA. (909) 861-2269. [www.ritekusa.com].

56

November 2007

Extended Temperature, Conduction-Cooled 3U CompactPCI SBC at 15W.

A new 3U CompactPCI SBC offers flexible, ruggedized high performance. The CPC7525 from Orion Technologies is available in both conduction-cooled and convection-cooled versions, and is designed for high performance in the most demanding applications. The flexible “Personality Module” and the ability to automatically detect whether to function as a system controller or an intelligent peripheral card allow the user to migrate from one application to another without hardware reconfiguration. The CPC7525 includes standard features such as a PMC slot (PrPMC and PCI-X capable) three Gigabit Ethernet ports, two serial ports, non-volatile RTC, 16 Kbits of EEPROM, PMC P14 and general-purpose I/O. Additional optional features include 128 Mbytes of onboard user flash and customer-specific I/O. Processors supported are IBM’s 750FX at 600 MHz and the 750GX running at 1 GHz. By incorporating the IBM 750 FX/GX PowerPC, Marvell’s Discovery III controller and the complement of I/O via the customer configurable “Personality Modules,” the manufacturer can adapt to almost any Military, Industrial or Commercial application. Volume pricing for a conduction-cooled 1 GHz 750GX starts at $3,017. Orion Technologies Merritt Island, FL. (321) 452-1670. [www.otisolutions.com].


Serial-to-Ethernet Gateway Connects Existing Equipment to the Internet

A plug and play, gateway module converts RS-232 protocol into TCP/IP protocol—and also includes DB-9 and RJ45 connectors for immediate use. The WIZ110SR from Saelig enables remote checking, managing and control of devices via Ethernet and TCP/IP by connecting existing RS-232 equipment at up to 230 Kbits/s to a network. WIZ110SR is a protocol converter that automatically converts and transmits data sent by serial equipment to TCP/IP data and converts the received TCP/IP data into serial data for the equipment.

Simple software set-up allows users to rapidly create a 10/100 Mbit/s Ethernet interface with a maximum serial rate of 230 Kbits/s. WIZ110SR is a compact 2” x 1.2” board with built-in system connectors to allow any device with serial inputs and/or outputs to be Ethernet/Internet-enabled for less than $29. WIZ110SR offers TCP, UDP, IP, ARP, ICMP, IGMP, Ethernet MAC and PPPoE protocols on a 10/100 Base-T auto-detecting Ethernet network. WIZ110SR is based on WIZnet’s TCP/IP- offload IC W5100 with built-in Ethernet driver stage. Three different modes of operation are supported: TCP server, TCP client and UDP. TCP guarantees data delivery, but UDP doesn’t require acknowledgement, so communication can be faster. Once the TCP client and server have established connection, data can be transparently transmitted bi-directionally. WIZ110SR’s IP Address can be manually assigned (static IP), or IP, subnet and gateway address can be acquired from the DHCP server automatically. Serial commands can be hardware- or software-triggered. Quantity one pricing is $29. Saelig, Pittsford, NY. (585) 385-1750. [www.saelig.com].

1 MHz to 2.7 GHz Spectrum Analysis in Handheld Unit

A new instrument utilizes a handheld computer to provide a 2.7 GHz spectrum analyzer in the palm of a hand. Weighing about a pound, the PSA2701T from Saelig offers sweep modes of continuous, single, peak hold and average (up to 256 sweeps) with unlimited storage for waveforms, setups and fully annotated screens under user-defined file names. PSA2701T is built around a Palm T|X handheld computer, so data transfer to a PC for analysis, documentation and printing is easy via the built-in mini USB connector. Additional features include Wi-Fi (802.11b), Bluetooth 1.1 and Infrared interfaces. Bitmap images of the whole screen can be stored for viewing or printing. PSA2701T offers a 93 dBm typical noise floor, -20 or 0 dBm reference level, a zero span mode with AM and FM audio demodulation, and resolution bandwidths of 1 MHz, 280 kHz, 15 kHz. The sweep parameters can be set as “centre plus span” or “start plus stop” to 1 kHz resolution. A “zero span mode” with AM or FM demodulation is also provided. Stored reference traces can be displayed simultaneously with live trace, clearly differentiated by color, the reference trace being automatically shifted and scaled to match the current sweep parameters when they are changed. Control of the compact 6.7” x 3.8” x 1.85” analyzer is by soft keys on the touch screen, or by using the handheld’s hard keys. Quantity one pricing is $1,999. Saelig, Pittsford, NY. (585) 385-1750. [www.saelig.com].

