Page 1

The magazine of record for the embedded computing industry

October 2013


ARM Wrestling: Tips for Optimizing Power Use Pick the Right Option for Mobile Wireless An RTC Group Publication

“Hobbyist” Boards to Impact Development

Fully Flight Qualified

Our Application Ready systems keep development time and costs grounded. X-ES systems are fully flight qualified to MIL-STD-810, MIL-STD-461, MIL-STD-704, and DO-160 specifications. We design, develop, manufacture, test, and support the systems and perform qualification under one roof in the U.S.

You’re cleared for take-off with fully qualified systems. That’s Extreme. Visit to see our flight qualified systems.

Extreme Engineering Solutions 608.833.1155


41 Quad-Port and Dual-Port Network Adapters with CostEffective Bypass Design

43 Feature-Rich COM Express Module with Fourth Generation Core Processor


48 Fast Embedded Memory Screams at 400 Mbyte/s





Small Modules Go Rugged

Developing for Low-Power Systems


Editorial Form Factors? “We don’t need no Steenking Form Factors!”



Industry Insider Latest Developments in the Embedded Marketplace


VITA 74: Small, Conduction-Cooled, Rugged and Modular Wayne McGee, Creative Electronic Systems

Wireless Connectivity for Mobile Devices

10 20 & Technology 40Products Newest Embedded Technology Used by TECHNOLOGY IN SYSTEMS Industry Leaders

Using Wi-Fi to Connect Embedded Systems with Mobile Devices

Small Form Factor Forum Bay Trail Closes the Gaps

EDITOR’S REPORT “Hobbyist” Boards Could Change Development

Mitch Dale, Microchip Technology

Optical Fiber Instrumentation


Fiber Optic Sensing Opens New Capabilities for Structural Analysis

ARM Wrestling: Tips 28Championship for Getting the Most out of ARM Cortex-M3 & M4 Microcontrollers Rasmus Christian Larsen, Silicon Labs

INDUSTRY WATCH High Level, High Performance

A Signal and Image 32VSIPL: Processing Standard Family

Anthony Skjellum, PhD, Run Time Computing Solutions

36LabView or C?

Simon Hogg, National Instruments

Alex Tongue, 4DSP

Boards Signal a Shift in Development Strategies 12“Dot-Org” Tom Williams

Digital Subscriptions Available at RTC MAGAZINE OCTOBER 2013


OCTOBER 2013 Publisher PRESIDENT John Reardon,


Bridge the gap between ARM and x86 with Qseven Computer-on-Modules

One carrierboard can be equipped with Freescale® ARM, Intel® Atom™ or AMD® G-Series processor-based Qseven Computer-on-Modules. conga-QMX6

ARM Quad Core



Intel® Atom™

AMD® G-Series


Art/Production ART DIRECTOR Jim Bell, GRAPHIC DESIGNER Michael Farina, LEAD WEB DEVELOPER Justin Herter,


Advertising/Web Advertising

congatec, Inc.


6262 Ferris Square | San Diego | CA 92121 USA | Phone 1-858-457-2600 |

Untitled-4 1

8/14/13 2:18 PM

Billing MSC Embedded Inc. Tel. +1 650 616 4068

Cindy Muir, (949) 226-2021

Qseven™ -

MSC Q7-IMX6 Compatible Modules from Single-Core to Quad-Core

ƒ Freescale i.MX6 Quad-, Dualor Single-Core ARM Cortex-A9 up to 1.2 GHz

HOME OFFICE The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Phone: (949) 226-2000 Fax: (949) 226-2050,

ƒ up to 4 GB DDR3 SDRAM ƒ up to 64 GB Flash ƒ GbE, PCIe x1, SATA-II, USB ƒ Triple independent display support

The MSC Q7-IMX6 with ARM

ƒ HDMI/DVI + LVDS up to 1920x1200

Cortex™-A9 CPU is a compatible

ƒ Dual-channel LVDS also usable

module with economic single-core CPU, strong dual-core processor or a powerful quad-core CPU with


as 2x LVDS up to 1280x720 1.1, OpenCL™ 1.1 EP

up to 1.2 GHz, and provides a very

ƒ UART, Audio, CAN, SPI, I2C

high-performance graphics.

ƒ Industrial temperature range


Editorial Office Tom Williams, Editor-in-Chief 1669 Nelson Road, No. 2, Scotts Valley, CA 95066 Phone: (831) 335-1509

ƒ OpenGL® ES 1.1/2.0, OpenVG™


Untitled-3 1

To Contact RTC magazine:

8/14/13 2:16 PM

Published by The RTC Group Copyright 2013, 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.

U.S. Postal Service Statement of Ownership, Management and Circulation Required by 39 USC 3685. 1) Title of Publication: RTC Magazine 2) Publication Number 1092-1524 3) Filing Date 10/01/2013. 4) Frequency of issue is monthly. 5) Number of issues published annually: 12. 6) Annual subscription price: n/a. 7) Complete Mailing Address of Known Offices of Publication: The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Orange County. 8) Complete Mailing Address of Headquarters of General Office of Publisher: The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Orange County, California. 9) Publisher: John Reardon, The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Orange County, CA 92673. Editor: Tom Williams, The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Orange County. Managing Editor: Sandra Sillion: The RTC Group, 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Orange County, CA 92673. 10) Owners: John Reardon, Zoltan Hunor. The RTC Group; 905 Calle Amanecer, Suite 250, San Clemente, CA 92673 Orange County. 11) Known Bondholders Holding 1 Percent or More of Total Amount of Bonds, Mortgages, or Other Securities: None. 12) Tax Status: The purpose, function, and nonprofit status of this organization and the exempt status for federal income tax purposes have not changed during the preceding 12 months. 13) Publication Title: RTC Magazine 14) Issue date for Circulation data: 9/1/2013. 15a) Extent and Nature of Circulation: average numbers of copies each issue during preceding 12 months (Net press run): 19,333. Number copies of single issue published nearest to filing date: (net press run) 18,000 15b) 1. Paid/requested outside-county mail subscriptions stated on form 3541. (Include advertiseršs proof and exchange copies)/Average number copies each issue during 12 months: 18,122; number copies of single issue published nearest to filing date: 16,938 b)2. Paid in-county subscriptions (include advertiseršs proof and exchange copies)/average number copies each issue during preceding 12 months/number copies of single issue published nearest to filing date: n/a. b)3. Sales through dealers and carriers, street vendors, counter sales and other non-USPS Untitled-6 paid distribution/average number copies each issue during preceding 12 months: n/a, number copies of single issue published nearest to filing date: n/a. b)4. Other classes mailed through the USPS/average number copies each issue during preceding 12 months: n/a, number copies of single issue published nearest to filing date: n/a. c) Total paid and/or requested circulation [sum of 15c. (1), (2), (3) average number copies each issue during preceding 12 months: 18,122 number copies of single issue published nearest to filing date: 16,938 d) Free distribution outside of the mail (carriers or other means)/ average number copies each issue during preceding 12 months: 1170; number copies of single issue published nearest to filing date: 1042. e) Total free distribution (sum of 15d. and 15e.)/ average number copies each issue during preceding 12 months: 1170, number copies of single issue published nearest to filing date: 1042. f) Total distribution (sum of 15 c and15e)/ average number copies each issue during preceding 12 months: 19,292 number copies of single issue published nearest to filing date: 17,980 g) Copies not distributed/ average number copies each issue during preceding 12 months: 20, number copies of single issue published nearest to filing date: 20 h) Total (sum of 15g and h)/ average number copies each issue during preceding 12 months: 19,312 number copies of single issue published nearest to filing date: 17,980 i) Percent paid and/or requested circulation (15c divided by 15g times 100)/ average number copies each issue during preceding 12 months: 93.9%, number copies of single issue published nearest to filing date: 94.2% 16) Publication of statement of ownership. Publication will be printed in November issue of this publication. 17) Signature and title of the editor, publisher, business manager or owner: Sandra Sillion (Managing Editor), Date: 10/3/2013. I certify that all information furnished on this form is true and complete. I understand that anyone who furnishes false or misleading information on this form or who omits material or information requested on the form may be subjected to criminal sanctions (including fines and imprisonment)and/or civil sanctions(including multiple damages and civil penalties). Sandra Sillion, Managing Editor

Solid or Spin... we go both ways

Ruggedized VPX Drive Drii v e Storage St Module Whatever your drive mount criteria, criteria everyone knows the reputation reputation, value and endurance of Phoenix products. The new VP1-250X, compatible with both solid state or rotating drives, has direct point-to-point connectivity or uses the PCI Express interface with the on-board SATA controller. It is available in conduction cooled (shown), conduction with REDI covers (VITA 48) and air cooled configurations.

We Put the State of Art to Work



8/6/12 11:56 AM

The magazine of record for the embedded computing industry

SUBSCRIBE FREE ONLINE • Spotlighting Trends & Breakthroughs • Design, Development and Technology of Embedded Computers • Archived Editions • Latest News in Embedded Community




Form Factors? “We Don’t Need No Steenking Form Factors!”


isruptive technology is notoriously hard to predict. If it were easy, it wouldn’t be so disruptive. Still, there appears to be a development on the horizon that could shake up the way a portion of the small form factor segment is made and marketed, especially when it comes to small, computationally very powerful mobile devices that are increasingly connected to the Internet of Things. With wirelessly connected 32-bit computer systems that include hefty memory, flash and rich complements of I/O now getting down to the size of credit cards, is there really a need for a specific form factor? At the same time, no one board module fits all needs; hardly any will use all the possible signals and connections that are available on today’s CPUs. Rather, they will want to optimize cost and board space by using just the USB ports, PCIe lanes, GPIO signals, A/D and D/A channels needed by the application and its potential enhancements. Trying to find a credit-card size module that fits all kinds of different requirements could be a real challenge. At the same time, the COM model of having the CPU, memory, etc., on the COM connected to a custom carrier board is less attractive when the goal is to fit everything into a small handheld package. The alternative is to identify the CPU and other components needed, verify a design with some sort of prototype, and then design the actual circuit board (small though it may be) from scratch. But what if there were another path for the OEM, one supported by the semiconductor manufacturers? What if semiconductor vendors made low-cost prototyping boards available for their chips that did bring out all the signals? Developers could then more easily create a prototype, develop the software and verify their design using only the subset of the available signals they need. But wouldn’t they then have to go off and design their specific board from scratch? Yes, unless . . . Unless they had access to the complete hardware design including schematics, bill of materials and even the CAD files for producing the circuit board. Then they could take their proven prototype and start with the whole design of the board, which they could then modify to remove and/or add the features required by their application, giving them a huge head start in producing the completed product. This form of open source hardware is starting



Tom Williams Editor-in-Chief

to appear in what is usually regarded as a segment for hobbyists, but it has potentials far beyond mere tinkering. There are indications that such a thing could happen in the form of what I am calling “dot-org” boards. These include such things as the ARM-based BeagleBoard line and the Raspberry Pi along with x86-based MinnowBoard and Gizmo Board. There is more detail about these and other products in this issue’s “Editor’s Report.” The boards and some of the supporting kits range in price from $35 to about $199 and appear to be priced and targeted for the hobbyist. They variously come with open source distributions of Linux and Android, Ubuntu and sometimes with an RTOS. The x86 variants can also load Windows. But the real value here is not trying to produce a credit card-sized PC; it lies in the potential for embedded development. They are also associated with user communities who blog and share ideas, designs and support tips. A growing number of user designs are freely available on the Web. At this point, the jury is out as to whether this “dot-org” board phenomenon will grow into a disruptive technology. That, however, could very well happen if one or more large semiconductor vendors see value in offering a selection of their processors that are appropriate for such applications on such low-cost prototyping platforms supported by the full open software/open hardware movements. These would of course have to be CPUs that meet the size, weight and power (SWaP) and cost characteristics for use in small, mobile applications. Alternatively, an entrepreneur or two could contract with semiconductor vendors to offer lines of such products. Again, no one expects that a “dot-org” board trade will destroy anything; if it happens it will fill a need that has not otherwise been satisfied. There is of course still plenty of need for the various small form factors, SBCs, COM modules, backplane designs and the wealth of other established technologies. Development decisions are still driven by cost, expected volume, “build or buy” and a host of other considerations. It just appears from this perspective that there are rumblings of opportunity to help developers faced with a more recent set of decisions and options as we move further into a completely connected and intelligent world.


INSIDER OCTOBER 2013 Updated 2.0 Specification of the Multi-Platform Qseven Standard SGET, the Standardization Group for Embedded Technologies e.V., has refined its 2.0 version of the Qseven standard. The standard, originally developed by the Qseven consortium in 2007, has been adopted by SGET e.V. in 2012. Qseven is a well-established, legacy-free standard for technology-independent small form factor computer modules (COMs), which includes standardized thermal/mechanical interfaces. Since its 1.2 version, which has been well adopted in the market, Qseven is the only standard that supports x86 and ARM technologies in COMcompatible environments. For developers of small form factor devices, this is a strategic and cost-saving benefit, as the decision for one or the other hardware does not affect the mechanical design of the devices. This enables riskfree changes between the platforms and a wider scalability by means of cost, performance and specific features. The updated Qseven specification (2.0 Specification and 2.0 Errata sheet) will be freely available for download on the SGET website according to the SGET terms of membership—both for Qseven developers as well as for users and carrier providers. Design Guides for the multi-platform releases 1.2 and 2.0 will be available this summer. Supported processor platforms for Qseven include ARM technology from Nvidia, Freescale and Texas Instruments as well as x86 technology from Intel and AMD. Other companies in the embedded computing industry are invited to join the SGET e.V. and contribute their ideas. Apart from embedded computing manufacturers on board and system level, chip and connector manufacturers, research and educational institutions as well as embedded system integrators, OEM solution providers and industrial users are also most welcome. SGET e.V., Munich, Germany. +49 (931) 418-3101. [].

Wireless Portable Plug and Print Patent Pending Printer for Smartphone/Digital Cameras

DoMark International, a consumer electronics company focused on designing, developing and marketing smartphone and tablet accessories, has acquired the IP and patent application for a unique cartridge printer for use with smartphones and tablets through DoMark’s associate company, Imagic. The Cartridge Printer—to be known as The Imagic Cartridge Printer—will complement DoMark’s product range of smartphone and tablet accessory products. The printer will allow durable fade-free photos to be printed from any smartphone or tablet simply by connecting wirelessly and printing, without the use of messy dye ribbons, ink, or sheets of chemically treated paper.



The plug-in renewable paper cartridge offers twenty 6x4inch post card sized prints of 100% digital reproduction quality. The cartridge is simple to replace and will be available online or at local stores. DoMark says the Imagic Printer can fill an important gap in the market by providing a combination of immediacy, quality, privacy and ease of use not delivered by PC desktop home printer or local store options.

Altera and Micron Lead Industry with FPGA and Hybrid Memory Cube Interoperability

Altera Corporation and Micron Technology have announced that they have jointly demonstrated successful interoperability between Altera Stratix V FPGAs and Micron’s Hybrid Memory Cube (HMC).

This technology achievement enables system designers to evaluate today the benefits of HMC with FPGAs and SoCs for next-generation communications and high-performance computing designs. The demonstration provides an early proof point that production support of HMC will be delivered with Altera’s Generation 10 portfolio, in alignment with market timing, and includes both Stratix 10 and Arria 10 FPGAs and SoCs. HMC has been recognized as the long-awaited answer to address the limitations imposed by conventional memory technology, and it provides ultra-high system performance with significantly lower power-per-bit. HMC delivers up to 15 times the bandwidth of a DDR3 module and uses 70 percent less energy and 90 percent less space than existing technologies. HMC’s ab-

stracted memory allows designers to devote more time leveraging HMC’s features and performance and less time navigating the multitude of memory parameters required to implement basic functions. It also manages error correction, resiliency, refresh and other parameters exacerbated by memory process variation. Micron expects to begin sampling HMC later this year with volume production ramping in 2014.

ARM Acquires Advanced Display Technology from Cadence

ARM and Cadence Design Systems have announced that the companies have signed a definitive agreement for the sale and transfer of Cadence PANTA display controller cores to ARM. The agreement enhances the companies’ long-standing ecosystem collaboration and strengthens their technical alignment. Cadence’s PANTA family of high-resolution display processor and scaling coprocessor IP cores was co-developed in conjunction with ARM and is targeted at advanced multimedia applications for high-end mobile devices with ultra-low power consumption. “Display technology is critical to the mobile consumer’s user experience,” said Pete Hutton, executive vice president and general manager, Media Processing Division, ARM. “The addition of the PANTA family of display cores to the ARM product portfolio will help our ecosystem of partners get to market quickly with high-end displays that are fully integrated with ARM’s leading Mali graphics and video solutions and protected with ARM TrustZone security.”

Martin Lund, senior vice president of Cadence’s IP Group, said, “ARM and Cadence work together closely on many levels, including IP integration, verification IP (VIP) for all ARM AMBA protocols, and high-performance design solutions optimized for ARM cores. As a result, both companies offer more tightly integrated solutions to our mutual customers.”

e-con Systems Forges Closer Ties with Cypress Semiconductor

e-con Systems, an embedded design services company specializing in development of advanced camera solutions, is a Cypress Design Partner at silver membership level. e-con Systems has worked with Cypress on its EZ-USB FX3 USB 3.0 peripheral controller from its launch and developed a USB UVC stack that is compatible with the Windows and Linux Operating Systems. e-con has developed products around the FX3 solution, leveraging its expertise in CMOS image sensors and understanding of USB UVC protocol. The Plug-n-Play UVC stack developed around the FX3 firmware provided by Cypress does not require any drivers to be installed on the host side and works seamlessly with any DirectShow(Windows)/ V4L2(Linux)-compliant applications for accessing and controlling the camera. By leveraging the highly configurable 32-bit, General Programmable Interface (GPIF II) 2.0 of FX3, e-con has developed an FX3 firmware framework that can interface any video source (CMOS/CCD image sensor, Thermal imager, NTSC/

PAL video source, HDMI/DVI video source) to a PC host on both USB 3.0 and USB 2.0 interfaces. e-con Systems’ See3CAM - USB 3.0 camera series, uses the Cypress FX3 controller and the first product of this family, an 1.3MP Global Shutter USB 3.0 UVC camera, See3CAM_10CUG has been already launched in the market. e-con has also customized its FX3-based USB3.0 camera designs for various customers for meeting their requirements.

