December 2019, Volume 21 â€“ Number 12 â€˘ cotsjournalonline.com
The Journal of Military Electronics & Computing
Artificial Intelligence Challenges the Industry Performance from Nvidia continues to change the world
The Journal of Military Electronics & Computing COTS (kots), n. 1. Commercial off-the-shelf. Terminology popularized in 1994 within U.S. DoD by SECDEF Wm. Perry’s “Perry Memo” that changed military industry purchasing and design guidelines, making Mil-Specs acceptable only by waiver. COTS is generally defined for technology, goods and services as: a) using commercial business practices and specifications, b) not developed under government funding, c) offered for sale to the general market, d) still must meet the program ORD. 2. Commercial business practices include the accepted practice of customer-paid minor modification to standard COTS products to meet the customer’s unique requirements.
—Ant. When applied to the procurement of electronics for he U.S. Military, COTS is a procurement philosophy and does not imply commercial, office environment or any other durability grade. E.g., rad-hard components designed and offered for sale to the general market are COTS if they were developed by the company and not under government funding.
SPECIAL FEATURES 16
Artificial Intelligence Challenges the Industry
By John Reardon, COTS Journal
Happy New Years – from your Friends at RTC Media.
The Inside Track
SYSTEM DEVELOPMENT 20
Performance from Nvidia continues to change the world
By John Reardon, COTS Journal
COT’S PICKS 24
Editor’s Choice for December
Cover Image A Powerpoint slide at a recent defense conference showed details of the Long-Range Hypersonic Weapon, or LRHW, a land-based U.S. Army system. The LRHW is one of multiple U.S. military projects to develop weapons that travel at hypersonic (faster than Mach 5) speeds. The Army is aiming to deploy the first experimental LRWH battery in 2023. COTS Journal | December 2019
The Journal of Military Electronics & Computing
INTERIM EDITORIAL DIRECTOR Glenn ImObersteg, firstname.lastname@example.org
DIGITAL MARKETING MANAGER Scott Blair, email@example.com
SENIOR EDITOR Rob Hoffman, firstname.lastname@example.org
ADVERTISING CONTACT John Reardon, email@example.com Aaron Foellmi, firstname.lastname@example.org
ART/PRODUCTION CREATIVE DIRECTOR & LAYOUT Designs by Dave, email@example.com DIGITAL MARKETING MANAGER Kyle Mac, firstname.lastname@example.org
FINANCE ACCOUNTING AND FINANCE Paige Reardon, Paiger@rtc-media.com
COTS Journal CORPORATE OFFICE RTC Media 3180 Sitio Sendero, Carlsbad, CA. 92009 Phone: (949) 226-2023 Fax: (949) 226-2050 www.rtc-media.com
COTS Journal | December 2019
PUBLISHED BY RTC MEDIA Copyright 2018, RTC Media. 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.
John Reardon, email@example.com VICE PRESIDENT Aaron Foellmi, firstname.lastname@example.org
John Reardon, Publisher
Happy New Years
from your Friends at RTC Media.
2020 is the decade of dreams and imagination. This is the decade where Artificial Intelligence will create defense systems once thought to be science fiction. Where technology will eliminate places to hide for our enemies; where boots on the ground will be enhanced with augmented realities. Quantum Computing and AI at the edge will tap huge reservoirs of knowledge to create highly sophisticated solutions. Ideas from smart bullets to defensive shields able to repel conventional attacks may become reality. Some technologies are simple fixes. Things like optical touch to allow a user to wear gloves when operating a touch screen. Some technologies are creative fixes. Things like modulating the clock speed of a processor to cut power needs. Some technologies are standards that will allow for high-speed interoperability of systems such as PCIe Gen 5 and 6 or 5G. Yet it is a compendium of these and others that will take us into a new world of defense. Some believe that AI will be a game changer. AI of course spurs the imagination, but it is made possible with support of all the other technologies that are 6
COTS Journal | December 2019
drawn together to build advance systems that take performance to a new high. So although AI is important, it would not be able to be utilized if it were not for the continued performance increases being achieved. One consideration that plays to the imagination is the use of AI to better secure AI systems. When detection of an irregular data stream is detected by an AI processor to protect itself from intrusion, the value of AI becomes more apparent. When highly coordinate topologies that compute at the Edge and yet draws upon the Cloud for reference and inference computing, the possibilities grow exponentially. These and other concepts excite the imagination on how the implementation of this and other technologies can be expanded to address the
most complex and sophisticated enemies. The ideas that Autonomous vehicles have spurred have brought about a whole new thought process for UAV’s. The use of Block chain to securely transfer Money has opened numerous discussions on how best to secure data. The use of Composable computing in Data Centers has contributed to the redistribution of compute regarding weapon systems. When you compile these and the many others that have occurred in the latter half the last decade, it has set the stage for this decade. This a decade where “fiction” will becomes reality, a decade where the “impossible” will seems a bit more possible.