Interface Delivers 10Gb Ethernet to Real-time and Embedded Applications

A 10Gb Ethernet (10GbE) networking interface for embedded, military and avionics applications incorporates Silicon Stack technology that offloads TCP/IP protocol stack processing to hardware, thus allowing wire speed transfers, minimal host processor overhead, very low latency and rock-solid determinism. As a result, the 10GbE XGE interfaces from Critical I/O permit users of high-performance systems to benefit from the low cost, interoperability and networking capabilities of Ethernet, even in such applications as radar, sonar, flight simulation and scientific applications. According to the company, 10GbE holds the promise of providing an order of magnitude increase in performance, offering an attractive alternative to other more specialized data networking technologies, such as InfiniBand and Serial Rapid I/O. The problem users run into, however, has to do with the software-intensive nature of Ethernet’s TCP/IP protocol stack. Many embedded systems have struggled to keep up with the stack processing associated with 1GbE connections, so dealing with 10GbE presents 10 times the challenge. The Silicon Stack technology is used as a solution to offload protocol processing to silicon and reduce host processor loading so as to deliver true 10 Gbit performance. The XGE 10Gb family is currently available in XMC form-factor with dual 10GbE ports, and employs an 8-lane PCI Express host bus interface. AMC and PMC versions are also planned. Native support for IPv4 and IPv6 is included. Critical I/O, Irvine, CA. (949) 553-2200. [www.criticalio.com]. November 2007

57


Products&TECHNOLOGY RFID Engine for Printers, Handheld Readers and Other Portable Apps

An embeddable RFID engine for printers, handheld readers and other portable RFID applications is specifically designed for mobile, portable and handheld RFID applications and is suitable for applications that require economy of power and space. The M5e-Compact from ThingMagic uses ThingMagic’s Mercury OS software and provides a transmit power of +23 dBm (200 mW). The M5e-Compact is EPCglobal Generation 2 DRM compliant, and is optimized for Generation 2 RFID tags. The embedded reader also combines multi-region capability into a single module, enabling immediate, locally compliant deployment of any M5e-Compact in North America, South America (countries adopting the FCC frequency plan), Europe (including TG34), South Korea, Japan and China. The ThingMagic M5e-Compact is available with a development kit that enables customers to quickly create testing and use environments for Generation 2 RFID tags (samples provided). This allows PCs to read and write RFID tag data, enables users to monitor the operation of the M5e-Compact control utility, and permits developers to reuse code contained in the utility to optimize specific applications. The M5e-Compact modules and development kits are available now on an OEM basis to manufacturers of label makers, print and apply machines, and other embedded, mobile and handheld RFID devices. ThingMagic, Cambridge, MA. (617) 499-4090. [www.thingmagic.com].

32-Channel, 16-Bit Analog Output Board for PowerDNA Cube Family

A new analog board provides 32 channels of 16-bit resolution analog output with a ±10 volt output range for the United Electronic Industries’ “Cube” architecture. The DNA-AO-332 board is capable of driving ±10 mA while per-channel digital offset and gain calibration limits gain/offset errors to ±450 µV and ±305 µV respectively. Compatible with UEI’s PowerDNA, UEIPAC and UEIModbus “Cubes,” the DNA-AO-332 provides up to 192 analog outputs in a single, 4” x 4” x 5.8” cube. A 1024 sample FIFO on each channel allows each D/A to be updated at 10 kHz without data loss. Double buffering the outputs combined with the use of low glitch D/As make the DNA-AO-332 an ideal solution for generating low frequency waveforms or providing highly accurate switched stimuli. All outputs may be set to update simultaneously if desired. Software for the DNA-AO-332 is provided in the UEIDAQ Framework. The Framework provides an API that supports all popular programming and operating systems including Windows, Vista, Linux and most real-time operating systems (e.g. QNX, RTX, RT Linux). Finally, the board is fully supported by LabVIEW, MATLAB/ Simulink, DASYLab or any application supporting ActiveX, OPC or Modbus TCP control. The DNA-AO-332 is priced at $2,400. United Electronic Industries, Walpole, MA. (508) 921-4600. [www.ueidaq.com].