GrammaTech Selected by the U.S. Navy to Improve Software Security

GrammaTech, a software developer specializing in software assurance tools and cybersecurity solutions, has announced that it has been selected by the U.S. Navy to develop a tool that will provide computer systems with the ability to understand and react to malicious attacks and then continue running safely. Protecting software from such attacks continues to be a challenge for critical systems. Since misbehaving software is not characterized by some universal pattern, it’s difficult to actively monitor systems to detect breaches and respond to them. In this project, GrammaTech researchers will use a combination of automatic program analysis and manual tuning techniques to develop a tool for creating a model of a system’s intended behavior, capturing its most important properties and determining what low level events must be tracked in order to observe the system’s critical behavior. “An important aspect of this tool is that it will be easy for developers to use,” stated Tim Teitelbaum, GrammaTech’s CEO. “As

the developer codes, the tool will capture his or her notion of what behavior is expected by creating a model that specifies a boundary the application shouldn’t cross. Our runtime monitors will then look for any unexpected behavior and take corrective action, even if the application has been compromised.” The development of this tool will provide security-critical systems with an extra layer of protection against attacks, including attacks that don’t involve unusual system call activity. The technology will be immediately useful to branches of the government, financial institutions and any companies whose systems require strenuous security protection.

Renesas a New Board Member to Drive Emerging Multicore and ManyCore Standards

The Multicore Association, a global non-profit organization focused on developing standards that help speed time-to-market for products that involve multicore implementations, has announced that Renesas Electronics has joined the consortium as an executive board member. “Renesas is a leading innovator in the areas of microcontrollers and SoCs for applications spanning automotive, audio-visual devices, communications and factory automation,” said Multicore Association President, Markus Levy. “Renesas involvement with the Multicore Association, especially focused on our new Software-Hardware Interface for Multi-Many Core (SHIM) standard, will provide very well-rounded guidance and a practical perspective on designing with multicore technology.”

“Renesas strongly supports the MCA and its mission to guide the long-term development of open standards for multicore solutions,” said Dr. Fumio Arakawa, Chief Professional, CPU Development Department, Core Technology Business Division, 1st Solution Business Unit of Renesas Electronics Corporation. The Multicore Association is working to enable the widespread adoption of multicore processor-based implementations by setting standards for how systems will be utilized and programmed. As a semiconductor design innovator, Renesas will help the Multicore Association drive standards for deploying multicore solutions in highperformance embedded systems. The Multicore Association provides three levels of membership, with executive board, working group and university members. The executive board helps determine the overall direction of the organization. Working group members are eligible to work on any of the working groups. University members may participate in any of the Multicore Association’s ongoing development work.




FORUM Colin McCracken

Bay Trail Closes the Gaps


t’s a new season for embedded x86 processors for intelligent systems. Now that select SKUs of the fourth generation Core i-series “Haswell” family have already found their way onto small form factor (SFF) embedded boards with long lifecycles, the new Atom “Bay Trail” platform has secured Linux and Android OS ports to leap over from some tablet design wins into automation, medical and communication edge markets. Lower thermal design power (TDP), faster RAM interface and better graphics deliver Bay Trail from the shackles of previous Atom generations. For SFF board manufacturers who’ve grown weary of supporting every generation of mobile-class Core i-series and Atom processor on the roadmaps, relief is finally on the way in terms of a processor that spans the entire range of previous Z-series, E-series, and even N-series families all with a single PCB design. Similarly at the mid-range and high end, only two PCB designs with a single board support package (BSP) allow Haswell to cover greater performance scalability than ever before. At last, system OEMs who seek an Intel solution aren’t staring at huge gaps at the low end and at the mid-range, which ARM and AMD have been exploiting. The gaps are closing. At the high end, Core i-series processors are always the heirs to the throne just by being the next step in the roadmap. As long as an OEM is careful about the power supply and thermal designs, it’s smooth sailing with abundant processor performance and resources. The situation for entry-level 32-bit and 64-bit processors could not be more different, though. Low-end processors live in a world of tight constraints. Competing requirements for size, weight and power, and yes, cost (SWaP-C) tend to disqualify all but a very limited set of processors for any given embedded system. Besides energy-efficient processor architectures, SWaP-C necessitates fewer I/O hooks such as PCIe lanes, SATA 2 ports and USB 3.0 ports. Even ARM system on chips (SoCs), known for low power consumption, serve up power-hungry I/O only sparingly in favor of diet-friendly serial ports and SPI and I2C buses. In this class, easy connectivity and low power are more important than raw bandwidth. Recently, AMD has come on strong in the low-end and mid-range with G-series “eKabini” Fusion SoC processors and



R-series “eTrinity” processors. Benefitting from the expensive acquisition of ATI Technology years ago, AMD has supported these great integrated GPUs with DirectX 11, OpenGL, and even OpenCL to use GPU cores for graphics or for code that can be broken down into parallel processing tasks. Intel was caught off guard this spring by AMD’s G-series SoC announcement, but has focused its extensive resources to close the gap. AMD needs to consider how to respond, perhaps in the pricing domain. For entry-level applications that need only modest graphics, G-series is somewhat overkill. SoCs based on the 32-bit ARM architecture dominate the low end, and are more than just a threat to Intel. These ARM parts keep inching up in performance each generation and now even outperform their more expensive Atom counterparts in some cases. Although from a features and product development standpoint, these processors are apples and oranges. Linux serves as the great equalizer. Software engineers can evaluate the performance of Linux on any cheap off-the-shelf board and influence the project direction or make decisions outright… it doesn’t even matter as much what the processor architecture is at the low end, as long as it runs Linux. Developers are no longer locked into the x86 camp or into the RISC camp according to the OS choice. OEMs who want an Android-style GUI have more processor choices available to them as well. Bay Trail is about more than just coming to a holiday sale near you. It’s a game changer for Intel in the embedded space. The tiny 22nm geometry with Silvermont microarchitecture enhancements affords Bay Trail better graphics, more RAM, performance scaling, and all for the “low, low price” of several watts. Finally. After previous Atoms were hamstrung to avoid cannibalizing lucrative notebook PC business, Intel has carefully considered market feedback and competitive pressures. The resulting processor family is truly a gem.

Got Tough Software Radio Design Challenges?

Unleash The New Virtex-7 Onyx Boards! Pentek’s OnyxŽ Virtex-7 FPGA boards deliver unprecedented levels of performance in wideband communications, SIGINT, radar and beamforming. These high-speed, multichannel modules include: ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡

A/D sampling rates from 10 MHz to 3.6 GHz D/A sampling rates up to 1.25 GHz Multi-bandwidth DUCs & DDCs Gen3 PCIe with peak speeds to 8 GB/sec 4 GB SDRAM for capture & delay Intelligent chaining DMA engines Multichannel, multiboard synchronization Ž ReadyFlow Board Support Libraries Ž GateFlow FPGA Design Kit & Installed IP Ž GateXpress FPGA - PCIe configuration manager OpenVPX, AMC, XMC, PCIe, cPCI, rugged, conduction cooled ‡ Pre-configured development system for PCIe ‡ Complete documentation & lifetime support

With more than twice the resources of previous Virtex generations plus advanced power reduction techniques, the Virtex-7 family delivers the industry’s most advanced FPGA technology. Call 201-818-5900 or go to for your FREE online Putting FPGAs to Work in Software Radio Handbook and Onyx product catalog.


EDITOR’S REPORT “Hobbyist” Boards Could Change Development

“Dot-Org” Boards Signal a Shift in Development Strategies Open source hardware joined with open source software may be signaling a shift in the way embedded development will take place. Completely available design data for prototyping, modification and manufacturing is attracting ever more attention. by Tom Williams, Editor-in-Chief


he prototype. It doesn’t have to be pretty. The board doesn’t have to be the right size and shape. It can have wires hanging off it. It can have instruments attached to it. But it better work. A working prototype is what we all know we need to show management to get the goahead for investment in the manufacture of a real product. Getting there can take many paths from software simulation to all kinds of cobbled-together hardware, or it can happen with a development kit supplied, for example, by a semiconductor manufacturer. Such development kits can be fairly inexpensive and are attractive. Quite often, however, you go out and spend $800 on a board product that closely (or not) resembles the system you have in mind and go from there. Of course at that point you lack schematics, a bill of materials (BOM) FIND the products featured in this section and more at



and a lot of other things that then must be done from scratch to get a design done. For a lot of developers, however, there may be some other ways opening up. Quietly in the background there have been a number of small board-level products appearing over the past five or so years that appear at first glance to be aimed mostly at hobbyists, but which have the potential to also be of great help to the “serious developer.” They have whimsical names like “Raspberry Pi, BeagleBoard, MinnowBoard or Gizmo Board.” And they appear to have attracted dedicated, nonprofit communities around them that share ideas and designs they have done. In addition, many of them are available as open source hardware platforms, which come with large offerings of open source software, including full versions of Linux. One is tempted to call these “dot-org boards” due to the online communities that have grown up around them. For all that, they come with serious processing power in terms of ARM or

x86 CPUs—some hitting the Gigahertz range—generous amounts of onboard memory, and with signals and I/O brought out that go well beyond what one would need for PC functionality. You can do serious embedded design with most of these. And the price is right. A number of these are even available over Amazon for under $50.

Small, Cheap and Very Powerful

One of the better known of these, which is marking its fifth anniversary, is the BeagleBoard. The original BeagleBoard, designed by Texas Instruments and now manufactured by Circuitco, is USBpowered and based on TI OMAP3530, an ARM Cortex-A8 CPU running at 600 MHz with 128 Mbytes of LPDDR RAM and 256 Mbytes of NAND flash with connections for DVD-I, I2C, JTAG, USB 2.0, RS-232, S-Video and more. A later version, the BeagleBoardxM, adds to all the features of the original more memory (512 Mbytes) a 1 GHz ARM processor, a 4-port hub with 10/100 Ethernet, more USB ports and HD video. Then TI moved from this more multimedia orientation to the BeagleBone series—the latest named BeagleBone Black—which is oriented more toward control applica-

FIGURE 1 The latest in the Beagle line, the BeagleBone Black includes a 1 GHz ARM Cortex-A8 and 512 Mbytes of RAM along with two 46-pin headers for an abundance of control signals, as well as a 3D graphics accelerator, all for $45.


tions with two 46-pin headers to bring out digital I/O signals. In keeping with its orientation toward small embedded applications, the BeagleBone is based on the ARM Cortex-M3 for power management. Finally, the latest offering, the BeagleBone Black has a 1 GHz Cortex-A8, 512 Mbytes of RAM and a 3D graphics accelerator (Figure 1). You can get it from Amazon for $49. Another popular ARM-based dot-org board is the Raspberry Pi, which was designed and is supported by the Raspberry Pi Foundation. It is based on a Broadcom SoC, the BCM2835, which contains an ARM 11 core and a Videocore 4 GPU capable of BluRay quality playback and 512 Mbytes of RAM. Although the Raspberry Pi does not appear to have a corporate origin, it has a very loyal following among its Foundation members and is available in the range of $35. There are also dot-org boards available in the x86 world, primarily from Circuitco and AMD. Circuitco, which as noted now manufactures the BeagleBoard line, also makes the MinnowBoard, which is based on a 1 GHz Intel Atom E640— with hyperthreading and virtualization technology—and the EG20T controller hub. The Atom processor also has an integrated graphics media accelerator. The MinnowBoard also boasts 1 Gbyte of DDR2 RAM and 4 Mbytes of SPI flash. It brings out a SATA2 connection, DVI via HDMI, analog audio, USB ports, a debug serial to USB conversion via a mini-USB port, and a variety of GPIO signals, PCI Express, Gigabit Ethernet and more (Figure 2). The latest and perhaps most ambitious entry into this arena is the Gizmo Board based on an AMD Embedded GSeries APU, a dual-core x86 design with an on-chip integrated general-purpose graphics processing unit (GPGPU). The Gizmo Board is part of the Gizmo Explorer Kit available from GizmoSphere, a consortium of five sponsoring companies that include AMD, Sage Electronic Engineering, Viosoft, Texas Multicore, Timesys and The board will be coupled with the AMD

FIGURE 2 The MinnowBoard allows development with an Intel Atom E640 running at 1 GHz and comes with a Linux distribution that is Yocto Project compatible. Expandability is available via user-designed MinnowBoard Lures.

embedded roadmap as processors become available. The current APU is rated at 52.8 GFLOP and a host of signals are brought out to two edge connectors, one high speed and one low speed. This includes, among others, SATA, USB, Display Port, PCIe, SPI, GPIO, PWM plus A/D and D/A. In addition there are onboard JTAG, VGA video, audio input/output, Ethernet and USB ports. The kit also contains an Explorer expansion board that can connect to the low-speed connectors. Here the user can have access to the signals by soldering components into a “sea of holes” and use of a small alphanumeric keypad and display. The kit also contains a Sage SmartProbe JTAG tool, a pre-installed copy of the SageBIOS and a 30-day trial license for their embedded development kit. A one-year extension is available for $299. A full commercial license is also available. The full Gizmo Explorer Kit is priced at $199. A GizmoSphere spokesperson explained that they aimed to price it at a point where a wife would not complain if her husband wanted to buy one and determined that more than $200 would be too expensive (Figure 3).

Open Source All the Way

That last little pricing observation points the way to what may be a very major development in developing embedded systems. If these boards appear to be appealing to the hobbyist, well, they are. But it goes beyond that because hobbyists alone do not comprise a significant market to justify this effort. However, the same tinkering mentality and often the same individuals are at work in major companies on designing new products or are thinking about taking their ideas to a start-up. And there are communities of bloggers associated with each of these where one can find help, comments, tips and ideas. That appears to be coming from a growing convergence of open source software with open source hardware. With the exception of the Raspberry Pi, we are talking about both here. And a little later we’ll look at an innovative variant in the form of a company called Gumstix. For the Beagle line, the MinnowBoard and the Gizmo Board, all the hardware documentation is freely available to the community including BOM, schematics, spec sheets and even the CAD files such as Gerber files for actually manufacturing the boards. RTC MAGAZINE OCTOBER 2013



FIGURE 1 The Gizmo Explorer Kit from Gizmosphere brings a host of useful embedded design signals out of the AMD G-Series accelerated processing unit (APU), which also integrates a powerful graphic/ numeric processing element. A variety of extensions and development tools are included with the kit.

So not far in the future, that clunky prototype will look a lot more elegant. A design manager can easily justify a few boards or kits for the team and they can get started with the selected CPU. The choices are fairly limited today but don’t expect that to last. There is also a wealth of software available, primarily several open source distributions of Linux but also Android and some RTOSs. QNX Neutrino is available for the BeagleBone Black for example. Of course, if they really need to use commercially available licensed software, there should be no real barrier either. So now the development team gets its prototype working and of course it is not the desired size or shape and there are a number of features available that are not used. Hardly anyone would use all of them. But they don’t have to start from scratch and design a new printed circuit board. Nor do they have to shop around to find a commercially available OEM board that most closely fits (but not quite) their needs. They can simply go to the CAD files and modify them saving work and optimizing their design. The communities that are growing up around the dot-org boards are adding value as they go by making designs available such as add-on peripherals like displays, camera interfaces, network interfaces, Wi-Fi and GPS modules and the



like. For the Beagle line they are called capes and for the MinnowBoard they are called lures. There is also a growing variety of designs that people are simply proud enough of that they want to share them. We mentioned an innovative variant who also needs to be included in this discussion and that is Gumstix. Gumstix has a line of board-level products built around its line of ARM-based COMs called Overo and DuoVero based on TI ARM Cortex-A8 (Overo) and Cortes-A9 (DuoVero) processors. Gumstix does not disclose the internal design of the COM modules beyond the external pin-out, but the designs of its boards based on these COMs is available to the community. But beyond that, user-modified designs based on Gumstix original designs are also available. Gumstix also has an online tool named Geppetto that allows users to take an existing design from the community database and modify it with exactly the features, size and shape desired for a given project. The user selects a design, clicks on “clone” and it is loaded into Gepetto for customization. Then for a $1,999 setup cost, Gumstix will verify the design, create the CAD files and either manufacture the boards or make the files available for contract manufacturing. Here again, developers do not start from scratch. Gumstix has also come up with a crowd funding plan that lets a number of participants form a collective effort to pool their money to purchase a product or service together, thus reducing the individual cost. Users create community campaigns for any electronic device designed with Geppetto, whether or not the design is shared publicly. Designers choose a quantity threshold and campaign duration (between 10 and 45 days) and share a catalog link with their community. Community members then pledge to buy boards, and after the threshold is reached the boards are built. Users divide the set-up fee based on the quantity of boards ordered, and as a part of the launch, Gumstix will waive the fee entirely for campaigns that reach over 50 subscribers. This dot-org board development is definitely something to watch. It has actually been around for a number of years, scampering around like small rodents

among ponderous T-Rexes. If it catches on with semiconductor manufacturers, it could have a broad impact on the embedded development OEM board business. AMD Sunnyvale, CA. (408) 749-4000. []. Circuitco Richardson, TX. (214) 466-6690. []. Gumstix Redwood City, CA. (650) 206-2464. []. Sage Electronic Engineering Longmont, CO. (303) 495-5499. []. Texas Instruments Dallas, TX. []. Texas Multicore Austin, TX. (512) 381-1100. []. Timesys Pittsburgh, PA. (412) 232-3250. []. Viosoft Ashland, MA. (508) 881-4254. []. []. []. []. []. [].