This a decade where “fiction” will becomes reality, a decade where the “impossible” will seems a bit more possible.
COTS Journal | December 2019
DecisionEdge’s CloudEVM Is The First To Deliver Seamless Integration From Primavera P6 To Empower For Aerospace and Defense (DoD) Companies and DOE Labs Empower Optimized Integration delivers large cost and schedule data sets faster and more accurately through intelligent automation, which cuts work time from hours to minutes.
DecisionEdge announced its CloudEVM is the first to provide seamless integration from Primavera P6 to Encore’s analytical tool Empower for both the Department of Energy (DOE) Laboratories and Aerospace and Defense Companies supporting the Department of Defense (DoD). This enables labs and companies to automatically upload and export cost and schedule data resulting in faster and more accurate processing of data through intelligent automation. “For the past five years our solution has worked successfully within companies in both the DOE and DoD to bring accurate
COTS Journal | December 2019
data into Empower in minutes rather than hours,” said Mark Tillema, DecisionEdge CEO and co-founder, and the developer of CloudEVM. “Their feedback allowed us to optimize CloudEVM for this application based on real-life use cases. We look forward to further enhancing the integration between CloudEVM and Empower in the future.” CloudEVM is the first to provide the new DOE PARSII CSV export format at Los Alamos National Laboratory, Y-12 National Security Complex, and the Pantex Plant. CloudEVM’s Earned Value Management System provides seamless integration with Microsoft Office Project, Excel, ERP and accounting systems, as well as P6. For DoD applications, CloudEVM integrates data from P6 directly to Empower, and for DOE applications, it sends
the data to the Project Assessment and Reporting System (PARS), and then to Empower. “This enhanced integration using the Empower Optimized format allows our mutual customers to accelerate their monthly business rhythm so that program/project analysis and quality assessments and corrections are accomplished prior to formal reporting,” said Gary Troop, President of Encore Analytics. “This solution provides Control Account Managers instant access to performance data, insight into potential quality issues, and the ability to communicate variance analysis and corrective actions quickly. Compliance Managers use the exact same data in real-time to review data-driven audit metrics for projects and/or portfolios with complete trend data. This allows them to communicate and share data quality status to their customers at any time.”
Global expert in simulation and validation opens branch office in San Jose
dSPACE has opened a Silicon Valley office in San Jose, California. As part of its strategic realignment, dSPACE – strong in software development and HIL solutions – has been investing in technologies such as artificial intelligence, machine learning, and sensor technology to support its customers with their dynamically-changing requirements for simulation and validation solutions. ”As part of the Silicon Valley ecosystem, we want to be a key contributor to customers and technology companies in providing new solutions to the dynamic challenges facing our industry. We are bringing our global expertise as a partner for simulation and validation moving closer to our customers,” said Martin Goetzeler, CEO of dSPACE. The new dSPACE West Coast regional office is located at 2540 N. First Street, Suite 201, San Jose, CA 95131.