58

November 2007

Host-to-Host Communications over PCI Express

A PCI Express (PCIe) x4 host-to-host networking technology allows up to 10 Gbit/ s data transfers over a x4 PCI Express link. The SuperSwitch from One Stop Systems is first appearing in a point-to-point solution for two PCs, but will soon provide networks of up to eight computers. The first offering, SuperSwitch 1 includes a PCIe x4 host adapter attached via a PCIe x4 cable to One Stop Systems’ CompactPCI Express x4 cable adapter, PCIe x4 cable adapter, or PCIe x4 Express Card. Using One Stop Systems’ ExpressNet utility software, data transfers between two systems occur at 10 Gbits/s. The ExpressNet software takes advantage of nontransparent bridging to communicate with the CPU on the other side. While it cannot communicate directly with the other CPU’s PCI Express peripherals, it can request data transfers to and from them via the other host CPU. One Stop Systems’ ExpressNet driver and utility software supports 64-bit CPUs as well as multi-threaded environments. In comparison to other network solutions, PCIe is less costly and operates at much higher speeds, with Ethernet and InfiniBand options costing as much as two-to-three times the PCIe solutions. Ethernet and InfiniBand also require bus-to-bus conversions creating increased latencies.

Available cable lengths are: 1m, 2m, 3m, 5m and 7m. SuperSwitch 2 includes connecting up to five processors using One Stop Systems’ PCIe x4 mezzanine cable adapter and four PCIe x4 cable adapters. SuperSwitch 2 enables data transfers at 20 Gbits/s. SuperSwitch 3 includes connecting up to eight CPUs or any combination of CPUs and I/O devices, using One Stop Systems’ 8-port 1U switch. It provides data transfers at 40 Gbits/s. SuperSwitch 1, including ExpressNet, lists for $1,695. One Stop Systems, Escondido, CA. (877) 438-2724. [www.onestopsystems.com].


Mass Storage Modules for VMEbus and CompactPCI®

COM Express Entry Offers Three CPU Options

Designed for applications including industrial, simulation/ training, test and measurement, gaming and transportation, the bCOM2-L1000, the first COM Express product from GE Fanuc Intelligent Platforms, offers a choice of Intel Core2 Duo, Core Duo or Celeron processors with up to 2 Gbytes of memory, and is supported by the CCAR-L1000 carrier board for customer evaluation of the bCOM2-L1000. The choice of three processors allows users to select an appropriate option depending on the power/heat dissipation and performance characteristics of the intended application. The Celeron M option—which has the processor soldered to the board for even greater reliability—is specifically designed for temperature-critical environments. Four serial ATA ports and an IDE port provide the bCOM2-L1000 with the ability to handle extensive, high-performance hard disk storage. Graphics performance for the bCOM2-L1000 is provided via integrated analog CRT and LVDS interfaces, DVMT 3.0, Intel’s Dual-Frequency Graphics Technology (which reduces integrated graphics chipset power) and dual independent display pipes. The ATX form-factor CCARL1000 complements the bCOM2-L1000 with a single x16 PCI Express slot, two x1 slots and four PCI 2.3 slots. Rear panel interfaces include mouse/ keyboard, serial, Ethernet, USB, VGA and audio, while onboard support additionally includes a parallel interface, CD-in, S/PDIF in/out and IrDA. edrock_04.indd

PMC CompactFlash Module Two Type I/ Type II CF Sockets

See the full line of Mass Storage Products at

www.RedRockTech.com

or call Toll-Free: 800-808-7837 Red Rock Technologies, Inc. 480-483-3777

1

2/2/07 1:21:52 PM

GE Fanuc Intelligent Platforms, (800) 368-2738. Charlottesville, VA. [www.gefanucembedded.com].

2

C OTS

Adapter Puts PMC into Standard PCI/PCI-X Slot

PCI is arguably the most successful bus to invade the embedded market, with numerous flavors and form-factors for every need. Dynamic Engineering’s latest PCI-X product is the PCIBPMCX1, an adapter/carrier converter card that offers the ability to install one PMC card into a standard PCI/PCI-X slot. The PCIBPMCX1 PMC card slot can be programmed for 3.3 or 5V operation by the user, and the primary PCI bus implementation is universal voltage. It is suitable for PCI or PCI-X operation with 32-bit or 64-bit data and 33, 66, 100 or 133 MHz clock. The PMC user I/O connector Pn4 is available on a SCSI II connector. The PCIBPMCX1 has a cooling cutout for increased airflow to the PMC. A cutout is preferred over a fan mounted to the PMC adapter for several reasons including: many components are not shielded against close proximity to electromagnetic fields, and the reduction in MTBF that an additional mechanical device represents. The PCI bus is interconnected to the PMC via a 64-bit 133 MHz-capable bridge. The bridge allows the PCI bus to operate with different parameters than the PMC card. Single unit price is $625 and quantity discounts are available. Dynamic Engineering, Santa Cruz, CA. (831) 457-8891. [www.dyneng.com].