Small Modules Go Rugged

VITA 74: Small, Conduction-Cooled, Rugged and Modular The VITA 74 standard brings the modularity of the VME and VPX formats to smaller systems that previously would have had to be custom built, saving time and cost. by Wayne McGee, Creative Electronic Systems


he VITA 74 committee was chartered by the VITA Standards Organization in May 2010 and soon attracted a diverse group of suppliers and potential customers for the new standard. The sponsors include Bustronics, CES, Molex, Samtec and Themis. The standard will consist of four specifications: the base specification and three clarifying specifications. VITA 74.0 is the base specification and the primary definition of the vision. The specification describes a plug-in module designed for backplane systems with one or more modules communicating over the backplane. I/O is routed via the backplane to connectors on the enclosure. VITA 74.1 describes a stand-alone enclosed computer module with cabled I/O. This is a special case implementation of the 74.0 specification. This specification has not yet been released. VITA 74.2 is the 74.0 rear transition module specification. This specification allows the user I/O to be brought out of the enclosure to standard connectors. This specification is intended to promote a development environment prior to the purpose-built backplane likely required for deployment. This has not yet been started. VITA 74.3 is a CPU module definition and is an optional standard I/O proposal to promote consistency and interchangeability of CPU modules. While de-



FIGURE 1 An example of a single-wide VITA 74 module.

sirable from a user standpoint, the widely varied needs for different I/O will require multiple profiles for varying classes and architectures of processor modules. This has not yet been started. We will now examine the base specification. The standard defines two different form factors for the modules. Both modules are 89 mm x 75 mm but differ in thickness (card pitch). They both support backplane pins for location and ESD protection. The 12.5 mm variant uses a single base card and a 4-row, 200-pin connector. This is intended to support peripheral functions such as I/O, GPS and storage and is shown in Figure 1. The 19 mm variant allows for a base as well as a NanoETXexpress-sized mezzanine card. It uses an 8-row, 400-pin connector and is intended to support single board computer, graphics, video and FPGA functions and is shown in Figure 2.

As the standard was being defined, deliberate efforts were taken to incorporate previously proven technology. The signaling methodology was similar to the VITA 46 (VPX) and VITA 65 (OpenVPX) standards. The connectors used were adopted from the VITA 57 (FMC) standard. FRU inventory records for discovery over IPMI were reused from VITA 42 (XMC). The 19 mm module is wide enough to permit use with the PICMG COM.0 revision 2.1 NanoEXTexpress specification, although this is not part of the VITA 74.0 specification. In designing small form factor modules and systems, careful consideration must be given to power dissipation and heat removal. VITA 74 modules are conduction-cooled, with 19 mm modules limited to 20W and 12.5 mm modules limited to 10W. Later revisions to the specification may allow these limits to be raised. The mechanical design utilizes a typical

FIGURE 2 An example of a double-wide VITA 74 module.


shown in Figure 5, a cubic configuration. The computing power provided by such a system, limited mostly by the power consumption, would be comparable to that of a tablet computer.

Applications and Deployment Benefits

FIGURE 3 An expanded view of the double-wide module with heat spreader and case.

skyline interface heat spreader with thermal gap pad material to remove heat from the chips. The heat spreader then connects to the module case, which provides additional heat spreading capabilities on three sides. An exploded view of a typical 19 mm module is shown in Figure 3. This same technique is applicable to both the VITA 74.0 and 74.1 modules. The external cooling provided to the system should be designed to keep the temperature of the module case at the heat exchanger thermal interface at or below 85째C. The backplane connector is a highspeed, rugged Samtec Searay series, the same as was used for VITA 57. This connector is not proprietary and is available from other connector manufacturers. The electrical bus partitioning is similar to VPX. There are three partitions. S0 is the utility plane containing power, control, clocks and management signaling. S1 is the bus fabric, which can be PCI Express Gen 1 or 2 with up to 16 lanes. S2 contains the user I/O and has 18 differential pairs or 36 single-ended lines in a full ground grid. The connector also provides separate locator pin/receptacles and ESD protection similar to VPX. An I/O transition module can be connected to the backplane to provide connectors for the various power and I/O signals needed for the system. An example of this is shown in Figure 4. When the various modules, the backplane and the I/O transition module are combined into a packaged system, the result could be much like the system

VITA 74 enables deployable systems with a smaller footprint and cost point than VME, cPCI or VPX. The increasing availability of standard modules in the growing ecosystem opens up many possibilities. Standard SBCs, video processor, storage and I/O modules are already available. Full systems with the functionality of a small PC, with additional video processing capability and adapted I/O, are already available as a rugged box, ready to be installed in any embedded or industrial environment. Thanks to the modular architecture, the processing functions, external interfaces and storage capacity can be combined into the optimal solution for each unit. Thanks to the small size and highperformance interconnects, multiple units can be installed in a distributed system, allowing system designers the freedom to decentralize their designs and provide dedicated subsystems with simplified and carefully segregated functionality. For safety-certified systems, this will make DO-178B certification easier to achieve. High-speed interconnects eliminate the need to have all functions in a single chas-

sis, eliminating one of the single points of failure. Placing the control function closer to sensors reduces fidelity issues with measuring low-level analog signals. Simplified dedicated controllers also make it easier to deploy redundant subsystems in critical areas. The VITA 74 system is smaller, lighter and less costly than a 3U VPX computer. This makes it ideal as a single or distributed subsystem controller for many different types of rugged vehicles. Figure 6 shows a typical distributed architecture for unmanned vehicles, whether flying, rolling or floating. There is a growing market worldwide for mini-UAVs that require smaller payloads. In addition to military usage, there will be increased deployment for security and border patrol as well as commercial uses such as aerial mapping. Transportation is another area of possible deployment. Buses and passenger cars for subways and trams in many places display what the next stop is and which connections will be available. Add a location-aware advertisement server, and the passenger is made aware of the merchants at the next stop while the carrier now derives a revenue stream from advertisements. The need for small rugged computers with a wide temperature range also exists in many industrial control applications. The growing domestic oil and gas market will see an increased need for down-hole

FIGURE 4 An example of a VITA 74 5-slot backplane with I/O transition module.




VITA 74: small size; low cost; low power; multi-GHz clock speeds; multiple cores, including dedicated DSP engines; highspeed DRAM interfaces; high degree of SoC peripheral integration; networking; peripheral interfaces; HD video capture, compression, decompression and display are highly prevalent in these processors. They even regularly endure environments that are comparable to many an embedded computing system. In summary, the VITA 74 specification brings to life a proven set of technologies to provide a superior size, weight and power (SWaP) performance package. Combined with an extended temperature range, extended shock and vibration capability and conduction cooling, the new specification brings enhanced computing power to a wide range of applications.

FIGURE 5 An example of how the VITA 74 modules and backplane can be packaged for deployment.

monitoring and data collection with the precision drilling controls needed to process shale oil deposits. For mining, many of the large vehicles contain multiple control systems and require an ultimate vehicle health monitor. In many cases, the airborne particulate matter makes non-sealed air-cooled computers unusable. VITA 74 systems are well suited for such an environment. Another application VITA 74 is well suited to address is 180° and 360° video capture. A quad input video capture card with four 90° cameras provides a complete situational awareness view. Many conventional police dash-cams are limited to a 90° forward view and lose sight of the officer if the person being stopped runs out of the field of view. A 360° system eliminates the evidential questions about conduct outside of the current field of view. While VITA 74 is a perfect fit for the ever increasing demand for smaller, more complex autonomous vehicles, the VITA 74 module itself is a perfect platform to take advantage of the processor revolution coming out of the commercial tablet and smartphone market. Demand for tablets and smartphones is predicted to rise to over 1 billion units per year by 2014 and this, in turn, is providing billions of dollars for mobile processor R&D. This



investment is clearly evident in complex System on Chip (SoC) processors like Samsung’s Exynos, Qualcomm’s Snapdragon, Apple’s A6, Texas Instruments’ Davinci and Intel’s Atom. And these never before heard of volumes, coupled with ever increasing commercial pressures, are driving these mobile processor platforms down to unheard of price levels. The predominantly ARM-based modern mobile SoC processor is a great feature match for

Creative Electronic Systems Geneva, Switzerland. +41 (0)22 794 74 30. [].

Mission Control

Processing Engine

Intelligent Display

FIND the products featured in this section and more at

Data Recorder

Sensor Fusion

Onboard Network

Smart Controller

Data Link

FIGURE 6 A typical distributed vehicle control architecture shown with optional intelligent display subsystem.

Intelligent Solutions Finder Search, Compare & Purchase Intel-Based Boards & Systems Featuring the MinnowBoard The MinnowBoard is an Intel® Atom™ processor based board which introduces Intel® Architecture to the small and low cost embedded market for the developer and maker community. It has exceptional performance, exibility, openness and standards. Find out more at

Purchase the MinnowBoard & many more products at: brought to you by


CONNECTED Wireless Connectivity for Mobile Devices

Using Wi-Fi to Connect Embedded Systems with Mobile Devices Wi-Fi modules are enabling a new generation of wireless embedded systems through Internet connectivity and the increased processing power of microcontrollers. Future module improvements will enable innumerable intelligent devices all over the world to communicate with each other and be controlled from smartphones. by Mitch Dale, Microchip Technology


mbedded systems are not limited to simple 8-bit controllers; many are sophisticated and self-contained computer systems. These advanced embedded systems can support color LED displays, vast data storage, significant computing power and myriad interfaces. One growing trend in these systems is to add wireless connectivity. The use of proprietary or standards-based radios to add wireless requires RF and protocol expertise. Unfortunately, most companies do not have the RF expertise or tools to develop their own wireless interface. Computers, smartphones and tablets comprise the majority of wirelessly connected devices. However, there is huge potential for the emerging market of embedded wirelessly connected clients, which is often called the “Internet of Things” or the “Wireless Connectivity of Things.” Some examples include temperature sensors, automotive diagnostics, blood-pressure monitors and many more. There are hundreds of applications within the smart home, energy grid, personal healthcare, medical and asset-tracking markets ready for the taking. While many wireless options exist, Wi-Fi has quickly become the de facto standard for wireless embedded systems because of its global acceptance and in-



FIGURE 1 Adding Wi-Fi to embedded systems enables them to tap into the vast installed wireless infrastructure.

teroperability. Additionally, with the availability of low-power, self-contained, industry-certified solutions, Wi-Fi fits into many markets; including those with mobile and battery-powered requirements (Figure 1).

Why Wi-Fi?

There are many forms of wireless communication, based on different frequencies, modulation techniques and protocols. However, Wi-Fi has several advan-




Bluetooth LE


Sub GHz

Internet-Connected Device






Smartphone Accessory












Low Power Consumption






High Data Rate






Interoperability Between Vendors






FIGURE 2 A comparison of the wireless options for embedded systems.

tages over the rest by directly interfacing to the World Wide Web. In this aspect, Wi-Fi leverages the Internet infrastructure, global connectivity and the interoperability of cloud services by exchanging data with computers or smartphones. Wi-Fi is a truly worldwide standard. Wi-Fi is ubiquitous anywhere you travel. There is an enormous established Wi-Fi infrastructure in place. According to the market research group Informa Telecoms and Media, the number of access points worldwide continues to grow by over 300% annually; and by 2015, there will be 5.8 million public hotspots worldwide. Another advantage of Wi-Fi is that people understand how to configure and use it. Unlike ZigBee or proprietary wireless protocols, the average user understands how to associate their mobile or embedded device clients onto a Wi-Fi network. Going forward, Wi-Fi should become even easier, as the Wi-Fi Alliance and smartphone manufacturers implement Hotspot 2.0 and mDNS. These extensions to the existing standards simplify the discovery, authentication and provisioning of devices on the Internet. Wi-Fi has achieved critical mass. The market for Wi-Fi devices is driven by the ever-expanding Internet and the prolifera-

FIGURE 4 Microchip’s RN-171 is an example of a pre-certified Wi-Fi module with onboard TCP/IP stack and services.

tion of smartphones and tablet connectivity. With Wi-Fi, devices can be connected anywhere, anytime. For example, a WiFi-enabled thermostat can be accessed anywhere from a cell phone or computer. The challenge for embedded designers is how to take advantage of the Internet infrastructure, Wi-Fi hotspots and cloud applications. Since each embedded Wi-Fi device can function independently, nodes anywhere on a corporate network or anywhere in the world can be accessed from a central server, via a URL. Alternative wireless options may have the advantages of range or power consumption, but they require a gateway to access the Internet. A gateway is an additional device that needs to be developed and supported. In addition to the added cost, it may also be a single point of failure in the network. Nevertheless, some applications, such as sensor networks, may use other wireless networks to configure, collect and forward data to the Internet. Even in these cases, gateways would likely include Wi-Fi, since it does not require the user to install additional Ethernet cables. While many wireless options exist, autonomous Wi-Fi embedded systems create new opportunities for innovation and expanded business models (Figure 2). For example, combining everyday products such as washing machines and air conditioners with wireless Internet connectivity creates easy-to-install, longterm servicing options for installers. Another example is the management of large trucking fleets using embedded Wi-Fi monitors. In this case, Wi-Fi allows reporting to happen whenever a truck enters the fleet parking lot or a distribution facility. Route, driving statistics and engine data are sent to a centralized server that manages scheduling, monitors driver safety and predicts maintenance problems. Previous fleet-management systems

were implemented with cellular radios, but they are being switched to Wi-Fi because it has no ongoing data charges.

Modules Make Embedded Wi-Fi Practical

Only a few years back, it was impractical for embedded system designers to implement Wi-Fi in small, power-sensitive devices. However, this all changed with the introduction of low-power, complete Wi-Fi system on chip components in 2010. This new generation of Wi-Fi silicon quickly became the building blocks for complete Wi-Fi solutions, in the form of certified modules that have antenna, RF, baseband and protocol stacks (Figure 3). Wireless connectivity requires both a hardware radio and a software protocol stack. In other words, the Wi-Fi alone is not sufficient for Internet connectivity. Modules come in two flavors, stack-onboard and stack-off-board. Stack-onboard modules include the processor on the module, or may have the processor integrated in the radio chip. These modules have a simple ASCII command interface for configuration, and natively support the majority of Wi-Fi networking protocols, such as UDP, TCP/IP, DHCP, DNS, TELNET, FTP, HTTP, XML, SSL, etc. Additionally, stack-on-board modules are the quickest approach to adding Wi-Fi, since they are simply a data interface to the embedded system. In other words, to send data over Wi-Fi to a server in the cloud, you only need to configure the IP address or URL to write the data to the module. The module takes care of the connection, packets and underlying Wi-Fi transport. On the other hand, stack-off-board modules contain the radio components and provide drivers for running the protocol stack on an external processor. Typically vendors will provide a protocol stack library. However, this approach requires more software integration, while allowing more flexibility. Stack-off-board modules typically support more types of peripherals and provide for additional Wi-Fi security and services. Microchip Technology Chandler, AZ. (480) 792-7200. [].