GenXComm announces Contract with Army Futures Command’s Army Applications Laboratory to Reduce Communication Signal Interference Adaptation of the Company’s Innovative RF Photonic Interference Cancellation Technology to Military Tactical Systems
GenXComm announced that the U.S. Army Futures Command has awarded a contract with its Army Applications Laboratory (AAL) to GenXcomm to adapt the company’s commercial analog interference cancellation technology to Army tactical systems. The project’s goal is to eliminate signal interference between Army systems and en-
able higher bandwidth, Simultaneous Transmit and Receive (STAR) communications with low Size Weight and Power (SWaP) analog radio frequency (RF) integrated circuits. The effort will result in a prototype cosite interference module integrated with Army tactical radios. The prototype will be tested and evaluated by Army U.S. Combat Capabilities Development Command (CCDC) Command, Control, Communications, Computers, Cyber, Intelligence, Surveillance and Reconnaissance (C5ISR) Center personnel for potential transition to operational Army systems. “We’re solving the problem of communication signal interference by applying GenXComm’s robust self-interference cancellation technology to enable true full duplex wireless communication,” said Dr. Sriram Vishwanath, Co-founder and CEO of GenXComm. “Our contract with the Army Applications Laboratory presents an opportunity to utilize our unique photonics platform for advanced signaling capabilities within tactical edge devices. We look forward to working with the team at Army Futures Command to shape the future of communications systems.” COTS Journal | December 2019
LYNX MOSA.ic™ Selected For F-35 Lightning II Mission System Avionics
Lynx Software Technologies announced that the F-35 Joint Strike Fighter Program Office has accepted the recommendation from Lockheed Martin (NYSE: LMT) to adopt the LYNX MOSA.ic™ framework to support the upgrade of the mission system avionics for the F-35 Lightning II as part of the “Technology Refresh 3” (TR3) modernization program. LYNX MOSA.ic™ will be deployed to support the development of key components of the next generation F-35 Lightning II avionics platform, including the Integrated Core Processor (ICP) being developed by Harris Corporation (NYSE: HRS). The ICP acts as the brains of the F-35, processing data for the aircraft’s communications, sensors, electronic warfare, guidance and control, cockpit and helmet displays. The
COTS Journal | December 2019
TR3 program is expected to generate large reductions in production and sustainment costs at the same time as delivering significant increases in computing power. A key element of the program is the adoption of commercial-offthe-shelf (COTS) technology and an Open System Architecture to enable the flexibility to add, upgrade and update future capabilities. Commenting, Gurjot Singh, CEO of Lynx Software Technologies, said, “Lynx is proud that the LYNX MOSA.ic™ framework has been selected for this very significant program. The open integration framework of LYNX MOSA.ic™ will support the overall objectives of the TR3 program to drive down development, sustainment and acquisition costs, and provide the necessary flexibility to implement system upgrades well into the future.” He continued, “LYNX MOSA.ic™ was selected because its advanced architecture fully met the needs of the project,
and because the road map and vision for LYNX MOSA.ic™ were well aligned with the future path of this program.” Will Keegan, CTO of Lynx Software Technologies, added, “The open integration framework of LYNX MOSA.ic™ will enable the TR3 subsystems to be cleanly architected from reusable software components that avoid proprietary dependencies, providing the program with more commercial options to manage supplier and manufacturing costs. LYNX MOSA.ic™ will provide additional flexibility to integrate components of varying degrees of complexity and quality, including open source components, without undermining architectural assurance properties. Finally, in providing a simpler foundation for hosting safety-critical applications, LYNX MOSA.ic ™lowers the cost, effort and risk of multicore certification compared with traditional SMP RTOS approaches.”
Crystal Group selected as CommScope Federal Partner of the Year
Crystal Group, Inc. was named CommScope’s 2019 Federal Partner of the Year at the annual Big Dogs Conference in Washington, D.C. on November 14, 2019. As CommScope’s sole supplier of rugged ICX switches for military and industrial applications, Crystal Group’s hardened and MIL-SPEC solutions deliver the most powerful networking capabilities to the warfighter and other tactical users. This mission-critical work enables IT security, reliability and performance in the harshest and most demanding environments. “As we approach the Big Dogs Conference each year, we look across our partners in the federal market for one company that truly stands out,” said Chris Collura, vice president of Federal Sales for CommScope. “We are proud to recognize Crystal Group, the
company that has single-handedly brought us into the rugged switching market and involved us in dozens of new program opportunities. We’re excited about what we’ve accomplished together so far, and the prospects for 2020 and beyond.” Crystal Group brings tactical superiority to an elite network of best-in-class, high-performance solutions and services spanning a range of markets. Over the past 18 months, the company has been awarded more than 30 different programs requiring a rugged ICX switch. “We take tremendous pride in delivering unparalleled solutions based on the latest innovations from partners such as CommScope,” said Robert Haag, vice president of Sales and Marketing for Crystal Group. “This award reaffirms the quality, value and reliability of our rugged and secure solutions. We look forward to building on this collaboration to strengthen and enable our customers’ IT infrastructures well into the future.”