Your product

cPCI PMC VME PCI ATX choice of OS

Standard & custom designs Extended temp. conduction or convection cooled SBCs Fast, flexible, reliable

ISO 9001:2000 Certified 321-452-1670 Visit us: www.OTIsolutions.com Untitled-1 1

November 2007

59

11/8/07 9:42:51 AM


, VIEWS NEWS & COMM E N T

November 2007

Embedded Computer Market Continues to Fly High

W

hat continues to be interesting about the embedded computer market—the total standard and stand-alone board-level computers, embedded PCs and modular computer subsystems—is the diversity of applications it continues to address. So, when we look at the growth of the market, we are looking at a growth in terms of penetration in an increasing number of applications, a scenario where the depth is relatively low, but the breadth—the number of applications—is growing at almost exponential rates. The result appears to be a market characterized by relatively small volumes per application, but an ever increasing number of different opportunities. This fits the more classic board-level embedded computer market with relative small volumes and high profit margins. Let’s take a quick look at the overall market. When we add everything up, we come up with a total merchant market for embedded computers that’s approaching $4.5 billion. Of that total, the defense and aerospace market takes up about $1.2 billon. Industrial Automation and control sucked up just under another billion. The balance is divvied up among the telecom infrastructure providers, medical instrumentation and diagnostics, retail and consumer, and a variety of other applications that stretch from scientific exploration such as the high-energy physics people and control areas (such as the binocular telescope control discussed in these pages a few months ago) to transportation including rail, truck and air-based control systems. And, of course, there are other segments that haven’t been defined or identified at this time. Also, what makes the market particularly difficult to measure is the number of systems developed and sent to ECMs (electronic contract manufacturers), which frequently get under the radar. Therefore, I estimate that the actual number is somewhat larger than what I’ve been able to measure—perhaps by as much as 1.5 times. The RTC editorial staff’s best estimates of how these settle out is seen in Figure 1. Last month we discussed the small formfactor board business saying that it was closing in on—and may have already exceeded—$1 billion in annual revenue. Though exact numbers are difficult to ascertain, it would appear that the small form-factor business is divided into almost all market seg-

60

November 2007

ments with the possible exception of telecom—at least until, or when, volume production of MicroTCA becomes a reality. It’s estimated some 15% to 20% of small form-factor business goes to the aerospace and defense industry.

Moore’s Law Survives—So Does Moore

The world has been lucky that no law of physics has so far stood in the way of Moore’s Law, says the 78-year-old cofounder of Intel, Gordon Moore, and the first to describe a formula for predicting the rate at which chips increase in capacity. He made the comment on the 50th anniversary of the founding of Fairchild Semiconductor, the company that spawned almost every successful semiconductor company in the U.S. He also commented that he is continually surprised by what people are able to do with “these things.” Every few years scientists predict Moore’s Law will come to a grinding halt. Early on it was thought the practical limits of lithography were going to inhibit continued growth. Then during the 1970s, cell sizes were reduced to dimensions that single alpha particle hits could erase information, and it was believed that that would bring an end to Moore’s Law. More recently, interconnect materials were to blame, and most recently power dissipation has been a limiting factor. However, despite all the obstacles, capacity continues to double about every 18 months. Gordon Moore obviously knew something.

Sun Servers Use Niagara 2 Processors

Back in August, Sun announced the availability of its UltraSPARC T2 (Niagara 2) to the merchant market as the first volume processor to offer eight cores with eight threads per core. The company claims it’s raised the bar for both raw performance and energy efficiency. Now the company has announced the first in a family of servers using the chip. The ability of UltraSPARC T2 to provide a 64-way system on a single chip lets the new Sun servers act like 64 separate servers.

Bubble Indicators?