Optical Fiber Instrumentation

Fiber Optic Sensing Opens New Capabilities for Structural Analysis Beyond finite element analysis, fiber optic sensors can be used to give graphic analysis for complex structures and soon may even be embedded in composite materials for real-time monitoring of stress and strain during actual operations such as in aircraft. by Alex Tongue, 4DSP


nderstanding how engineering structures respond to loads and their environment is of paramount importance for their successful design and reliable operation. Analyzing the strains, stresses, temperatures and deflections in a bridge during rush hour, a composite aircraft’s wing spar during a storm, or a high-end bicycle during the Tour de France, is what allows engineers to predict a structure’s lifetime, increase its safety and optimize its performance. Armed with a complete picture of these quantities from the early design stages to the end of a product’s lifecycle, designers would undoubtedly create safer, stronger and more efficient engineering structures. 4DSP has teamed up with NASA Dryden to create a fiber optic sensing instrument capable of providing this information in real time. Their distributed sensing technology provides strains, temperatures, stresses, out-of-plane deflections and three-dimensional shape; all in a small and lightweight package. Fiber optic sensing has emerged in recent years as an effective solution for characterizing structures in harsh environments. Optical fiber is extremely lightweight, immune to electromagnetic interference, and is resistant to extreme tem-



FIGURE 1 4DSP RTS150 8-channel system.

peratures and corrosive chemicals. The sensing elements within these fibers are called Bragg gratings. These elements are fabricated using high-intensity UV laser light to locally destroy the silica-oxygen bonds within the core of single mode opti-

cal fiber. The disruption caused by these broken bonds slightly decreases the speed at which light travels through the core, or it increases its index of refraction. If this process is done periodically over a short distance along the fiber, it creates a Bragg


FIGURE 2 RTS150 Sensing Engineering Strain.

grating. The interesting and useful characteristic these elements exhibit is that they act as selective optical mirrors; they reflect a specific wavelength and allow all others to pass. The grating’s periodicity is what determines the wavelength that it reflects. Thus, when these fibers are stretched, compressed, or undergo thermal expansion, the periodicity of the grating and its reflected wavelength change proportionally. By using a light source to interrogate them and a detector to sense the reflected signal, Bragg gratings turn plain optical fiber into passive and robust strain gages and thermocouples. Bragg grating sensing systems have actually been around for quite some time. Off-the-shelf instruments combining grated fiber and reflected optical signal demodulation techniques such as wavelength division multiplexing or time division multiplexing are offered by several established manufacturers. However, severe processing limitations associated with these technologies have hindered their widespread adoption in industry. For example, wavelength division multiplexing requires that each grating on a single fiber reflect its own unique wavelength. To accommodate the expected shifts due to strain and temperature changes,

each grating must also be allocated its own wavelength band. This puts a limit on the number of sensors one can have on a single fiber since light sources have limited bandwidth. Time division multiplexing allows the gratings to be written at the same wavelength. However, it runs into extreme processing speed limitations with an increasing number of sensors on a single length of fiber. Low sensor count and sample rates, while useful in some scenarios, aren’t an effective solution for large scale distributed sensing. These roadblocks have been recently overcome through collaboration between NASA Dryden and 4DSP. Together they have developed and implemented a new processing technique that expands on the capabilities of existing Bragg grating technologies. Their efforts led to 4DSP licensing the patented technology and integrating their platform into an off-the-shelf product. Their RTS150 is a multichannel optical sensing instrument capable of interrogating a significantly larger number of sensors at higher rates than past Bragg grating systems (Figure 1). Utilizing the newly developed demodulation technique, engineers can now simultaneously monitor up to 65,536 optical strain gages and thermocouples, each at 100 samples per

second.This opens the door to a vast number of structural problems being solved by distributed optical sensing. The fundamental physical parameters that define the state of a structure are stress and strain. These are the quantities that determine if a bridge will collapse, whether a pressurized tank will burst, or how long an aircraft can continue to be in service. Traditionally, the stresses and strains in large and complex structures such as aircraft and spacecraft are modeled using various finite element methods (FEM). These are used to simulate different loading scenarios and engineers use their results to determine structural soundness. FEM models are also used to locate critical points to monitor with conventional foil strain gages during testing and operation. While finite element methods have been a groundbreaking technology in structural engineering, there are still some problems associated with using them. Modern structures are becoming increasingly more complex as new materials and technologies continue to be discovered and introduced to industry. Despite this increasing complexity, FEM models are often simply assumed to be accurate, and, if they are validated, it’s done using a few discrete points with strain gages. Any practicing engineer will know a lot of unexpected things can occur between any two discrete sensing points. Distributed optical sensing allows engineers to determine precisely what happens over a continuous sensing length and acquire finite element-like experimental data (Figure 2). Fiber can be applied along the entire length of a wing spar, wound around the perimeter of stress-concentrators such as doors and windows, or fiber grids can be applied to large planar sections to obtain a far more detailed picture of a structure’s actual behavior. With the ability to sense stress, strain and temperature at spatial resolutions less than a millimeter over entire structural components, this new technology provides the means for FEM models to be fine-tuned, and can help establish the utmost confidence in a model’s accuracy. Distributed fiber optic sensing also provides new solutions for carbon fiber composites. Their unique fabrication processes allow Bragg grating sensors RTC MAGAZINE OCTOBER 2013



to easily be embedded straight into the structure. This technology enables realtime embedded monitoring of highly flexible composite structures from the early design stages to the end of their operational lifecycle. Optical sensing’s small spatial resolution and high strain accuracy allows engineers to discover, locate, quantify and track the various complex failure modes carbon fiber composites exhibit. Anomalies like delamination between layers, matrix cracking, fiber breakage and fiber buckling all produce stress and strain concentrations at their origin. Because 4DSP and NASA’s new technology turns the entire length of fiber into one continuous strain gage, these concentrations would appear along the fiber’s output strain profile. As delaminations or cracks propagate under increasing loads or fatigue, the concentration point at the crack front can be tracked along the fiber and quantified at each point along the way. Optical fiber is also resistant to temperatures ranging from cryogenic to hundreds of degrees Celsius. Embedded fibers can be monitored during the high temperature cure phase of composite fabrication to determine the through-thickness residual stresses and strains. Continuous monitoring of composite structures could be a far reaching advancement in industry. Commercial airlines will always know the state of fatigue damage on critical components, allowing them to fine-tune maintenance schedules and minimize aircraft downtime. Composite rotor blades, skins, beams and pressure vessels across various industries can benefit from knowledge of the physical state of a structure or structural component during operation. This new technology also enables out-of-plane measurements. The majority of sensing instruments in the past have only been able to provide in-plane quantities. Strain gages and traditional fiber Bragg grating systems sense strain only along the primary directions of a structural component—along the length of a beam or in the plane of a plate or shell. From this, mechanical stress and planar deformations are readily obtained. However, there are very few solutions for ob-



taining distributed out-of-plane measurements such as deflections or applied loads, especially during operation of a structure. As industry continues to move toward lighter and more flexible structures, outof-plane information is becoming a critical factor in design and operation. Transfer functions have been developed for use with this new technology that utilize the high spatial density planar strain measurements along a fiber bonded to a structural component. These algorithms yield both out-of-plane deflections and applied loads, all in real time. When applied to a wing spar, engineers can simultaneously obtain the strain along its length and the deflection of the aircraft’s wing. This can provide information on flutter, natural frequencies and mode shapes, lifting loads and moments. Three dimensional shape sensing of a continuous fiber is another new application enabled by distributed fiber optic sensing. These shape sensors can accurately determine deviations from a straight line, or they can be manipulated into complex shapes with tight twists and bends with radii less than half of an inch (Figure 3). This unique application can

FIGURE 3 3D shape rendering.

be beneficial to such fields as petroleum engineering, where oil pipelines may contain blockages that need to be located and inspected, or in satellite technology where an antenna is undergoing complex bending in multiple directions. While the 2-D shape sensing solution requires the fiber optic cables to be bonded to an underlying structure, the 3-D shape sensor is an independent entity, and can be used by itself or integrated with an existing structure such as a catheter used for invasive surgery. With a diameter down to 450 micrometers, these sensors can be put through complex, tightly turning tracks like arteries, pipelines and mining boreholes to accurately determine the shape of their path. Human limb movement can be recorded far more accurately than current digital image correlation techniques, or the shape of the tether to an underwater vessel can be tracked and used to determine the vehicle’s location and whether or not the umbilical is wrapped around an unseen obstacle. The future of fiber optic instruments lies in miniaturization and speed. As the size of the interrogation system is reduced, the number of applications that


can use it increases tenfold. Arriving in 2014, the next generation product will be the size of a 5-inch cube, more than capable of fitting within a typical flight instrumentation box. With such a small weight penalty, the future may see every commercial airliner being constantly monitored by distributed fiber optic sensing. Or, a system may be installed in the housing of every wind turbine around the world, continuously monitoring blade shapes and structural integrity. In the long-term, 4DSP intends to reduce the size to the equivalent of a deck of playing cards. Another leap forward to be taken is an order of magnitude increase in sample rates. While the current rates have already left past fiber optic sensing technologies behind, 4DSP has its eyes on sensing high frequency vibrations such as those on rotor blades. Paired with the continuous sensing capabilities, high sampling rates would also allow engineers to study stress wave propagation in a whole new way. Distributed sensing offers a unique solution to a wide spectrum of engineering problems and appears to have a very bright future.

CompactPCI & Graphics/ Video Boards Gallery Featuring the latest in CompactPCI & Graphics/ Video Boards Product technologies ADLQM87PC - Industry’s Smallest 4th Gen. Intel Core i7 SBC • 4th Generation Intel® Core™ QUAD and DC Processors; PCIe/104; up to 8GB DDR3L • HD4600 GPU; Over 2x Performance Increase over 2nd Gen Intel Core i7 GPU • Onboard PCI-Express Mini Card Socket v1.2 Compatible with miniPCIe and mSATA Modules • Integrated Trusted Platform Module (TPM) v1.2 • USB 3.0; 4x SATA 6Gb/s; DisplayPort, eDP, HDMI, VGA, 2x GbLAN, 2x COM, 16-bit GPIO • Mil-Std 810 Ruggedization and -40°C to +85°C Options Available

ADL Embedded Solutions Inc.

Phone: (858) 490-0597 Web:

Schroff 450/40 ATCA Chassis Supports Next Generation ATCA Board Requirements • Minimize time-to-market for critical, high-availability applications • 2, 6, and 14 slot backplanes • 40 Gbps (10 GBASE-KR) transmission rate – bussed or radial IPMB • Increased power availability of 450W per slot • Enhanced cooling capabilities of 400W per slot in the front • Improved cooling capabilities of 50W per slot in the RTM • Both integrated AC or DC Redundant Power Entry Module options available in all slot counts

4DSP Austin, TX. (775) 830-2059. [].

Pentair Equipment Protection Solutions (PE) Schroff

Phone: (401) 732-3770 Web:

USB Wi-Fi Modules 802.11b/g/n Compliant • • • • •  • •  • • •

Radicom Research, Inc

USB 2.0 hot swappable interface Compatible with USB1.1 and USB2.0 host controllers Up to 300Mbps receive and 150Mbps transmit rate using 40MHz bandwidth Up to 150Mbps receive and 75Mbps transmit rate using 20MHz bandwidth 1 x 2 MIMO technology for exceptional reception and throughput 2 U.FL TX/RX antenna ports Wi-Fi security using WEP, WPA and WPA2 Compact size: 1.0” x 1.0” x 0.25” (Modules) Windows 2K, XP, Vista, Win7 support Linux 2.4/2.6 support Phone: (408) 383-9006 Web:



PCI Express, PCI, and ISA Experts RTD Designs and Manufactures a Complete Line of High-Reliability Embedded Products & Accessories

AS9100 and ISO9001 Certified 4JOHMF#PBSE$PNQVUFST t t t t t





t t


1FSJQIFSBM.PEVMFT t t t t t t t t t



t t t


Copyright © 2013 RTD Embedded Technologies, Inc. All rights reserved. All trademarks or registered trademarks are the property of their respective companies.


#VT4USVDUVSFT t t t t t

1$*F 1$*&YQSFTT 1$* 1$Plus 1$

The products above are just a sampling of RTD’s board-level and ruggedized packaging solutions. From low-power to high performance, RTD can tailor a system for your mission-critical application. Visit to see our complete product list.

AS9100 & ISO9001 Certified


TECHNOLOGY DEPLOYED Developing for Low-Power Systems

Championship ARM Wrestling: Tips for Getting the Most out of ARM Cortex-M3 & M4 Microcontrollers

features, along with performance and features that help products stand out in crowded markets.

ARM Cortex Basics

While by no means complete, this modest collection of tips and tricks should show how to exploit some of the Cortex-M series’ lesser-known features to your advantage in your next design. by Rasmus Christian Larsen, Silicon Labs


any embedded developers are familiar with the ARM Cortex processor architecture, but few have the opportunity to become intimately acquainted enough with this popular architecture to take full advantage of its unique features and capabilities. This is especially true for the new ARM CortexM4 processor, which boasts an improved architecture, native digital signal processing (DSP) capabilities and an optional floating-point accelerator, which a savvy programmer or hardware engineer can exploit to their advantage. Let’s take a closer look at some of the more interesting (and often overlooked) features found in Cortex-M3-based microcontrollers (MCUs) as well as in new M4 variants. Since many target applications for Cortex-M-based MCUs are portable and derive their power from batteries or energy harvesting systems, most of the ideas we will explore involve techniques for reducing a design’s overall energy con-



sumption. In many cases, however, these energy conservation techniques are also helpful tools for designing processoroptimized applications that provide more cost-effective solutions, more processing margin available for upgrades and new


Cortex-M4 Nested Vectored Interrupt Controller

Wake Up Interrupt Controller Interface

CPU (with DSP Extensions)

FPU Data Watchpoint

Code Interface Memory Protection Unit SRAM & Peripheral Interface

Much like the original 16-bit processor cores created by Advanced RISC Machines (ARM) in the 1980s, the ARM Cortex series is based on a Harvard-style RISC machine with a modest silicon footprint that enables high performance as well as code and memory efficiency. The architecture has evolved considerably over the past decade, branching into three distinct sub-families (or profiles) created to meet the requirements of a particular application space: • A-profile products are optimized for high-performance open application platforms. • R-profile processors include features for enhanced performance and relibility in real-time applications. • The M-profile processor series was developed for use in deeply embedded MCUs in applications where performance must be balanced with energy efficiency and low solution cost. Popular applications for the Cortex-M series include smart metering, human interface devices, automotive and industrial control systems, white goods, consumer electronics products and medical instrumentation.

Bus Matrix

Flash Patch & Breakpoint ITM Trace ETM Trace

Nested Vectored Interrupt Controller

Wake Up Interrupt Controller Interface CPU

Debug Access Port Serial Wire Viewer, Trace Port

Data Watchpoint

Code Interface Memory Protection Unit

Bus Matrix

SRAM & Peripheral Interface

FIGURE 1 Comparison of the Cortex-M3 and M4 Processor Cores.

Flash Patch & Breakpoint ITM Trace ETM Trace

Debug Access Port Serial Wire Viewer, Trace Port


ATCLK domain


ATB slave port

ATB interface

Asynchronous FIFO


APB interface

Register bank

Trace out serizlizer


PCLK domain

FIGURE 2 The dedicated ARM Cortex SWO interface saves I/O pins and speeds up debugging.

The Cortex-M3 vs. Cortex-M4 Story

The idea behind the Cortex-M3 architecture was to design a processor for cost-sensitive applications while providing high-performance computing and control. These applications include automotive body systems, industrial control systems and wireless networking/sensor products. The M3 series introduced several important features to the 32-bit ARM processor architecture including nonmaskable interrupts, highly deterministic, nested, vectored interrupts, atomic bit manipulation and optional memory protection (MPU). In addition to excellent computational performance, the CortexM3 processor’s advanced interrupt structure ensures prompt system response to real-world events while still offering low dynamic and static power consumption. The Cortex-M3 and M4 processors (Figure 1) share many common elements including advanced on-chip debug features and the ability to execute the full ARM instruction set or the subset used in THUMB2 processors. The Cortex-M4 processor’s instruction set is enhanced by a rich library of efficient DSP features including extended single-cycle 16/32-bit multiply-accumulate (MAC), dual 16-bit

MAC instructions, optimized 8/16-bit SIMD arithmetic and saturating arithmetic instructions. Overall, the most noticeable difference between M3 and M4 is the optional single-precision (IEEE-754) Floating Point Unit (FPU) available with the M4.

Serial Secrets Stimulate Slick Solutions

The success or failure of an embedded design often rests on finding the right balance between system performance, energy consumption and solution cost. In many cases, developers can use the Cortex-M processor’s unique features to optimize for product cost or energy appetite while maintaining, or even improving, its performance. For example, the Cortex-M core has native serial I/O capabilities that can be used to save energy, simplify development and free up peripherals to be used for other application tasks. Besides the traditional Serial Wire Debug functions, ARM Cortex-M-based microcontrollers also offer an instrumentation trace interface through their singlepin Serial Wire Viewer Output (SWO), as shown in Figure 2. This port can be used to pass “printf-format” debug messages

directly to application code. SWO allows the debug messages to be viewed directly from any standard IDE. Additionally, these messages can be viewed through a standalone SWO viewer such as Segger’s J-Link SWO Viewer software or the energyAware Commander from Silicon Labs. Since the SWO output is built into the core hardware itself, this is an inherent benefit of the Cortex-M core. SWO doesn’t waste any of the MCU’s regular UARTs, which might already be committed to the application. Another important advantage of SWO-based debugging is that it allows the MCU to maintain an active debug connection when it enters its lowest sleep modes where, in most cases, the logic for traditional debug connections is inoperative. The instrumentation trace of the SWO can also be used for sampling the program counter to help IDEs create statistics on how much time is spent in each of the program functions. These statistics can be combined with current measurements to help fine-tune a design’s energy consumption. Cortex-M-based MCU vendors are beginning to recognize this benefit, and some manufacturers have already incorporated power profile and current measurement hardware into their development platforms for this purpose. For example, all starter and development kits for the EFM32 Gecko MCUs from Silicon Labs include live power measurement outputs, which can be coupled with the program trace in the energyAware Profiler tool. Figure 3 shows how this allows the designer to pinpoint which program functions are the highest energy drains and allows fast debugging of other energyrelated problems.

Sleep Smart and Make Every µW Count

The ARM Cortex-M processor’s Sleep-on-Exit instruction is another “twofer” feature that can save both CPU cycles and energy. This is especially useful in interrupt-driven applications where the processor spends most of its time eiRTC MAGAZINE OCTOBER 2013



ther running interrupt handlers or sleeping between interrupt events. When entering an interrupt service routine (ISR), the MCU must spend several instruction cycles pushing the present thread’s state onto the stack and then “popping” it upon return. In applications where the processor returns directly to its sleep after an ISR, a conventional MCU must still recover its stored state information before the thread code can put the device to sleep. Likewise, its state must be pushed to the stack again when the next interrupt wakes the device. When an ARM Cortex-M-based microcontroller’s Sleep-on-Exit is enabled, the device will enter sleep directly after the ISR finishes without returning to the thread (Figure 4). This preserves the processor in the interrupt state, saving the precious machine cycles normally required to push the normal state onto the stack during wake-up. Eliminating the stack push and pop cycles saves both the

time and energy otherwise consumed by unneeded instruction cycles, as well as any code a conventional MCU would need to manage the stack between its sleep and wake states. And, should the processor be awakened by a halt debug request, the unstacking process will be carried out automatically.

Run Faster, Sleep Deeper with the ARM Cortex-M4

Like many MCUs, Cortex-M3/M4 processors can often achieve energy savings in interrupt-driven applications by running at a relatively high clock rate. This counterintuitive but commonly used energy-saving tactic works well if the processor spends much of its time in a sleep mode where the savings from its reduced active time far outweighs its slightly higher operating current. Put simply, expending 10 percent more power for

FIGURE 3 Software and hardware tools that pinpoint which functions are using the most current eliminate the need for oscilloscopes and multimeters and enable fast debugging.