Bob Haag, VP, Sales and Marketing for Crystal Group, accepted the U.S. Federal Partner of the Year award at CommScope/Ruckus’ annual Bug Dogs conference. CommScope representatives pictured with Haag are Phil Castillo, VP, North American Region, Enterprise Sales, Raelyn Kritzer, VP, Worldwide Channel and Stephen Kowel, SVP, Global Sales Enterprise COTS Journal | December 2019
PICMG COM-HPC technical subcommittee approved new highperformance Computer-on-Module specification
congatec announces that the PICMG COM-HPC technical subcommittee approved the pinout of this new high-performance Computer-on-Module specification. The new COMHPC standard is now entering the home stretch for the ratification of version 1.0 of the specification, which is scheduled for the first half of 2020. C o m p u t e r - o n - Mo d u l e manufacturers and carrier board designers who are active in the COM-HPC workgroup can now embark on first edge computing designs based on this pre-approved data, with the expectation to bring them to market in time with the launch of new high-end embedded processor generations from Intel and AMD next year. PICMG president Jessica Isquith is delighted with the progress of the COM-HPC specification: Within PICMG we are currently working on the next generation Computer-On-Module standard which is of utter importance for the embedded and edge computing world. Next to the physical footprint, the pinout is the most essential milestone. It could only be pre-approved so quickly because we managed to get all key market players, including semiconductor manufacturers such as Intel, around one table in the COM-HPC technical subcommittee, thereby also making sure that the standard will be the best fit possible for future processor generations. Committee chairman Christian Eder is confident that the specification can be officially ratified before the next high-end embedded 12
COTS Journal | December 2019
processors hit the market: A new Computer-on-Module specification is a complex task that involves many stakeholders. However, we officially started our work back in October 2018 and are on schedule to release new COM-HPC modules, carrier boards and solution platforms in time with the next high-end embedded processor generations. They will extend the existing PICMG COM Express module standards with new solutions that move in the direction of headless edge server and more multifunctional edge client solutions.
With the adoption of the pinout, all committee members now have a solid working basis from which to offer interfaces supporting up to 100 GbE and PCIe Gen 4.0 and Gen 5.0 as well as up to eight DIMM sockets and high-speed processors with more than 200 watts on standardized COM-HPC modules, and to work on
standard-compliant carrier board designs. Members of the PICMG COM-HPC committee include the University of Bielefeld and Adlink, Advantech, Amphenol, AMI, congatec, Elma Electronic, Emerson Machine Automation Solutions, ept, Fastwel, GE Automation, HEITEC, Intel, Kontron, MEN, MSC Technologies, N.A.T., Samtec, SECO, TE Connectivity, Trenz Electronic and VersaLogic. Adlink, congatec and Kontron are also committee sponsors. Christian Eder, Marketing Director
of congatec, is the chairman of the COM-HPC committee. He was previously involved as draft editor in the development of the current COM Express standard. Stefan Milnor from Kontron and Dylan Lang from Samtec support Christian Eder in their respective functions as editor and secretary of the PICMG COM-HPC committee.