Many of us remember the “dot-com” bubble of the end of


NEWS, VIEWS & COMMENT

Warren Andrews Associate Publisher

Semi Equipment Spending

Dollars in Billions 1.2 1 0.8 0.6 0.4 0.2 0

A

D&

Figure 1

s

m

Co

l

ica

d Me

C

A&

Tr

rt

ic

tif

po

s an

S

n cie

l

tai

Re

The embedded computer market—standard and stand-alone board-level computers, embedded PCs and modular computer subsystems—is reaching up toward $4.5 billion led by defense and aerospace, followed by communications, then automations and control, retail and transportation.

the last millennium. And while there are any number of Monday-morning quarterbacks willing to tell us what happened, no one has quite determined how to predict the next bubble. While the government and private industry monitor thousands of things from consumer confidence to inventory levels to forecast the future, some Silicon Valley veterans have apparently developed some leading indicators of their own. According to the Wall Street Journal, one such index is the goofy-names index. Some names that top that list are Orgoo, Inc. and Zipidee, Inc. to name a few. And there are others such as BillMonk, Obopay, Inc. and ZocDoc, all going enterprises. But who knows. In the past bubble, there were certainly many, many losers, but how about some with goofy names that are still around today? What about Google, stock symbol GOOG? Or Yahoo? Does a goofy-names index give us any clue as to when the next bubble is about to burst? Or which company will be the next Google? In terms of market cap, Google ranks in the top 10 companies worldwide.

There has been some correlation in the past between chip sales and growth of the embedded market. Then there are the semiconductor equipment makers such as Novellus, Applied Materials, KLATencor and others that are considered leading indicators and, among other things, users of embedded computers. However this season, there seem to be some mixed signals. According to the Semiconductor Industry Associates in its recent report, it showed that chip sales climbed almost 5%. One of the big gainers was NAND flash memory, which saw a sales surge of 48% from the previous year. It’s guestimated that the acceleration of handheld devices such as cell phones, personal MPEG players, PDAs, palm-top computers and combinations of such has fueled the growth of flash. And while chip sales look encouraging, the companies making the equipment that builds the chips may not be faring as well. Researcher Gartner, Inc. reports 2008 spending on capital equipment for chip makers is expected to decrease somewhat over 4%. However there are mixed signals. Samsung Electronics eked out a scant 1% increase in net, yet plans to increase capital spending in its semiconductor division by 25% adding $1.5 billion to the already robust $5.9 billion planned for the year. But on the other side of the fence, memory maker Micron Technology is looking at a 2008 reduction in capital expenditure to $2.5 billion down from $4 billion in 2007. Micron, whose primary business is DRAM, has seen an inventory glut this year with a drop in ASPs.

A&D Workforce Study

According to Aviation Week’s recent employment report, the aerospace and defense industry has over 40,000 job openings in all categories. The shortage of skilled workers has the industry’s largest employers working to be the best choice for employees. These companies say that the U.S. is beginning to “feel the pinch of being talent constrained,” looking beyond boarders for critical talent. The Boeing Company was singled out as one of the top employers. According to the report, some 2,000,000 people applied to the company for work last year. Current employment rolls have the company with 158,000 employees. November 2007

61


is to research the latest datasheet from a company, speak directly with an Application Engineer, or jump to a company's technical page, the goal of Get Connected is to put you in touch with the right resource. Whichever level of service you require for whatever type of technology, Get Connected will help you connect with the companies and products you are searching for.

www.rtcmagazine.com/getconnected

Advertiser Index Get Connected with technology and companies providing solutions now Get Connected is a new resource for further exploration into products, technologies and companies. Whether your goal is to research the latest datasheet from a company, speak directly with an Application Engineer, or jump to a company's technical page, the goal of Get Connected is to put you in touch with the right resource. Whichever level of service you require for whatever type of technology, Get Connected will help you connect with the companies and products you are searching for.

www.rtcmagazine.com/getconnected

Company

Page

Website

Advanet Technologies............................................................................................44...........................................................................................www.advanettech.com Aptyc....................................................................................................................45..................................................................................................... www.aptyc.com BittWare................................................................................................................18................................................................................................. www.bittware.com

Products

End of Article

Concurrent Technologies Plc..................................................................................10..................................................................................................... www.gocct.com Corvalent..............................................................................................................49....................................................................................................... corvalent.com Data Device Corporation........................................................................................33................................................................................................. www.ddc-web.com

GetLogic Connected with companies and Get Connected Digital AG.....................................................................................................26............................................................................................. www.digitallogic.com products featured in this section.

with companies mentioned in this article.