20 percent less time represents an overall energy savings. This technique can be applied to any Cortex-M series processor, and applications that involve compute-intensive tasks can also benefit from the Cortex-M4 processor’s added capabilities. Its singlecycle DSP instructions and optional floating point accelerator can greatly reduce the number of execution cycles required for functions such as digital signal conditioning, filtering, analysis or waveform synthesis. Some applications simply need the processing horsepower of a DSP. For example, some security systems employ a device that senses glass breakages using acoustic analysis. Breaking glass is accompanied by a distinctive series of sounds and vibrations that culminate in a resonance at the characteristic natural frequency of the glass, in this case around 13 kHz. Most systems employ a sensor interface that only wakes up the processor when telltale frequencies are detected. However, designs using a Cortex-M4 DSP-enabled CPU achieve additional energy savings by performing the actual glass break analysis more quickly than software-based solutions. Even greater energy savings can be realized in these applications using M4based MCUs that include advanced sleep modes and autonomous peripherals that perform many routine tasks while the CPU remains asleep. For instance, the CortexM4 equipped Wonder Gecko MCU has five distinct low-energy modes including a 20 nA shut-off state and a 950 nA deep sleep mode (running real-time clock, with full RAM and register contents retained and brown-out detector enabled). The same features that enable energy savings can also yield other advantages. For example, applications such as ultrasonic/acoustic water meters, which must operate for years on a small battery, require the MCU to remain in sleep mode as long as possible. In addition to helping to reduce the MCU’s wake time, the CortexM4 DSP and floating-point math instructions also eliminate the need for expensive ultrasonic flow transducers by using sophisticated filtering functions to extract the necessary information from the output of inexpensive acoustic sensors. In this


application example, the Wonder Gecko MCU’s peripherals provide additional energy savings by acting as an analog state machine that wakes the Cortex-M4 processor only when needed. In addition, the EFM32 Gecko and Wonder Gecko MCU families from Silicon Labs provide examples of how the choice of an ARM-based MCU with


the right combination of I/O, accelerators and other advanced peripherals can improve a design’s performance, energy consumption and solution cost.

MDK-ARM™ is the complete software development environment for ARM® Cortex™-M series microcontrollers.

Silicon Labs Austin, TX. (512) 416-8500. [].





ARM DS-5 DS-5™ Professional is a full featured software development solution for all ARM Powered® platforms.


Continue to next instruction




Reset Value






Send Event on Pending; wakes up WFE if a new interrupt is pending, regardless of whether the interrupt has priority higher than the current level.







Enable SLEEPDEEP output signal when entering sleep mode.





Enable SLEEPONEXIT feature.



FIGURE 4 The ARM Cortex-M Sleep-on-Exit capability reduces power consumption by avoiding unnecessary program execution and by reducing unnecessary stack push and pop operations. Courtesy of “The Definitive Guide to the ARM Cortex-M3.” © ARM Ltd. AD381 | 08.13

Untitled-4 1



9/3/13 9:29 AM



High Level, High Performance

VSIPL: A Signal and Image Processing Standard Family For high-performance, portable, highly productive embedded computing applications, the standards provide a path to implement powerful functions at levels of abstraction that can migrate to newer processors, operating systems and vendors by Anthony Skjellum, PhD, Run Time Computing Solutions


he Vector and Image Signal Processing Standards (VSIPL) is a coherent set of de facto standard application programmer interfaces for signal and image processing emphasizing floating point computation, which was recently adopted by the Object Management Group (OMG). It is a mature specification family with three sub-standards and a set of subsets or profiles. The most widely used of these is C-API, simply called VSIPL, which includes signal processing and linear algebra functionality. It was developed in the late 1990s and was updated over the past 15 years. It has broad and deep support in the embedded High Performance Computing (eHPC) community and is used in deployed defense programs as well as in research laboratories, and to a lesser extent in academic research related to mil/ aero applications. Currently the OMG High Performance Embedded Group (OMG HPEC) houses, maintains and enhances these standards. High-performance, high-productivity middleware implementations of VSIPL enable significant mil/aero applications worldwide. Predictability and performance with high portability are key design aspects of the VSIPL standards. The arrival of more and different proces-



sors and attached processors ranging from vector units to GPUs and embedded signal processing units increases the value and relevance of VSIPL, and an eco-system based on it, rather than diminishing it. Here, we explain the origins, technologies, profiles and example uses of VSIPL, and its continuing and growing relevance in embedded and real-time high-performance computing. The history of VSIPL is based on sound, effective industrial practice and fundamental algorithms in signal processing, image processing and linear algebra, with an emphasis on floating point and performance portability. The efforts resulted in an initial specification dating back to 1997. The goal of VSIPL is to make significant tradeoffs in the performance-portability-productivity (PPP) space, which is illustrated in Figure 1. Traditionally, engineers use the maxim of “choose 2” relating to “good,” “fast” and “cheap.” VSIPL and its related standards support strong portability, good performance and reasonable productivity of the programmer. Depending on how they are used, these standards allow for developer tradeoff in the PPP space and provide a maintainable basis for developing platform-specific code of ultimate

performance that can continue to evolve if economically justified.

Origins and Initial Standardization

The VSIPL standard was motivated and funded by the Defense Advanced Research Projects Agency (DARPA), and further advanced by the U.S. Navy during the latter half of the 1990s.The original concept was to abstract, standardize and support both performance and portability across a variety of computer architectures, drawing on the inspiration of the widely copied Floating Point Systems (FPS) signal-processing library. By that time, FPS was a historical reference point, but several companies had similar libraries in the high-performance (floating point oriented) signal-processing space, and this approach was found meritorious. Major features and advances of the formal specification that were developed and that became VSIPL 1.0 include: • Detailed application programmer interface rules for multiple precisions, based on usability from the C 1989 standard language. • The incorporation of an objectbased set of abstractions, rather than application programmer interfaces


with “fat argument lists” or “fat interfaces.” These are common to FPS, vendor libraries and mathematical library software such as Basic Linear Algebra Subprograms (BLAS) and LINPACK of the same era. Through the incorporation of abstract data types, options within the library for storage and performance trade-offs were enhanced. • A memory-model abstraction involving memory blocks and views, which enables both library-allocated, and user-allocated memory, with flexible memory layouts, and clear state transition between user and library control of such memory. • The incorporation of earlybinding semantics for “heavy” operations such as Fast Fourier Transforms (FFTs), enabling “planning” of the operations as a function of the parameters such as size, precision and in-memory storage. • Compliance points—profiles— supporting commonly understood subsets to allow early adoption and support of highly constrained systems, notably the CoreLite profile (127 functions) most closely mimicking industrial practice at the time, and the Core Profile, which included substantial linear algebra. The API properties and abstractions of the VSIPL standard coincided in a timely way with the introduction of the AltiVec family of vector-enabled processors from the Motorola 7400 PowerPC (now Freescale). VSIPL was able to hide the complexity of AltiVec within its APIs, while still delivering high performance in a number of implementations of the standard. The multi-year dominance of AltiVec-based processors in embedded High Performance Computing helped ensure wide adoption of VSIPL and its use in a number of defense and mil/aero programs in the United States, Europe and Asia/Pacific regions. VSIPL again proved its utility a decade later when Freescale, the successor to Motorola in AltiVec-based processors, temporarily stopped offering a future roadmap for AltiVec-enabled processors. It proved easy for programs and algo-

Portability (Good)

Performance/Scalability (Fast)

Productivity (Cheap)

FIGURE 1 Project Management Triangle applied to VSIPL.

rithms written to VSIPL and targeted to AltiVec implementations of the library to retarget to other processors. This took the form of optimized implementations of VSIPL for x86 and x86-64 SSE/AVX. More issues about cross-platform portability are discussed later in this article.

Extensions to the Original Standard, C++ and Image Processing

The VSIPL specification underwent reasonably small transitions over the period of 1998-2012, principally corrections and incremental additions, including ancillary functions such as interpolation, sorting and clarifications of the existing functionality. The DARPA and U.S. Navy funding eventually came to an end, and the High Performance Embedded Working Group Software Initiative supported by them became an entirely volunteer group over time. In addition, an image-processing library was proposed in the late 1990s, but it was never formally standardized for reasons that remain unclear after over a decade. One full implementation of this image functionality is known to exist within the commercial VSI/Pro library and it is used in certain defense applications. Standardization of the image functionality remains an option for the new OMG standards process. A strong effort for message passing and parallelism abstraction developed during the late 1990s and early 2000s

with the addition of a C++ standard. Originally prototyped and later commercialized by CodeSourcery (later Mentor Graphics Sourcery Tools), a number of developers chose VSIPL++ and parallel VSIPL++ over the C-API VSIPL at the time. Additionally, it became an important vehicle for higher-level middleware used in parts of the community such as the Parallel Vector Tile Optimizing Library (PVTOL), a higher-level library for signal processing developed at MIT Lincoln Laboratory. However, unlike VSIPL, which has many vendor implementations and a good open-source implementation, VSIPL++ standards have primarily only the CodeSourcery implementation, which as of 2013 is no longer commercially available.

The New OMG Process

During 2011-2012, the ad hoc High Performance Embedded Working Group, which was an entirely voluntary group with origins in the U.S. Defense industry, government agencies and VSIPL implementers, decided to align itself in part with the Object Management Group. Their intention was to seek adoption of VSIPL 1.3, VSIPL++ 1.1 and VSIPL++ Standard 1.1 with a recognized standards body. The key goals were to support longterm viability, bring new participation into the standards, and provide a deeper imprimatur of a public standards body to these well-developed specifications. A number of companies, a handful of universities and many of the original standardizers championed the move to OMG “ownership” of the standards. These included RunTime Computing Solutions, Georgia Institute of Technology, Mentor Graphics, MIT Lincoln Laboratory, GE Intelligent Systems, Mercury Computer Systems, CSP, Lockheed Martin Northrop, and other organizations and individuals with various ties to VSIPL and/or VSIPL++. The effort culminated in 2012 with adoption by OMG. At that time, with minor exceptions, VSIPL 1.3 became OMG VSIPL version 1.4, and likewise, the two C++-based standards were incremented to indicate initial OMG support.




Range of Algorithmic Coverage

The range of coverage of the VSIPL algorithms covers both the C and C++ versions of the standard. They come in the following categories: • Vector operations (element by element, dot product, extensions of BLAS level 1) • Matrix-vector operations (such as matrix-vector multiplication, analogous to BLAS level-2) • Matrix operations (solvers such as LU, QR, SVD, analogous to LAPACK) • Fast Fourier Transforms, Windowing Options • Convolution and Correlation Operations • Finite Impulse Response and Infinite Impulse Response Filters • Sorting, Interpolation, Copy, Inter-precision Copy and other data management







Untitled-3 1

Safe computers for rail, road and air, up to SIL 4/DAL-A Modular box and panel PCs for industry & transportation  Powerful system solutions on CompactPCI®/PlusIO/Serial  Rugged, standard Computer-On-Modules (ESMexpress®, ESMini™)  EN 50155- and e1-certified Ethernet switches and fieldbus interfaces

OCTOBER 2013 RTC MAGAZINE 9/26/13 9:25 AM

All of these operations work with abstract data typess, memory management that abstracts storage format (when desired) and allows the internal implementation to make tradeoffs of space, speed and accuracy within commonly accepted bounds. Importantly, the operations are also specified for multiple data types, including optimizable precisions such as integers of varying widths, float and double, as well as precisions that are not widely used now, and might be utilized in the future, such as long double. Notably, many VSIPL programs utilize only float and integer precisions, based on legacy architecture support and the design of legacy applications based on them. The VSIPL and VSIPL++ standards are flexible standards with a bright future. They support existing and earlier architectures well. As part of OMG they are poised to expand in ways that ensure their continued relevance. The VSIPL (C-API) standards have brought vendor and user acceptance and multiple implementations, while the C++ API standards have strong community interest, though they lack a currently supported commercial implementation. Because of the openness of the Object Management Group process, the move of the VSIPL standard family to OMG in 2012 means that all interested parties can join and

help advance and influence the standards now and in the foreseeable future. The trade-space of performanceportability-productivity illustrated in Figure 1 remains important to designers and implementers of large-scale signal and image processing. Its importance in the current era affects both economic and technical dimensions of software engineering that drive practical systems of high performance in which the software is maintainable and can be retargeted to newer systems in the future. The major lessons of the VSIPL application programmer interfaces is that a relatively small group of people can produce a major impact on the quality of a large class of computer applications in the embedded High Performance Computing space and their ability to evolve. VSIPL has certainly driven standardization and has advanced COTS computing, because programming to VSIPL enables developers to retarget to different processors, operating systems, vendors and inter-generational system changes as well. VSIPL met its original goals and continues to enhance the quality and lower the cost of important applications in embedded computing with mil/aero and defense applications in particular. RunTime Computing Solutions Birmingham, AL. (205) 314-3595. [].

Why Should Researching SBCs Be More Difficult Than Car Shopping? INTELLIGENTSYSTEMSSOURCE.COM IS A PURCHASING TOOL FOR DESIGN ENGINEERS LOOKING FOR CUSTOM AND OFF-THE- SHELF SBCs AND SYSTEM MODULES. Today’s systems combine an array of very complex elements from multiple manufactures. To assist in these complex architectures, ISS has built a simple tool that will source products from an array of companies for a side by side comparison and provide purchase support. PLATINUM SPONSOR







High Level, High Performance

LabView or C? The next time someone asks you whether LabView trumps C, feel free to answer “42.” It may be the only response that will get the discussion heading in the right direction. by Simon Hogg, National Instruments


hy is LabView better than C? A LabView Product Manager gets asked this question a lot. Honestly, it is the wrong question to ask. It becomes a valid question with a little nuance and application context, for example, “Which is better for this task, under these constraints?” Without this detail, it’s like asking why bread is better than flour. If you want to build a measurement or control system, then NI LabView system design software is a tool that can save you the risk, expense and inconvenience of building your own from lowlevel languages like C. I’m not suggesting that LabView is a “better” programming language than C—especially considering that large portions of LabView are written not only in G but also in C and C++. Rather, they have different strengths that programmers should understand to succeed. The relationship between LabView and C is similar to bread and flour. If you want to make a sandwich, start with bread. If you want to bake a cake, start with flour. Baking bread with flour from scratch can be expensive and time-consuming—especially if you just want a quick snack. Using pre-baked bread gives you a foundation to build your custom sandwich on. When it comes to baking a cake, flour is essential and bread is not even an option. Similarly, you might find it challenging to



FIGURE 1 C is ubiquitous, high performance and helps maximize the system creator’s choice.

decide which programming language is best for your task. It comes down to using the right tool for the right job.

C Gives You Low-Level Control

C is often better for applications with tight resources that must be closely managed. Since C is a relatively low-level language in this age of object-oriented, memory-managed languages, you need to consider and specify even the smallest

details, such as memory assignments and threads. A good programmer uses this low level of control to eliminate the overhead present in most high-level implementations. By targeting and compiling for the specific machine you intend to run on, instead of a common runtime or virtual machine, it is possible to also take advantage of target architecture or host operating system properties to achieve greater performance.


NI programmers wrote most of the LabView libraries in C or C++ for this reason. For operations like file I/O and analysis to be as fast in LabView as they are in C, they are written in low-level languages and optimized for each of the platforms and operating systems that LabView supports. While C is undoubtedly low level by today’s standards, it still remains one of the most portable languages available. Well-written and standards-compliant C source code can be recompiled for a wide range of hardware—from the smallest of embedded targets to giant supercomputers—with minimal changes. This frees the system creator to choose the most appropriate hardware from the spectrum of available processor architectures, operating systems, price points and form factors. The ubiquity of C (Figure 1) also means that almost any peripheral hardware device will have a C driver or interface. This again helps provide the system creator with increased flexibility when putting together a system.

Efficiency Versus Control

At some point, developer efficiency trumps the need for hand-optimized code and complete choice with regard to every detail. Relinquishing a little control to stand on the shoulders of those who have solved similar problems can benefit many projects in terms of quality and productivity. Programming languages are constantly progressing toward higher levels of abstraction. This helps you focus on the problem at hand instead of the minutia of the computing. For some, higher levels of abstraction make do-it-yourself programming feasible. As an example, domain experts, scientists and doctors often have the insight and intelligence to make game changing advances in their fields but lack the programming experience to create reliable implementations. High-level tools such as LabView make it possible for these experts to contribute key functionality rather than going through a series of specification-implementation-validation cycles with a programmer. For others, higher levels of abstraction allow for greater reuse of well-tested and supported functionality that is baked into the abstraction layers. Just because

FIGURE 2 LabView allows you to skip building a foundation and go straight to customization.

you can write a memory manager or networking stack doesn’t necessarily make it a good idea. Off-the-shelf foundational pieces are typically more affordable in the long run and offer better documentation, debugging tools and compatibility with third parties than custom implemented designs. High-level code by definition hides some of the implementation details. This isn’t always a good thing when you are trying to wring out every clock cycle of performance. The further you are from raw CPU instructions, the more likely you will encounter unnecessary CPU and memory usage due to imperfect optimizations in generic abstraction layers.