Dialog Semiconductor and Flex Logix Establish Strategic Partnership for Mixed Signal Embedded FieldProgrammable Gate Arrays (eFPGA) Tweet
Dialog Semiconductor plc (XETRA:DLG) and Flex Logix® Technologies, Inc. announced a strategic agreement for Dialog to license Flex Logix’s EFLX® Embedded Field-Programmable Gate Array (eFPGA) technology for use in high volume semiconductor Integrated Circuits (ICs), and the EFLX Compiler to program these embedded FPGAs. EFLX is a low-power fully functional FPGA that is integrated into SoCs, microcontrollers and other standard or custom ICs, rather than serving as a standalone chip. EFLX cores can be used to upgrade I/O protocols, change encryption algorithms to improve security, enable elements of software-defined ra-
Great River Technology joins the SOSA consortium ARINC 818 video protocol is moving forward through Industry Global leader of ARINC 818 video protocol and 10+ years European partner of TECHWAY, Great River Technology announces its membership with the SOSA (Sensor Open Systems Architecture) consortium,
dio or accelerate data center algorithms. The EFLX arrays are programmed using Verilog or VHDL; and the EFLX Compiler takes the output of a synthesis tool such as Synopsys Synplify and handles packing, placement, routing, timing and bitstream generation. When loaded into the array, the bitstream programs it to implement the desired RTL function. “Adding eFPGA functionality to our products will give our customers the flexibility to keep pace with rapidly changing market needs,” said Davin Lee, Senior Vice President and General Manager, Advanced Mixed Signal Business Group at Dialog Semiconductor. “By partnering with Flex Logix and leveraging its EFLX eFPGA, there is a massive opportunity to increase the configurability of future Dialog products within several of our target markets, such as IoT, computing, storage and mobile.” “We are proud to partner with Dialog further deepening its industry-wide relationships. The goal of this membership is to leverage ARINC 818’s proven performances in Avionics applications to integrate other high-speed sensors and components requiring a robust, low-latency, high-bandwidth transport protocol. As sensor systems increase in number, applications, cost and complexity, users need to address issues such as affordability, versatility and capabilities. Sensor systems should be rapidly reconfigurable and reusable by a greater number of stakeholders. The SOSA consortium is introducing this culture of re-use by creating modular open system architectures dedicated to military and commercial sensor systems that adjust stakeholder interests. These architectures use widely supported, consensus-based, nonproprietary standards. The SOSA initiative enables enhance and accelerate the deployment of affordable, capable, interoperable sensor systems.
Paul Grunwald, Chief Systems Architect at Great River Technology
“Originally, ARINC 818 was intended for cockpit displays. In the last few years, we have seen the protocol adopted for cameras, headsup displays, enhanced vision and degraded en-
Semiconductor to deliver additional programmability and flexibility for their customers,” said Geoff Tate, CEO of Flex Logix. “This is just the beginning of a long-term relationship and innovative product roadmap for Dialog and Flex Logix.”
Davin Lee, Senior Vice President and General Manager, Advanced Mixed Signal Business Group at Dialog Semiconductor
vironment displays, and now high-speed EOIR and radar applications. The features that make ARINC 818 good for cockpit displays such as robust error checking, low-overhead and low latency, and high bandwidth, along with a proven pedigree with millions of hours in hundreds of aircraft make it ideal for sensors. We are looking forward to working with SOSA and its members to evaluate if ARINC 818 makes sense to be part of the standard.” says Paul Grunwald, Chief Systems Architect at Great River Technology.
Interested in getting your copy of
J O U R N A L
3180 Sitio Sendero, Carlsbad, CA. 92009 Phone: (949) 226-2023
www.rtc-media.com COTS Journal | December 2019
Curtiss-Wright Defense Solutions to Expand Position in U.S. Navy Shipbuilding and Upgrade Markets with Acquisition of 901D Holdings, LLC Will add leading supplier of U.S. Navy Ruggedized Shipboard Enclosures and Electronic Subsystems to existing portfolio of U.S. naval fleet products and services for DoD maritime customers
Curtiss-Wright’s Defense Solutions division announced today that it expects to significantly enhance its position in and commitment to the U.S. and international markets for ruggedized shipboard enclosures and electronic subsystems following Curtiss-Wright Corporation’s announced intention to acquire 901D Holdings, LLC. The acquisition is expected to close in December 2019, subject to the receipt of regulatory approval and other closing conditions. 901D is a leading designer and manufacturer of mission-critical integrated electronic systems, subsystems and ruggedized
Microdata Telecom Announces the Acquisition of Kaelus Microdata Telecom, a supplier of high-quality RF Conditioning solutions, announces that it has acquired the Kaelus business from Infinite Electronics, Inc., a US-based global supplier of electronic components. Kaelus is a globally recognized leader in test and measurement instruments, cell-site filters, combiners, and tower-mounted amplifiers. Formed in 2010, Kaelus provides the telecommunications industry with high-quality testing and conditioning products that enable global communication. Kaelus brings strong engineering capabilities, world-class manufacturing operations, a diversified set of products, strategic customer relationships with many of the world’s largest telecom operators and is poised for future growth. “The acquisition strengthens Microdata
shipboard enclosure solutions supporting every major U.S. Navy platform. The addition of 901D’s product portfolio will greatly expand the breadth of the Defense Solutions division’s products and services, and will enable it to better serve its customers in the U.S. and worldwide.