ELMA Electronic, Inc..............................................................................................32...................................................................................................... www.elma.com www.rtcmagazine.com/getconnected www.rtcmagazine.com/getconnected Emerson Network Power .......................................................................................27................................................................ www.emersonembeddedcomputing.com GE Fanuc Embedded Systems.................................................................................4.................................................................................. www.gefanucembedded.com Harting, Inc...........................................................................................................11...................................................................................................www.harting.com Get Connected with companies mentioned in this article.

www.rtcmagazine.com/getconnected Harting, Inc. EPT....................................................................................................23.............................................................................www.harting.com, www.ept.com Get Connected with companies and products featured in this section. www.rtcmagazine.com/getconnected

Hybricon Corporation.............................................................................................14.................................................................................................www.hybricon.com IEI Technology.......................................................................................................20........................................................................................... www.usa.ieiworld.com Intel......................................................................................................................25.......................................................................................................www.intel.com Kontron America....................................................................................................64..................................................................................................www.kontron.com McObject LLC........................................................................................................49............................................................................................... www.mcobject.com MEN Micro, Inc......................................................................................................15.............................................................................................. www.menmicro.com Microsoft Windows Embedded.............................................................................. 2,3..............................................................................www.microsoft.com/embedded Motorola...............................................................................................................63................................................................................................ www.motorola.com MVACEC................................................................................................................39.................................................. www.mvacec.com/www.mountainviewalliance.org National Instruments..............................................................................................8............................................................................................................www.ni.com One Stop Systems.................................................................................................35.................................................................................... www.onestopsystems.com Orion Technologies,Inc...........................................................................................59............................................................................................www.otisolutions.com Pentek, Inc............................................................................................................17................................................................................................... www.pentek.com Performance Technologies.....................................................................................29.......................................................................................................... www.pt.com Phoenix International..............................................................................................6................................................................................................. www.phenxint.com Real-Time & Embedded Computing Conference......................................................40......................................................................................................www.rtecc.com Red Rock Technologies, Inc...................................................................................59........................................................................................... www.redrocktech.com Sensoray Company................................................................................................41................................................................................................www.sensoray.com TRI-M Systems......................................................................................................24...................................................................................................... www.tri-m.com VadaTech..............................................................................................................21................................................................................................www.vadatech.com

RTC (Issn#1092-1524) magazine is published monthly at 905 Calle Amanecer, Ste. 250, San Clemente, CA 92673. Periodical postage paid at San Clemente and at additional mailing offices. POSTMASTER: Send address changes to RTC, 905 Calle Amanecer, Ste. 250, San Clemente, CA 92673.

62

November 2007


THE ADVANTAGES ARE CLEAR Motorola is the clear choice for MicroTCA™. Motorola’s cutting edge MicroTCA products can help cut your costs, risk and design cycle while increasing your flexibility. Whether your applications need specific I/O, flexible packaging, ruggedization or a choice of processors, we can customize solutions to meet your needs. That’s why our MicroTCA products are quickly being adopted for network-centric applications across telecommunications, defense, aerospace, industrial and medical industries. Looking for a clear advantage over the competition? HELLOMOTO™ See why Motorola should be your first choice for MicroTCA solutions at: motorola.com/computing/MTCA

MOTOROLA and the Stylized M Logo are registered in the US Patent & Trademark Office. MicroTCA and the MicroTCA logo are trademarks of PICMG. All other products or service names are the property of their respective owners. © Motorola, Inc. 2007. All rights reserved.


Feeling locked in?

ACCESS EDGE

flexible

DATA CENTER

OPEN MODULAR SOLUTIONS

ENTERPRISE

CORE TRANSPORT

Design your next application with Kontron. IMS • IPTV • WIRELESS • CONTENT DELIVERY NETWORKS

Don’t get locked in with ‘one-vendor-fits-all’ ATCA platform providers.

Kontron and the Kontron logo are registered trademarks of Kontron AG. All other trademarks are the property of their respective owners. ©2007 Kontron America, Inc.

Why, you ask? Because no one ATCA vendor can do it all. That is why you demand an ATCA platform provider that is open to pre-validate, pre-test and, of course, provide the exceptional flexibility to integrate even third-party hardware and software. At Kontron, we hear you loud and clear. We draw from an expanding portfolio of GbE and 10GbE ATCA platform elements, AMC processor, storage and I/O modules, plus extensive integration services of thirdparty hardware along with carrier-grade OS and HA middleware. Talk today with a Kontron system architect to help you from start to finish, and experience firsthand why seeing your application go to market quickly is important to us. Now that’s flexible.

kontron.com/OMS

1-888-294-4558 - info@us.kontron.com — EMEA: +49 0 800 7253756 - Asia: +886 2 2910 3532

Switch Boards Blur the line with Routers  

And it’s also surprisingly affordable. More time, more space, more money to do other things with. Does life get any better? News, Views and...

Read more
Read more
Similar to
Popular now
Just for you