Parallel Execution and Real-World I/O

No matter what the implementation language, high-level system design and low-level implementation must inevitably split. In measurement and control applications, programming is just one task of a system designer. Engineers often don’t have time to keep up with or rewrite old software to support the advancements in computing and measurement hardware, operating systems and so on. They add value by figuring out how to acquire, ma-

nipulate and present real-world data—not by coming up with new ways to handle memory allocations and thread pools. LabView users can build on top of tested, supported and maintained libraries of lower level code from NI. C users have the option of implementing, supporting and maintaining their own lower level libraries, purchasing them à la carte from various vendors or relying on open-source solutions. Syntax-wise, C is a general-purpose language that excels at sequential execution of instructions as fast as the CPU can handle them. This is perfect for pure computation, when only one task is being executed at a time, or on microcontrollers with only one execution thread. The graphical syntax in LabView, on the other hand, is optimized for the parallel execution of tasks that have real-world timing constraints and run on multicore and heterogeneous systems. LabView is more than just a programming language and associated libraries. The LabView integrated development environment (IDE) is best when used with NI hardware to provide an integrated development experience. The software is aware of available hardware resources and can present available I/O




channels and execution targets as dropdown menus and project items. In addition, the compiler can catch many incorrect hardware or driver configurations during editing, avoiding costly and hardto-debug runtime errors. Furthermore, LabView can also be used to program the FPGAs on certain NI hardware, opening up the high-performance, high-reliability world of custom hardware to people who would otherwise be dissuaded by the need to learn additional languages and tools (Figure 2). Many projects end up late or over budget because people underestimate the effort required to stitch together parts from disparate sources. Using a platform-based approach means hardware drivers return data in the same format as the analysis libraries consume, and UI widgets display technical data in the same format that the analysis libraries produce—eliminating the need to piece together components. Ultimately, this approach makes it possible to gain a significant boost in developer productivity as long as everything you need is available within the NI platform. The downside of this tight hardwaresoftware integration is that you give up the ability to choose any hardware or

Untitled-2 1



OS combination. LabView can target full desktop operating systems (e.g. Windows, Mac OSX or certain Linux distributions) or NI embedded hardware running LabView Real-Time. However, it also requires a larger runtime environment than C, which makes it impractical for low-end embedded hardware. Predictably, peripheral hardware from other vendors isn’t as well integrated in the LabView environment as NI’s own hardware although any hardware that works with C can also be used in LabView by calling into the existing C libraries. This brings up another important consideration—interoperability. In many situations the best solution may be to use both tools—for example, LabView to architect the system and C to implement the optimized critical paths. The answer to the question, “Which is better: LabView or C?” might as well be “42.” To draw from The Hitchhiker’s Guide to the Galaxy, the answer isn’t meaningful until you know which question you are asking or what problem you are trying to solve. LabView and C are both useful tools that, in the hands of skilled users, can solve almost any problem. LabView tends to be better for rapidly developing high-level test, measure-

ment and control applications, and C is more apt for lower level implementations where a myriad of choices is beneficial to the system creator. National Instruments Austin, TX. (512) 794-0100. []

9/26/13 9:22 AM

Embedded Engineers Invited Its here...The perfect technological storm to enhance embedded functionality RTECC shows you how intelligent and embedded systems are changing our world and the technology we deliver.

Register today at Seattle, WA Tuesday, November 5, 2013 Double Tree by Hilton Seattle Airport 18740 International Blvd. Seattle, WA 98188 8:30 AM – 2:30 PM

Portland, OR Thursday, November 7, 2013 Crowne Plaza Lake Oswego 14811 Kruse Oaks Drive Lake Oswego, OR 97035 8:30 AM – 2:30 PM

Platinum Sponsor Don’t miss special sessions and hands on labs presented by AMD and their partners in both locations. SIlver Sponsors


TECHNOLOGY Mini-ITX SBC with Intel Fourth Gen Processors to Serve Industrial Control & Other Apps

A high-performance, scalable Mini-ITX embedded board is powered by a selection of Intel Celeron and 4th Generation Core i7/5/3 BGA processors. The MB-73320 from Win Enterprises features multiple digital video outputs, including HDMI, DVI-D, 24-bit LVDS, as well as VGA output to support an array of display options. In addition to features often found on embedded Mini-ITX products, such as PCIe, mPCIe, LAN, etc., MB-73320 provides increased integration flexibility with up to 6 COM Ports (i.e., 1x RS-232-RS-485, 5x RS-232) and integrated Intel RAID firmware with support for RAID 0, 1, 5, 10; plus support for mSATA storage. Features include a rich assortment of I/O with HDMI/DVI/ VGA and 24-bit LVDS. There are two Intel GbE LAN ports plus six COM, 10 USB ports, LPC and SMBus. Serial ports include one RS-232/422/485 and five RS-232. There are four SATA interfaces with RAID plus HD Audio. In addition, the module offers PCIe X16 & X1 slots and two Mini-PCIe sockets. The board’s single-voltage variable input of 8-32 VDC enables installation opportunities across a variety of environments. The MB-73320 serves as the basis for small-footprint, graphics-intensive, solution-level products in industrial automation, communications, military, medical and digital security surveillance applications. WIN Enterprises, North Andover, MA. (978) 688-2000. []

Programmable USB Port Power Controllers for Active Connectors and 12W Charging

A new family of power controllers offers advanced USB-based charging capabilities for designing host devices, such as laptops, tablets, monitors, docking stations and printers, as well as dedicated AC-DC power-supply and charging products such as wall adapters. The UCS100X family from Microchip Technology—the UCS1001-3, UCS1001-4 and UCS1002-2—is an expansion of its UCS1001 and UCS1002 series. These new controllers offer higher current and priority charging for smartphones and tablets. The UCS100X have also added support for active cables, such as the Apple Lightning connector, along with 12W charging. The UCS1002-2 features a built-in current sensor that can report on the amount of charging current. This allows a system to optimize its charging current and appropriately allocate power. Additionally, the UCS100X can support future USB product designs via a flexible method for detecting and creating charging emulation profiles. This allows designers to update their systems as new products are introduced to the market, while providing compatibility with a wider range of existing products. The UCS100X family is supported by Microchip’s new UCS1001-3/4 Evaluation Board (Part # ADM00540, $24.99) and UCS1002-2 Evaluation Board (Part # ADM00497, $90.00), both of which are available today from any Microchip sales representative or authorized worldwide distributor. The UCS1001-3, UCS1001-4 and UCS1002-2 are available now for sampling in a 20-pin QFN package. Pricing starts at $0.90 each, in 5,000-unit quantities. Microchip Technology, Chandler, AZ. (480) 792-7200. []



PC/104 SBC Using DMP Vortex86DX2 Single-Chip Solution Reduces Power/Space/Cost

A new PC/104 form factor single board computer (SBC) features an extensive I/O feature set and full ISA bus support. The CoreModule1-86DX2 from Adlink Technology is aimed at control applications that require power efficiency, small form factor design, longevity and industrial grade ruggedness. The CoreModule1-86DX2 enables a performance boost for legacy PC/104 systems that continue to utilize real ISA bus and supports applications across markets, including transportation, automated manufacturing and aviation. Based on DMP’s Vortex86-DX2 single chip solution, which integrates a powerful yet efficient CPU with graphics controller, audio controller and other functionalities, the CoreModule1-86DX2 provides all of the standard peripheral connections of an embedded PC on a printed circuit board with dimensions of 96 mm x 90 mm (3.775” x 3.550”). The CoreModule1-86DX2’s rich I/O includes 2 Ethernet ports, 4 serial ports, 3 USB ports, 8 A/D inputs, 8 GPIOs and PS/2 connectors for keyboard and mouse. A SATA 1.5 Gbit/s interface is also provided to allow for connection with a disk or optical drive. The CoreModule1-86DX2 is also able to withstand 11.95 Grms vibration and 40G shock (MILSTD-202G Method 214A/213B) and operate over an extreme temperature range of -40° to +85°C. The ADLINK CoreModule1-86DX2 also supports Smart Embedded Management Agent (SEMA) functions, including watchdog timer, temperature monitoring and fail-safe BIOS support, further ensuring the system’s operational reliability. The CoreModule1-86DX2 runs DOS, Windows XP/7 and Linux operating systems and is powered by a 5V-only supply. System expansion can easily be realized over the PC/104, Mini-PCI Express and I²C bus connectors. Adlink Technology, San Jose, CA. (408) 360-0222. []


PMC Module Carrier with PCI Express Bus Interface for PC-Based Embedded Systems

A PMC carrier card provides an easy and efficient solution to interface a PMC mezzanine module to a PC across the PCI Express (PCIe) bus. Engineers can plug Acromag’s FPGA modules or other PMC modules onto its new APCe8670 carrier card to perform a variety of signal processing functions. A bridge chip handles the PCI-X to PCIe conversion between the plug-in PMC module and the host computer. The carrier’s PCIe x4 interface supports up to four serial lanes for rapid data transfer. These carriers are suitable for high-performance industrial and scientific research computing systems. They are also useful to test advanced systems for defense and aerospace applications that will be deployed later on rugged embedded computer platforms such as VME, VPX, or CompactPCI. Three PMC PCI-X bus sockets interface through a bridge to the carrier’s PCIe x4 edge connector. The other PMC socket routes 32 LVDS I/O to a rear SCSI port for board-to-board communication. An integrated fan provides a constant airflow across the PMC module since most PCs cannot provide sufficient cooling to dissipate heat generated by large FPGA devices. The carrier card is also ready for conduction cooling with a thermal frame. Software development tools are available for VxWorks, Linux and Windows environments. These function libraries provide example routines or DLL driver support with C source code to save time. Pricing starts at $750. Acromag, Wixom, MI. 248-624-1541. []

LVDS Digital I/O Module for Control and Data Capture Applications A digital I/O module connects low-voltage differential signaling (LVDS) I/O to a Virtex-6 FPGA. This Cobalt module provides 32 LVDS differential input or output pairs plus LVDS clock, data valid and flow control signals via an easy access front panel 80-pin connector routed to the Virtex-6 FPGA. LVDS used in the Model 71610 from Pentek is a general purpose digital interface operating at very high speeds over inexpensive twisted-pair or flat ribbon copper cables. It is popular for many common control or data capture applications such as high-speed video, graphics, video camera data transfers and general purpose computer buses. The standard built-in data capture and data generation facilities of the Model 71610 make it a suitable turnkey solution for developing and deploying custom FPGA processing algorithms. Data can be pulled from devices such as sensors, data converters, custom digital systems or spectrum analyzers and then processed, decoded or formatted in the Virtex-6 FPGA before delivery to the host system processor. Likewise, signals from the host can be processed and formatted by

the FPGA before delivery to the LVDS output. The Model 71610 is a VITA 42 XMC (switched mezzanine card) module that can plug into carriers such as VPX, CompactPCI, AMC (Advanced Mezzanine Card) and PCI Express. The Model 71610 supports x4 PCIe on the primary P15 XMC connector. The secondary P16 XMC connector supports dual 4X or single 8X user-installed gigabit serial interfaces, such as Aurora, PCIe or Serial RapidIO. The Model 71610 has an optional PMC P14 connector with 20 pairs of LVDS connections to the FPGA for custom I/O to the carrier. Pentek GateFlow FPGA Design Kits allow users to develop and install their own custom IP for data processing. The kits include full VHDL source code for factory installed intellectual property (IP) functions, so developers can easily add new functions by incorporating additional FPGA IP modules. Pentek’s ReadyFlow board support packages, with high-level C-callable library functions and device drivers, are available for Windows and Linux operating systems. Pricing starts at $9,495 and the Model 71610 is available in XMC, VPX, CompactPCI, AMC and PCI Express and rugged form factors. Pentek, Upper Saddle River, NJ. 201-818-5900, []

Quad-Port and Dual-Port Network Adapters with Cost-Effective Bypass Design

American Portwell’s BPC-51242 is an Intel I350-AM4-based quad-port copper Gigabit Ethernet adapter with superior bypass functionality provided by Portwell’s Gen 1.6 bypass interface. The BPC-51120 is an Intel I210-based dual-port copper Gigabit Ethernet low profile adapter with Portwell’s Gen 3 bypass interface that benefits from being software-controlled. The BPC-51242 is suitable for high-end server appliances and inline network systems that need to maintain connectivity with fail-over/bypass functionality. BPC-51120 is targeted for the enterprise data center, high-performance computing and the embedded environment. The BPC-51242 supports a quad-port copper Gigabit Ethernet interface via Intel’s I350AM4 MAC+PHY Ethernet Controller; Intel SR-IOV functionality; built-in Watchdog Timer (WDT) to bypass Ethernet ports when host system hangs or suffers power failure; PCIe x4 (and ready for PCIe v2.0 5GT/s solution). The BPC-51120 leverages Intel’s I210 (formerly Springville) Ethernet controller. It is fully compliant with 1000base-T and provides Audio Video Bridging (AVB) support with power management technologies; EnergyEfficient-Ethernet (EEE) and direct memory access (DMA); IEEE 1588 precision time protocol circuitry synchronization. BPC-51120 offers Generation 2.1 (2.5GT/s) PCIe x4 or PCIe x1 through its PCI x4 golden finger and presents a compact-168.15 mm x 68.9 mm-low profile, half-length footprint. Both new BPC-51242 and BPC-51120 network adapters provide a completed intrusion prevention solution provided by Portwell’s bypass design that supports three different bypass modes—normal, bypass and open—to diffuse the cost and inconvenience of a system crash or power failure. American Portwell, Fremont, CA. (877) 278-8899. []




Ultra-Slim 8.4-Inch Fanless Touch Panel Computer Adopts Atom Processor N2600

A new 8.4-inch fanless touch panel computer with super slim and light design is targeted as a self-service kiosk in shopping center, supermarket and factory & building automation fields. The GOT5840T-832 from Axiomtek supports an energy-efficient Intel Atom processor N2600 1.60 GHz with the Intel NM10 Express chipset. This rugged platform adopts an 8.4” SVGA TFT LCD with resolution up to 800 x 600 pixels, and features fanless operation and a rugged IP65 dust/waterproof front bezel. The durable panel system also provides a PCI Express Mini Card slot and a built-in WLAN antenna for wireless network connections. This new addition to the fanless touch panel computer series is being reformed to be a highly integrated and easy-to-use platform with a thickness of 45 mm and a weight of 1.3 kg. The industrial-grade front bezel is NEMA 4 / IP65 compliant with the advantages of light weight, high degree of hardness, and anti-corrosion ability to be designed to work in wide range heavy-duty industrial fields. The GOT5840T-832 also offers excellent vibration resistance up to 2 Grms (with CompactFlash) for stable operations. The new 8.4-inch fanless touch panel computer offers a PCI Express Mini Card slot and a fixed rotational WLAN antenna (optional) for wireless network connection in real time. By just plugging in the WLAN Mini Card, users can have instant access to wireless LAN/GPRS/GSM/3G. It provides users with rich I/O interfaces including two COM ports (RS-232 & RS-232/422/485), one Gigabit Ethernet, two USB 2.0 ports and an audio port (Line-out). It comes with one DDR3 SO-DIMM slot with up to 2 Gbytes of system memory. This rugged unit supports two types of power input: 10 ~ 30V DC power input with terminal block and 60W AC power adaptor with screw-type connector. For flexible installation, the GOT5840T-832 can be panel mounted, wall mounted, VESA mounted, or desk mounted. It is compatible with Windows 7 Embedded Standard and Windows 7 (32-bit), which allows developers to quickly and easily bring devices to market.

Full HD Autofocus 8MP USB 3.0 Camera Streams Uncompressed 1080p

MicroTCA.4 Chassis Offers Rear I/O Options

An 8U MicroTCA.4 chassis is designed for High-Energy Physics and other applications that require rear I/O. The VT811 chassis from VadaTech is targeted to specifically address customers’ concerns with other MicroTCA.4 chassis in the marketplace, including fan tray alignment, chassis weight, cable management and more. The VT811 features a lightweight aluminum construction with integrated cable ducts below the card cage. This allows the cables to be easily protected and routed to the rear of the chassis. The fan trays utilize Teflon strips, which make insertion/extraction smoother and easier. Plus, the trays have shrouded blind-mate connectors for both the male and female ends, which prevent damage and ease guided insertion. The VT811 has a 30-layer impedance controlled backplane with slots for 12 double-wide AMC modules in the mid-size, plus 2 MCHs, and 4 PMs. The clock traces are laid out to give equal track length from MCH to each AMC slot, easing latency equalization. The rear of the enclosure allows for 12 double-wide Rear Transition Modules (RTMs) to be inserted. The fan trays are arranged with sixteen 2-inch fans each above and below the card cage in a push-pull configuration. The smaller, powerful fans ensure each slot gets optimal airflow, avoiding hot spots. Other features of the VT811 chassis include radial I2C bus to each AMC, an integrated and pluggable JTAG Switch Module (JSM) and Telco Alarm. Ports 2-3, 12-15 and 17-20 are connected among the slots per the MicroTCA.4 recommendation.

A new camera solution supports video streaming at resolutions of up to 1080p@30fps and includes an eight megapixel autofocus camera module, the e-CAM80_MI8825_ MOD, with an OV8825 CMOS image sensor. The UVC-compliant See3CAM80 from e-con Systems is plug-and-play in both Windows and Linux. In Windows, the camera is exposed as a DirectShow device and in Linux, as a V4L2 capture source. The See3CAM_80 supports VGA at 30, 720p30, 1080p30 preview resolutions for high-end video recording, video analytics and HD video conferencing applications. e-con is also bringing the full 8 megapixel still image capture capability to regular PCs for very high resolution imaging applications. Above all, the See3CAM_80 can stream the full 8MP resolution video at around 11fps for applications where full image resolution required, but at a lower frame rate. All of the above are uncompressed video and streaming is through the USB3.0 interface, which supports about 5 Gbit/s bandwidth. The See3CAM_80 is suitable for customer applications such as document cameras/ visualizers, high-resolution OCR applications for large sheet sizes, medical and scientific applications such as tissue vision, pathology examination and microscopic applications where high resolution streaming—even at a low frame rate—is required. It makes full use of the potential of its OmniVision’s OV8825 CameraChip sensor. The OV8825 is a 1/3.2” optical format, 1.4 micron pixel raw image sensor with a four-lane MIPI CSI-2 interface and supports 720p60, 1080p60 and also 8MP at 24fps. Third-party Linux and Windows applications can use the standard UVC protocol to access the e-CAM80_MI8825_MOD camera module’s built-in ISP. The ISP supports autoexposure (AEC), AGC, auto-white balance (AWB) and image property controls such as brightness, saturation, AWB effects, sharpness and contrast. Special features, including manual focus, single-trigger and autofocus are available through the UVC extension interface, which is also provided by e-con.