“With the acquisition of 901D, Curtiss-Wright’s Defense Solutions division firmly establishes its position as a leading designer
Telecoms’ presence in the global telecommunications marketplace by a broader portfolio, customer base and more channels to market, while also providing attractive design and manufacturing synergies,” says Mats Gustavsson, Executive Chairman of Microdata Telecom.
and manufacturer of MIL-SPEC Naval Shipboard ruggedized enclosure and electronic subsystems,” said Lynn Bamford, Senior Vice President and General Manager, Defense and Power. “Combined with our current mission-critical maritime solutions, this acquisition further expands our ability to provide shipboard solutions on all U.S. aircraft carriers, submarines, surface combatants, and amphibious ships.” About 901D Founded in 1999, 901D provides complete design, analysis, integration and testing of MIL-SPEC subsystems and ruggedized COTS products. 901D’s products include qualified electronic system cabinets, modular console and workstation designs that support a wide array of surface ship and submarine applications. It provides customers with a single source for complete, integrated subsystems ready for software installation. The acquired business will operate within Curtiss-Wright’s Defense segment. 901D is headquartered in Airmont, New York, and has approximately 85 employees.
ship same-day. This divestiture will allow us to better focus on this core and strategic growth areas.”
Kaelus and Microdata Telecom have a 3-year history of collaboration and joint business interests and share a culture that emphasizes innovation, flexibility, and performance. “We complement each other very well and we are excited about what this combination will deliver to our customers, partners, and employees around the world,” says Peter Sandberg, Chief Executive Officer of Microdata Telecom. “Microdata Telecom has a strong reputation in the industry and is an excellent strategic fit for the Kaelus business and its talented employees,” says Penny Cotner, CEO of Infinite Electronics. “Infinite’s strength is in serving engineers’ urgent needs for electronic components and assemblies, with a broad portfolio of products available to
Penny Cotner, CEO of Infinite Electronics
COTS Journal | December 2019
Artificial Intelligence Challenges the Industry By John Reardon, COTS Journal It is a truly exciting time to be a part of the defense technology business. As the consumer market drives the use of AI technologies for use in house hold products from vacuum cleaners to doorbells, the defense market is challenged by a global race to transfer these and others to the defense systems of tomorrow. For more than four decades, the defense market has drawn technology from the desktop market and the handheld market to be used in the most advanced systems possible. Using the economy of scale, we were able to draw from high volume processers and open architectures to create very advanced systems. Today although, we are still drawing from this consumer market, the concepts that are coming forth are no longer based on a single hardware structure tied to a uniquely independent software platform. Today the whole application is addressed in a way that ties application to composable system architecture. Today’s systems consist of a multi-dimension-
COTS Journal | December 2019
al configuration from the Cloud to the Fog and then out to the Edge. The Cloud affords the benefit seen in the consumer market by enabling the compute requirements to scale dynamically. Today’s Cloud based systems can dynamically scale to the needs. This composable feature not only is a cost saving, but also extends the performance to extreme compute needs.
detect IEDs – General Microsystems. It is a RADAR array detecting an incoming threat – CEMTROL. As we move forward to the front line of defense, the need for truly real-time response increases exponentially.
Fog Computing resembles a cashe of compute ability between the Cloud and the Edge. This might be on a platform such as an aircraft circling high above an area of hostilities or it might be in a command center managing a drone. As the defense market relies on real-time response with the lowest latencies possible, this cashe of compute is perfect for command and control center near the front. The Edge is the pointy end of the sword. It is the sensor on the front of the troop transport to
Image 1 - This is an example of an array of servers and systems that GMS offers for harsh environments
So as we move from the Cloud to the Fog and then on to the Edge, the environmental concerns increase in ways that were never considered by those who reside in the consumer market. It was from companies in our space that understand that designing a system for a defense application was a challenge. Challenges like fitting a system into the nose of an aircraft, that can withstand -40c to +70c, and can operate under significant power constraints was an example of what was needed. It was companies in our space that can take high resolution, real-time video from multiple sources and identify threats and react preemptively – One Stop Systems.
Image 3 - Xilinx has created a collaborative library of design and support tools that aid in the use of their FPGAs - VITIS
cation. Image 2 - One Stop Systems has delivered high performance storage arrays for airborne applications
high performance compute engines in a short amount of time. Xilinx expects to add to the library using a collaborative model to affirm that their FPGA’s are not hindered by re-creating the wheel with each appli-
Mercury Computer Systems, in a strong alliance with NVidia has incorporated GPU’s into their deployed systems. Although the
power demands and the space requirements place these systems more in the Fog, Mercury continues to drive to the Edge with compute intense solutions that are driven on the principal of SWAP – Space, Weight and Power. Mercury has a rich history with digital signal processing, but their recent move to build custom ASICs reflects the depth of IP this company posses. Another company that is putting high performance compute systems into the field is Aitech. This company has shown its talents for packaging advance GPUs into ATR Boxes. Their environmental understanding has allowed them to design systems that go from Rugged through Mil to Space. If extreme performance in a package to endure the harshest environments is needed, Aitech has an array of solutions.
This three dimensional chessboard of Cloud, Fog and Edge is made all the more interesting when you apply the Machine Learning and Artificial Intelligence. This application a multidirectional understanding that will increase an accuracy of a decision based on a compendium of knowledge. Until this point we have consumed information in a serial stream, much like we read a book from left to right. A software solution such as BERT - Bidirectional Encoder Representations from Transformers – from Google is reported to increase accuracy by twenty percent along with lower latencies. This is accomplished by a predictive algorithm, which reflects the beginning of something that will change computing forever. In a recent release from Xilinx, their VITIS software suite offers AI and Machine Learning tools to assist developers in the use of their FPGA’s. By offering these libraries to the market, Xilinx has reduced design times greatly, allowing applications to be addressed with
Image 4 - Mercury has taken a lead in developing advance systems from the ground up for the most demanding applications COTS Journal | December 2019
Image 5 - Aitech with the support of Nvidia has utilized the Jetson class GPU in ATR enclosures able to withstand the
From the images captured via a FLIR Thermal image camera or LIDAR in an autonomous vehicle, General Microsystems has proven that their solution is field-tested and meets vehicle standards such as VICTORY. GMS’s Titan Rack mounted server’s come in an array of configurations that use the latest Intel Xeon processors and offer numerous configurations of I/O and Storage. These are but a few of the companies that offer solutions that will form the foundation for tomorrow’s high performance AI compute
COTS Journal | December 2019
engines. And although we briefly touched on software, companies such as Lynx Image 6 - Lynx takes a new approach at MOSA compliant OS with their MOSAic Software Works with their Mosaic will also 2020 may not be exactly what the futurist contribute to this energy of bringing this three dimensional world to life by addressing predicted at the turn of the Century, but the the security demands of this highly complex promise of AI does ignite the imagination of networks with numerous areas of security what could be! concerns. And do this with an elegance that enables the certified reuse of code.