VadaTech, Henderson, NV. 702-896-3337 []

e-con Systems, St. Louis, MO. (636) 898-8788. []

Axiomtek, Taipei City, Taiwan. +886-2-2917-4550. []




Rugged COM Express Type 6 Modules with Fourth Gen Core CPUs and Removable Memory also serves as a conduction plate to help wick A Type 6 Basic COM Express module offers a choice of Intel’s fourth-generation Core i7 or i5 (Haswell) CPU and Intel’s 8-Series QM87 PCH chipset, formerly known as Lynx Point. An innovative SODIMM hold-down method ensures secure connections while still allowing users to remove or upgrade the memory. Ruggedly designed for use in defense, aerospace and industrial applications, the XCOM=6400 Express module from Acromag features an extra-thick circuit board, advanced thermal management and extended operating temperature ranges. The XCOM-6400 provides heat sink capabilities not available on traditional COM Express designs. Conduction-cooled rails establish a new design opportunity for carrier cards. Additional heat management technologies include heat spreader plates plus options such as cooling fins and a fan. The module sets a new approach for shock and vibration by implementing a SODIMM hold-down mechanism. Soldering down the memory is no longer necessary. A screw-down latch securely fastens the memory in place to prevent it from shaking loose and

away heat. Removing the latch provides user access to two memory slots that support up to 16 Gbytes of high-speed DDR3L available in 1 x 4 Gbyte, 2 x 4 Gbyte and 2 x 8 Gbyte configurations. Enhanced graphics enable smoother playback of high-quality images. Better power efficiency reduces heat and allows smaller, lighter, more portable designs. Acromag offers a high-performance quad core i7 CPU (2.4 GHz, 47W) and more efficient dual core i5 CPU (1.6 GHz, 25W). The platform controller hub is Intel’s 8-Series QM87 PCH chipset, which provides faster connectivity and flexibility with integrated I/O technologies. For graphics, the module uses an Intel integrated graphics processor with support for a 3x digital display interface (DVI or DisplayPort) and eDP interface (x2). Audio is supported through an HDA interface. The LAN Port is a Gigabit Ethernet Medium Dependent Interface (MDI). Other I/O interfaces include four SATA III ports (6 Gbit/s), a PEG / general-purpose PCIe x16 interface with bifurcation/trifurcation support, four USB 3.0/2.0 ports and four USB 2.0 ports. Users can access the SPI bus, LPC

bus, SMBus (system) and I2C (user). There are also four general-purpose outputs and four general-purpose inputs. Security and encryption capabilities are facilitated through use of a Trusted Platform Module (TPM). A variety of model configurations start at $1995. Acromag, Wixom, MI. 248-624-1541. []

FIND the products featured in this section and more at

Feature-Rich COM Express Module with Fourth Generation Core Processor

An advanced COM Express Type 6 module adopts the fourth generation Intel Core processor (formerly codenamed Haswell) and delivers breakthrough CPU performance, stunning graphics and improved security functions. The Express-HL from Adlink Technology is among the first products in an array of industrial embedded systems delivering the computing power of Intel’s newest generation Core processor family. The new module is well suited for intelligent systems innovations in a variety of market segments, such as retail, medicine, gaming, transportation, defense, communications and industrial automation. The Adlink Express-HL adopts the PICMG COM.0 rev 2.1 form factor with Type 6 pin-out, a pin-out specification enabling a robust feature set, which—with the introduction of the fourth generation Intel Core processor family—can be fully utilized. The Express-HL supports the fourth generation Intel Core 2- and 4-core mobile processors (i7/i5) with the Mobile Intel QM87 Express Chipset and up to 16 Gbyte dual channel DDR3L SDRAM at 1600 MHz system memory. The fourth generation Core processors also offer a huge performance gain for floating-point-intensive computations by adding new instructions to Advanced Vector Extensions (AVX); these advanced instructions are especially beneficial for digital signal and image processing applications, such as medical imaging or radar. To fully tap the benefits of 4th generation Intel Core processors, the Express-HL provides rich image output options, including LVDS, analog VGA, and multiple DDI (for DP or HDMI) originating directly from the CPU to provide a considerable increase in band-

width and resolution over the previous generation Core processor. The new module can support three independent displays, all of which can be DDI (compared to past generation processor requirements of CRT/DDI or LVDS/DDI combinations), and a PCIe x16 port is available for external video cards. The Express-HL also provides a full set of high-speed I/O interfaces—with seven PCIE x1, four USB 3.0 ports and four USB 2.0 ports, as well as four SATA-III (6 Gbyte) ports. The Express-HL module is provided with a Smart Embedded Management Agent (SEMA) controller, which monitors BIOS, power, temperature, watchdog and board information. The module also supports a wide voltage input range of 8.5V~20V, and an operating temperature range from -40° to +85°C. ADLINK Technology, San Jose, CA. (408) 360-0222. []




Microcontroller Integrates 16-bit ADC, 10 Msps ADC, DAC, USB and LCD

A new family of microcontrollers (MCUs) is an analog system on a chip that integrates a full analog signal chain, including an on-chip precision 16-bit ADC and 10 Msamples/s 12-bit ADC, plus a DAC and dual operational amplifiers (op amps), along with eXtreme Low Power (XLP) technology for extended battery life in portable medical and industrial applications. This PIC24FJ128GC010 family offers a combination of analog integration and low power consumption, reduces system cost and noise, and improves the signal throughput in applications such as portable medical monitoring devices, (e.g., blood-glucose meters and blood-pressure monitors), as well as in industrial applications such as portable monitoring devices including voltage and current monitors, gas sensors and high-speed sensor arrays, among others. The PIC24FJ128GC010 family includes an integrated LCD display driver that provides the ability to drive up to 472 segments with information-rich user displays that include scrolling alphanumeric banners. Integrated USB supports the uploading of clinical data for medical equipment, and can act as a service/data port for industrial equipment. Capacitive touch sensing is supported with an on-chip mTouch peripheral. The integration of a 16-bit ADC, USB and LCD into a single low-power MCU allows for very small form factor, battery-powered applications. The PIC24FJ128GC010 family represents a significant cost reduction over a multi-chip implementation, enabling lower noise, faster throughput, smaller PCB size and faster time-to-market. The new Microchip family appears to have brought together a set of features that can be highly integrated, reducing cost, BOM, board area and power consumptions, and is widely applicable in a host of applications that can take advantage of high-precision analog processing, high computational power, low power consumption and mobility. Designing precision analog circuits is a challenge. The new family integrates most of the analog signal path to eliminate the chance of introducing noise from onboard components and circuits. Thus designers get consistent analog performance across applications. Microchip also offers a comprehensive starter kit on which customers can build their software, hardware and sensors—drastically improving their development time by eliminating the need to design a board. The PIC24FJ128GC010 family is supported by Microchip’s MPLAB Starter Kit for PIC24F Intelligent.Integrated.Analog. This kit is focused on the family’s integrated analog to preserve signal integrity. It provides 95% of what designers need to develop a handheld analog prototype—all they need to do is add sensors. Microchip Technology, Chandler, AZ. (480) 792-7200. [



Mini-ITX Motherboard with AMD Embedded R-Series APU

A new Mini-ITX Embedded motherboard is powered by the AMD embedded RSeries APU, which enables a high definition visual experience for this small form factor with low power. The MB-73330 is designed for a variety of digital image applications with single or multiple displays. Two GbE LAN ports enable networked capabilities in application areas that include advertising, medical, informational, infotainment, transportation and industrial control. Robust expansion capabilities can support incremental features desired by the OEM. MB-73330 provides up to 16 Gbytes of DDR3 SDRAM. Features include the AMD R-Series APUs, PGA (FS1r2) socket with AMD A75 chipset and up to 8 Gbyte DDR3. The module incorporates 2x GbE LAN along with x16 PCIe, Mini-PCIe and Half-Size Mini-PCIe. There are three HDMI ports, one DP and a CFast socket. Other I/O include two SATA interfaces, two COM ports and USB 3.0. Power input is DC 8V to approximately 32V. WIN Enterprises, North Andover, MA. (978) 688-2000. []

FPGA Mezzanine Clock Distribution and A/D Modules

Two new FPGA Mezzanine Cards (FMC – VITA 57.1) are designed for multichannel data acquisition and high-speed signal processing and recording. The FMC168 from 4DSP is a digitizer FMC featuring eight ADC channels with 16-bit resolution and 250 Mega samples per second sampling rate per channel. Using the newly released ADS42LB69 from Texas Instruments, the FMC168 offers a high level of performance for software defined radio, beam forming and wireless communication infrastructure equipment. The second released product, the FMC408, is a clock distribution FPGA Mezzanine Card with eight synchronized clock outputs and eight synchronized Pulse Per Second signals. Operating at frequencies up to 4 GHz while preserving excellent jitter characteristics, the FMC408 makes it possible to build high-speed multi-channel solutions with ease. The FMC408 combined with several FMC168 offers 64 fully synchronous A/D channels that can be enclosed in a single ruggedized chassis or 3U server. FMCs can be mounted on PCI-Express, VPX, CompactPCI or AMC FPGA carrier cards. With advances in FPGA device densities and signal processing capabilities on the latest Xilinx Virtex-7 and Kintex-7 platforms, it has become possible to process gigabytes of data in real time and on-the-fly. When coupled with a highspeed CPU and SSD array for storage, FMC-based solutions offer the flexibility, DSP power, scalability and ruggedization levels associated with high-end systems in the Telecommunications, Aerospace and Defense industries. 4DSP, Austin, TX. (775) 830-2059. []


PCI Express Digital I/O Module with Data Rate up to 100 MHz

A new high-speed digital I/O module supports up to 100 MHz data rate and 400 Mbyte/s throughput, with parallel and bi-directional 32-channel high-speed I/O lines and 8-channel application function I/Os. Ideally suited to large scale and high resolution digital pattern data acquisition and generation applications, the PCIe-7360 from Adlink Technology delivers high-end support for complementary metal–oxide–semiconductor (CMOS) image sensor capture for image performance verification, ADC/DAC testing, digital image transfer to large-scale printer/plotter applications and digital video recording/playback. Adlink’s PCIe-7360 utilizes a PCI Express x4 interface, with up to 400 Mbyte/s throughput and data rate up to 100 MHz guaranteed, with programmable parallel data width (8/16/24/32bit). The sampling and updating clock can derive from an internal or external timer pacer, supporting up to 80-step clock phase shift to ensure accurate data sampling and timing updates and prevent invalid timing problems during data transition.

In addition to high-speed I/O, the PCIe7360 features extra 8-channel programmable function I/O connections, which can be configured to I2C/SPI master, trigger IN/OUT, external clock IN, digital input/output sample clock exporting, and handshake modes to communicate, control, or synchronize with external devices or devices-under-test (DUT). The logic level of both high-speed and function I/O is software programmable and supports three options, including 1.8V, 2.5V and 3.3V, for the flexibility to direct the interface with external devices and DUT if different logic levels are required. The PCIe-7360 supports Windows 8, Windows 7 and Windows XP operating systems, and is fully compatible with third-party software such as LabVIEW and Visual Studio.NET. Adlink Technology, San Jose, CA. (408) 360-0222. []

XMC Quad Channel sFPDP Interface Card for High-Performance Sensor Interface

The Titan Serial FPDP XMC from Galleon Embedded Computing is designed to address demanding applications while keeping a close eye on the cost and power budgets. The Titan is designed as a native XMC card, which makes it suitable for a wide range of applications. It can easily be integrated in systems based on PCI Express backbones, such as VME and VPX processing boards, cPCI and PCIe card edge based systems. Typical applications are radar and medical imaging, where there is a need to move large amounts of streaming data between processing nodes. A x8 PCI Express interface to the host processor allows high bandwidth for full channel utilization, up to 425 Mbyte/s per sFPDP channel. An onboard 1 Gbyte DDR3 SDRAM buffer memory is provided for use as an ultra-deep FIFO in applications needing data stream elasticity. The sFPDP IP core is optimized for Xilinx Virtex 6 LX75T providing maximum performance in a power- and cost-optimized FPGA. Several variants for front and rear I/O are available, including a model with recessed optical transceivers for internal optical cabling in conduction-cooled systems with no front panel I/O. The Titan Serial FPDP XMC is available in conduction-cooled and air-cooled variants, supporting full range -40° to +85°C operational temperature range. The XCC is tested according to MIL-STD-810. Volume pricing starts at $3,995. Galleon Embedded Computing, Katy, TX. (832) 786 5007. []

Desktop Network Communication Appliance Provides Scalability with Low Running Costs

A new network communication security appliance is a desktop unit with a compact footprint based on the low-power embedded 22nm Intel Atom processor C2000 SoC, formerly codenamed “Rangeley,” with up to 8 cores and includes Intel QuickAssist acceleration technology. The new CAD-0230 network appliance from American Portwell features up to 16 Gbyte DDR3L SO-DIMM, two Intel I211 GbE ports and four GbE ports from SoC Quad MAC, two bypass segments on SoC quad ports, one SATA and CFEX storage interface, one mini-PCIe slot and reserved PCIe x1 connector, RJ-45 type console and two USB ports, 40W DC 12V power adapter, optional rack mount and a compact footprint of just 210 mm (W) x 210 mm (D) x 42 mm (H) or 8.27˝ x 8.27˝ x 1.65˝. The CAD-0230 network appliance is an attractive solution for security networking applications such as network intrusion prevention/detection; spyware control; content filter; P2P control; instant message recording for system integrators, OEM customers, software developers; and SOHO and remote/satellite office. Networking and security applications continue to grow in complexity, so systems need increased computational resources to handle increased workloads that can now include cryptography. That’s where Intel QuickAssist Technology really helps. It is specifically designed to optimize the use and deployment of algorithm accelerators on the kinds of applications where CAD-0230 is the suitable choice because it makes it easier for developers to integrate built-in accelerators into their designs. When combined with Atom processor C2000 SoC, it makes CAD-0230 an extremely powerful tool. The CAD0230 is flexible enough to provide an effective solution not only for enterprises with large networks but also is an efficient low-cost security solution for small office and home office environments. American Portwell Fremont, CA. (877) 278-8899. []

FIND the products featured in this section and more at




Newest Intel Dual-Core Processors on COM Express Modules

A high-end COM Express Type 6 module family is based on the just-introduced dualcore variants of the fourth generation Intel Core processors (code name “Haswell”). The powerful MSC C6B-8S COM Express modules from MSC Embedded are characterized by highest computing and graphics performance with simultaneously low power dissipation. In addition to the first module with Intel Core i7-4700EQ (2.4 GHz, 3.4 GHz in turbo boost mode) quad-core processor, four more economical Intel Core i3 and i5 variants with two processor cores are now available. The COM Express products are offered with Intel Core i5-4400E (2.7/3.3 GHz), i54402E (1.6/2.7 GHz), i3-4100E (2.4 GHz) or i3-4102E (1.6 GHz) processor. The thermal design power (TDP) is 37W or 25W. Depending on the type, the processors support the Intel AMT 9.0 Technology, Intel 64, the Intel Virtualization Technology, VT-d Virtualized I/O, Intel’s Trusted Execution Technology, the Intel Advanced Encryption Standard and the Intel Turbo Boost Technology. The Intel Advanced Vector Extensions (AVX) 2.0 Technology allows for high-end imaging applications.

Low-Cost Entry to Fourth Generation Intel Core Processors on COM Express

Congatec has announced expanded processor support for its conga-TS87, a pin out Type 6 COM Express module based on the fourth Generation Intel Core 2-chip solution. The most significant innovation is the introduction of Intel Advanced Vector Extensions (Intel AVX) 2.0 for improved floating point instructions in signal and image processing. In addition to the existing quad-core Core i7-4700EQ processor, other variants such as Core i3-4100E (TDP 25W) are now also supported at an attractive entry price. Extended and consistent scalability is achieved by the new Core i5-4100E processor. A total of five processors of the fourth Generation 2-chip solution are now available on COM Express. The focus of fourth Generation Intel Core 2-chip processors is on increased performance per watt. This is particularly evident in the instruction set and vector processing unit expansions, where the performance has almost doubled. Sophisticated calculations such as simulations and CAD or imaging/im-



The powerful 4600 (GT2) Intel HD graphics integrated into the processor die offers a significantly improved video and graphics performance over the third generation of Intel Core processors. Accelerated coding and decoding functions for high-resolution videos, DirectX 11.1 and OpenGL 3.2 are supported. OpenCL 1.2 allows for the additional use of the graphics engines for applications with extensive use of floating-point computations. The new MSC C6B-8S module family integrates the improved QM87 Intel 8 series platform controller hub (PCH). Two fast dual-channel DDR3L (1333/1600) SDRAM modules with a maximum storage capacity of altogether 16 Gbyte assure a great computing performance and a low power consumption. An Infineon Trusted Platform Module offers a hardware-based security functionality according to the Trusted Computing Group (TCG). The MSC C6B-8S modules have comprehensive display support offering three DisplayPort 1.2 with multi-stream-transport or HDMI digital display interfaces with a resolution of up to 3800 x 2400 pixels to connect three independent displays. Furthermore, the modules provide two embedded DisplayPort interfaces, along with a LVDS and CRT interface. The processor modules feature four fast USB 3.0 and four USB 2.0 ports, seven PCI Express x1

age processing applications for tomography, radar or optical inspection benefit greatly from this performance improvement. As an additional advantage, the processing units of the integrated GT2 graphics can be used via OpenCL programming when the 3D graphics is not used elsewhere. Thanks to the new variable coder and 4K2K support for large displays, it is possible to tackle even the most demanding multimedia and gaming applications. The Advanced Encryption Standard New Instructions (Intel AES-NI) set is more relevant than ever. It enables the offloading of particularly compute-intensive packaging and encryption routines of the well-known cryptographic algorithm AES (Advanced Encryption Standard) into hardware. This results in high-performance encryption without putting a significant burden on the CPU cores. 4K2K pixel resolution of up to 3840 x 2160 with DisplayPort and 4096 x 2304 with HDMI is natively supported in the processor. In addition, it is possible to daisy chain up to three independent display interfaces via DVI, LVDS and VGA. Seven PCI Express 2.0 lanes, PCI Express 3.0 graphics (PEG) x16 lanes for

channels, a PCI Express graphics (PEG) x 16 interface, a LPC bus, Gbit Ethernet, HD audio and four SATA interfaces at up to 6 Gbit/s. The high-performance platform runs under the Microsoft Windows 7, Windows 8 and Linux operating systems. BIOS firmware is based on the UEFI platform Aptio from AMI. In addition to the embedded modules, MSC offers starter kits and carrier boards to ensure a quick evaluation of the COM Express modules and corresponding cooling solutions like a passive and an active heat sink. The MSC C6B-8S module family is especially suitable for highsophisticated applications, for example in image recognition and processing, for displaying multiple high-resolution (up to 4k) videos or for the control of large screen displays. MSC Embedded, San Bruno, CA. (650) 616-4068. []

high-performance external graphics cards, four SATA ports with up to 6 Gbit/s and RAID support and one Gigabit Ethernet interface enable fast and flexible system extensions. Active fan control, LPC bus for easy integration of legacy I/O interfaces and Intel High Definition Audio complete the feature set. congatec San Diego, CA. (858) 457-2600. []


PCI Express Boards Enable Robust Synchronous Serial Communications

Two new single-port PCI Express serial I/O adapters are designed for avionics, satellite, radar and other applications that require robust synchronous communications. Both boards from Sealevel Systems use the Zilog Z85230 Enhanced Serial Communications Controller (ESCC) for maximum compatibility with a variety of interfaces and protocols. The 5102e multi-interface board is configurable for RS-232, RS-422, RS-485, RS-530, RS-530A or V.35, while the 5103e has an RS-232 interface. Both boards support data rates up to 128 Kbits/s. A digital phase lock loop (DPLL) circuit is included in each design for reliable communication when using NRZI or FM encoding. Sealevel’s SeaMAC synchronous serial driver provides HDLC/ SDLC protocols as well as certain configurations of monosync, bisync and raw modes for critical communication development and support. Software utilities are included to speed application development and to assist in field installation. The 5102e is available immediately from stock priced at $459 for low-profile and full-height PCI Express slots. The 5103e and 5103eS begin at $449. Standard operating temperature range is 0 - +70°C, and extended temperature versions operating from -40° to +85°C are available. Like all Sealevel I/O products, the 5102e and 5103e are backed by free technical support and a lifetime warranty. Sealevel Systems, Liberty, SC. (864) 843-4343. []

Fanless Slim-Type Network Appliance with Atom N2600 and 4 Gigabit LANs

A fanless slim-type network appliance platform for small office and home office applications supports low-power Intel Atom processor N2600 dual core 1.6 GHz with the Intel NM10 chipset. Four Gigabit Ethernet ports are integrated into the NA341 from Axiomtek with one pair LAN bypass function. It comes with a high-bandwidth DDR3 SO-DIMM slot with memory maximum up to 2 Gbyte and features one CompactFlash socket for storing event log; one Mini PCIe slot and two rotational WLAN antennas are available for wireless/3G/LTE network connection. This fanless embedded platform is very slim and light can be easily fit to any space-limited environment. The NA341 is an attractive choice for VPN, content filtering, UTM, network security gateway and firewall. In order to address the various network application needs, the NA341 features one CompactFlash slot, one console port, two USB 2.0 ports and one mini PCIe slot. For easy maintenance, there are front-facing LED indicators on the panel for power and HDD activity monitoring. The unit also supports mainstream Linux kernel 2.6 operating systems. Axiomtek, City of Industry, CA. (626) 581-3232. []

Camera Software Suite Supports USB 3.0

An extensive software package for Basler cameras allows you to quickly start up a Basler camera with all its features. The latest release now supports pylon 4 USB3 Vision and Windows 8. Basler pylon 4 includes all drivers necessary to establish the various camera interface standards (USB3 Vision, GigE Vision, IEEE1394, Camera Link). It also has a complete software development kit (SDK) with numerous programming examples and current programming languages. In addition, an easy-to-use tool lets the user develop his or her own camera application with just a few lines of code. Because the programming interface offered by pylon is uniform for all Basler cameras, regardless of which camera interface is used, the existing programming code can be used without modification for Basler USB3 Vision cameras as well. Basler pylon 4 is also the first version to provide official support for Windows, inclusive Windows 8 and Linux. The new pylon 4 Camera Software Suite can be downloaded free of charge from the Basler website. Basler, Ahrensburg, Germany. +49 4102 463 500. []

Experiment! Kits for Creative, Easy and Low-Cost Evaluation

A line-up of kits is aimed at creative evaluation for a low price. Developers can use the Experiment! Kits from IAR Systems to evaluate the embedded development tool chain IAR Embedded Workbench and the included microcontroller, as well as design, start development, integrate and test their applications. First to be introduced are a game controller kit and a magnetometer kit. The game controller kit includes a game controller, a 1.8 x 1.6” color LCD and example projects including a snake game. The magnetometer kit can be used to identify magnetic fields using the included magnetometer module and to output information on a 1.53 x 1.38” LCD. The kits include a size-limited license of the tool chain IAR Embedded Workbench and can be used with IAR Systems’ in-circuit debugging probe I-jet as well as other JTAG probes, and are also available in versions including an IAR J-Link Lite USB-JTAG/SWD debug probe. Both kits are based on the ARM Cortex-M3 STM32L152VB microcontroller from STMicroelectronics and feature several connectors: JTAG 20 pin, a small 19-pin trace connector, two UEXT/UXT connectors and a USB connector. Power is supplied through the JTAG, trace or USB connector. The board also has a prototyping area, user and reset buttons and power and user LEDs, as well as designated power/current measuring points that can be used with IAR Systems’ I-scope probe to enable detailed current and power measurements correlated with the source code in IAR Embedded Workbench. The game controller kit is priced at $119 and the magnetometer kit is priced at $94. IAR Systems, Uppsala, Sweden. +46 18 16 78 00. []




Fast Embedded Memory Screams at 400 Mbyte/s

Samsung Electronics is now producing embedded multimedia card (eMMC) 5.0 devices—in 16 Gbyte, 32 Gbyte and 64 Gbyte densities for next-generation smartphones and tablets that feature an interface speed of 400 Mbyte/s. The lightning-fast eMMC PRO memory provides exceptionally fast application booting and loading. The chips will enable much faster multitasking, web browsing, application downloading and file transfers, as well as highdefinition video capture and playback, and are highly responsive to running large-file gaming and productivity applications. Samsung’s eMMC PRO memory chips, being produced in 16, 32 and 64 Gbyte versions, are based on Samsung 64Gb 10nm class NAND flash technology. The new Samsung chips support the eMMC version 5.0 standard now nearing completion at JEDEC—the largest standards-setting body in the microelectronics industry. In 32 Gbyte and 64 Gbyte densities, the new memory solution has a random read speed of 7000 IOPS, and a random write speed of 7000 IOPS (in cache on mode, without host overhead). In addition, these chips read sequentially at 250 Mbyte/s and write sequentially at 90 Mbyte/s. As fast eMMC devices at more than 10 times the speed of a class 10 external memory card (which reads at 24 Mbyte/s and writes at 12 Mbyte/s), the new mobile memory greatly enhances the movement from one application to another in multitasking activities. Samsung’s 16 Gbyte, 32 Gbyte and 64 Gbyte eMMC 5.0 devices come in 11.5 x 13 mm packages, making them ideal for mobile devices where space on the printed circuit board is extremely limited. The eMMC PRO 5.0 chips come with Samsung’s own intelligent NAND controller and firmware to provide a vertically integrated mobile DRAM solution. Samsung Electronics, Giheung, South Korea. []



Dual Channel 3G-SDI Video/Audio Capture Card for Uncompressed Video Streaming

A new SDI video/audio capture card based on the PCI Express x4 interface offers features that enable 2-channel acquisition of 3G-SDI for low latency and uncompressed video data signals up to 1920x1080p/60fps. The PCIe-2602 from Adlink Technology provides lossless full color 4:4:4 video and up to 12-bit video data for critical applications such as medical imaging, intelligent video surveillance and analytics, and broadcasting. The PCIe-2602 supports all SD/HD/3G-SDI signals and operates at six times the resolution of regular VGA connections. It also provides unrivaled video quality with lossless full color YUV 4:4:4 images, for sharper and cleaner images. With up to 12-bit pixel depth, the PCIe-2602’s extreme image clarity and smoother transitions from color to color boost image detail for unprecedented support of high-end critical medical imaging, such as PACS (picture archiving and communication system) endoscopy, as well as broadcasting. The PCIe-2602 features low latency uncompressed video streaming, CPU offloading, and support for high-quality live viewing for video analytics of real-time image acquisition, as required in casino and defense environments. When combined with a suitable 75Ω coaxial cable, PCIe-2602 signals can be transmitted over 100 meters, requiring no significant modification of existing analog-based CCTV systems, with employment of existing coaxial cable networks representing significant savings. The PCIe-2602 is equipped with RS-485 and digital input & output—accommodating external devices such as PTZ cameras and sensors—supports Windows 7/XP operating systems, and comes with Adlink’s ViewCreator Pro utility to enable setup, configuration, testing and system debugging without requiring any software programming. All Adlink drivers are compatible with Microsoft DirectShow, reducing engineering efforts and accelerating time-to-market. ADLINK Technology, San Jose, CA. (408) 360-0200. []

Multi-Fabric Switch Enables Development of Complex, Scalable, High Performance Systems

A multi-fabric switch and PMC/XMC carrier card is designed to enable the development of complex, scalable, high-performance 3U VPX systems in today’s increasingly connected military/ aerospace world. The PEX431 from General Electric Intelligent Platforms is characterized by significant flexibility and complements GE’s recent announcement of the SBC326 3U VPX single board computer based on the fourth generation Intel Core i7 (Haswell) processor. For customers who need a multi-fabric network switching capa- bility that includes PCI Express Gen 2.0/Gen 3.0 and up to nine ports total of 1000BaseBX and 1000BaseT Gigabit Ethernet, the PEX431 supports three OpenVPX payload profiles. The PEX431 can also make a significant contribution to reducing SWaP (size, weight and power). It provides both control plane and data plane switching on a single board and, by providing the ability to be a carrier card for an XMC mezzanine card, potentially reduces the number of slots required to implement a given system—or allows more functionality to be configured in the same number of slots. Many of today’s high-performance embedded computing (HPEC) systems deploy multiple processors and multiple I/O modules within a single chassis. A single PEX431 enables six processors to be interconnected: for even higher performance systems with more than six processors, multiple PEX431s can be daisy-chained. The PEX431 is also supported by GE’s software capability that allows VPX PCI Express peer-to-peer connectivity. Without this, customers are required to “hand-craft” a solution, with consequent impact on program cost and time-to-market. General Electric Intelligent Platforms, Huntsville, AL. (780) 401-7700. []


Wonder Gecko MCU Development Kits Speed DSP Smart Sensor Design

Development kits and application software demonstrations support the EFM32 Wonder Gecko microcontroller (MCU) family from Silicon Labs. The kits were developed by Energy Micro, recently acquired by Silicon Labs. The Wonder Gecko MCU line is based on the ARM Cortex-M4 processor core, which provides a full DSP instruction set and includes a hardware floating point unit (FPU) for faster computation performance. The development kits and software examples are designed to help embedded engineers leverage 32-bit digital signal control with the high-performance CPU and extremely low standby power modes. To speed up the design time, the EFM32 development kits include a built-in J-Link debugger and come with software examples using each kit’s built-in features including an audio pre-amplifier equalizer that digitizes the audio connector signal with the MCU’s on-chip analog-to-digital converter (ADC) and subsequently generates the output via a digital-to-analog converter (DAC). There is also an audio frequency analyzer using the kit’s audio connector and performing a Fast Fourier Transform (FFT) to display a frequency plot on the development kit’s LCD. An application example using the kit’s onboard light sensor for 10-500 Hz FFT analysis is also included These software demonstrations also enable designers to evaluate the differences between hard and soft floating-point operations and compiler optimization, as well as the CPU cycle count. The example projects are coded using algorithms that are part of the Cortex Microcontroller Software Interface Standard (CMSIS) DSP function library, which includes complex FFT, finite impulse response (FIR) filters, matrix and vector operations, and statistical analysis. CMSIS provides a vendorindependent hardware abstraction layer for ARM Cortex-M processors. Silicon Labs’ complimentary Simplicity Studio software suite includes all the necessary CMSIS, board support package (BSP) and documentation for the development kits including a Wonder Gecko white paper highlighting the 32-bit processing, DSP and FPU performance benefits of the EFM32 Wonder Gecko MCU family. The white paper also illustrates how the Wonder Gecko MCUs achieve high levels of 32-bit performance while delivering best-in-class energy efficiency. Silicon Labs, Austin, TX. (512) 416-8500. []

Exceptional Positioning Accuracy in GNSS-Enabled PCI Express Mini Card

A new PCI Express Mini Card uses a global navigation satellite system (GNSS) receiver to handle data transmissions from both GPS and its Russian counterpart, the GLONASS system. The PX1 from Men Micro, which supports both active and passive antennas via a U.FL connector, provides superior satellite-based communication worldwide. A gyroscope sensor on the new Mini Card enables dead reckoning functionality, ensuring accurate position identification even when a satellite signal is interrupted, such as driving through a tunnel. This functionality, combined with the board’s GNSS receiver, makes the PX1 attractive for fleet management applications where commercial vehicles, such as trains, buses, ships and airplanes, travel across wide geographic ranges. Various satellite-based augmentation systems (SBAS) that help improve the accuracy, reliability and availability of the GNSS information are also supported by the PX1 for increased positioning accuracy. In addition to handling data generated from the current GPS and GLONASS satellite systems, the Mini Card has been developed to support communications on the pending European Galileo system, set for launch in 2014, as well as the Chinese Compass system that is on track for global availability in the future. Pricing for the PX1 is $238 USD.

6U VME SBC Features Fourth Generation Intel Core i7 architecture

A new 6U VME single board computer is based on the quad-core fourth generation Intel Core i7 architecture (Haswell) processor. The XVR16 6U VME rugged single board computer from GE Intelligent Platforms has improved capabilities that will allow it not only to address existing command/control applications, but also to be deployed in more demanding High Performance Embedded Computing (HPEC) signal processing applications such as intelligence, surveillance and reconnaissance (ISR), sonar and radar. The XVR16 offers existing users of the XVR14 and XVR15 significantly more processing power, graphics performance, functionality and I/O capability. Importantly, it does this within the same power envelope as its predecessors, enabling simple “drop in” replacement. Fourth generation Intel Core i7 architecture brings with it numerous improvements when compared with predecessor generations. These include Intel Advance Vector Extensions (Intel AVX 2.0) enhancements, which provide a significant performance improvement in floating-point-intensive computations, which are a key part of digital signal and image processing applications. Media and graphics performance is also greatly enhanced, capable of delivering 2D/3D graphics with high-quality playback enabling a compelling visual experience. Improved security comes in the form of Intel AES New Instructions (Intel AES-NI) enhancements, which use hardware to accelerate data encryption and decryption. The XVR16 features up to 16 Gbytes of memory and benefits from the 4th generation Intel Core i7 processor’s support of PCI Express Gen3 technology and USB 3.0, providing even greater bandwidth for on-board and off-board connectivity and enabling the high bandwidths required by today’s applications. It supports two onboard PMC/XMC expansion sites.The XVR16 is available in five build versions from air-cooled to fully rugged, providing cost-effective solutions in environments from benign to extremely harsh. General Electric Intelligent Platforms, Huntsville, AL. (780) 401-7700. []

FIND the products featured in this section and more at

MEN Micro, Ambler, PA. (215) 542-9575. [] RTC MAGAZINE OCTOBER 2013


Advertiser Index GET CONNECTED WITH INTELLIGENT SYSTEMS SOURCE AND PURCHASABLE SOLUTIONS NOW Intelligent Systems Source is a new resource that gives you the power to compare, review and even purchase embedded computing products intelligently. To help you research SBCs, SOMs, COMs, Systems, or I/O boards, the Intelligent Systems Source website provides products, articles, and whitepapers from industry leading manufacturers---and it's even connected to the top 5 distributors. Go to Intelligent Systems Source now so you can start to locate, compare, and purchase the correct product for your needs.

Company Page Website Advanced Micro Devices, Inc............................................................................................. 52................................................................................................ American Portwell............................................................................................................. 51............................................................................................................ ARM, Ltd.......................................................................................................................... 31.................................................................................................................. Congatec, Inc..................................................................................................................... 4.............................................................................................................. Dolphin Interconnect Solutions............................................................................................ 7.......................................................................................................... Extreme Engineering Solutions............................................................................................ 2.............................................................................................................. Intel.................................................................................................................................. Intelligent Systems Source................................................................................................. 35................................................................................... Lauterbach........................................................................................................................ 38........................................................................................................ MEN Micro, Inc................................................................................................................. 34......................................................................................................... MSC Embedded, Inc........................................................................................................... One Stop Systems, Inc...................................................................................................... Pentek, Inc........................................................................................................................ Phoenix International.......................................................................................................... 5............................................................................................................ RTD Embedded Technologies, Inc................................................................................... 26, Real-Time & Embedded Computing Conference.................................................................. 39................................................................................................................ CompactPCI & Graphics/Video Boards Gallery Showcase................................................... 25........................................................................................................................................

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.



RTC Magazine  

October 2013

Read more
Read more
Similar to
Popular now
Just for you