Performance from Nvidia continues to change the world. By John Reardon, Editor
TensorRT 7 – Accelerate End-to-end Conversational AI with New Compiler NVIDIA’s AI platform is the first to train one of the most advanced AI language models — BERT (Bidirectional Encoder Representations from Transformers) — in less than an hour and complete AI inference in just over 2 milliseconds. This groundbreaking level of performance makes it possible for developers to use state-of-the-art language understanding for large-scale applications they can make available to hundreds of millions of consumers worldwide. Early adopters of NVIDIA’s performance advances include Microsoft and some of the world’s most innovative startups, which are harnessing NVIDIA’s platform to develop highly intuitive, immediately responsive language-based services for their customers. Limited conversational AI services have existed for several years. But until this point, it has been extremely difficult for chatbots, in-
telligent personal assistants and search engines to operate with human-level comprehension due to the inability to deploy extremely large AI models in real time. NVIDIA has addressed this problem by adding key optimizations to its AI platform — achieving speed records in AI training and inference and building the largest language model of its kind to date. “Large language models are revolutionizing AI for natural language,” said Bryan Catanzaro, vice president of Applied Deep Learning Research at NVIDIA. “They are helping us solve exceptionally difficult language problems, bringing us closer to the goal of truly conversational AI. NVIDIA’s groundbreaking work accelerating these models allows organizations to create new, state-of-the-art services that can assist and delight their customers in ways never before imagined.”
How Bert Works
BERT uses Transformer that learns the contextual relationship between words in a text. Transformers include two separate mechanisms – an encoder that reads the text input and a decoder that produces a prediction for the task. In directional modes, which read text left-to-right, the Transformer encoders read the entire sequence of words at once. By reading the entire sequence at once it allows the model to learn the context of the words based on all of its surrounding (left and right of the word). Using two training strategies, BERT over comes the directional approach that limits context learning and creates a richer predictive model. 20
COTS Journal | December 2019
TensorRT-7 TensorRT™ is an SDK for high-performance deep learning inference. It includes a deep learning inference optimizer and runtime that delivers low latency and high-throughput for deep learning inference applications. TensorRT-based applications perform up to 40x faster than CPU-only platforms during inference. With TensorRT, you can optimize neural network models trained in all major frameworks, calibrate for lower precision with high accuracy, and finally deploy to hyperscale data centers, embedded, or automotive product platforms. TensorRT is built on CUDA, NVIDIA’s parallel programming model, and enables you to optimize inference for all deep learning frameworks leveraging libraries, development tools and technologies in CUDA-X for artificial intelligence, autonomous machines, high-performance computing, and graphics. TensorRT provides INT8 and FP16 optimizations for production deployments of deep learning inference applications such as video streaming, speech recognition, recommendation and natural lan-
guage processing. Reduced precision inference significantly reduces application latency, which is a requirement for many real-time services, auto and embedded applications. You can import trained models from every deep learning framework into TensorRT. After applying optimizations, TensorRT selects platform specific kernels to maximize performance on Tesla GPUs in the data center, Jetson embedded platforms, and NVIDIA DRIVE autonomous driving platforms.
TensorFlow 2.0 with Tighter TensorRT Integration To help developers build scalable ML-powered applications, Google released TensorFlow 2.0, one of the core open source libraries for training deep learning models. In this release, developers will see up to 3x faster training performance using mixed precision on NVIDIA Volta and Turing GPUs. TensorFlow 2.0 features tighter integration with TensorRT, NVIDIA’s high-performance deep learning inference optimizer, commonly
COTS Journal | December 2019
used in ResNet-50 and BERT-based applications. With TensorRT and TensorFlow 2.0, developers can achieve up to a 7x speedup on inference. This update also includes an improved API, making TensorFlow easier to use, along with higher performance during inference on NVIDIA T4 GPUs on Google Cloud. For use in multiple GPUs, developers can use the Distribution Strategy API to distribute training with minimal code changes with Kerasâ€™ Model.fit as well as custom training loops.
Using MATLAB and TensorRT on NVIDIA GPUs MathWorks released MATLAB R2018b which integrates with NVIDIA TensorRT through GPU Coder. With this integration, scientists and engineers can achieve faster inference performance on GPUs from within MATLAB. The high-level language and interactive environment MATLAB provides enables developers to easily create numerical computations and algorithms with various visualization and programming tools. A new technical blog by Bill Chou, product manager for code generation products including MATLAB Coder and GPU Coder at MatWorks, describes how you can use MATLABâ€™s new capabilities to compile MATLAB applications into CUDA and run on NVIDIA GPUs with TensorRT. The example workflow includes compiling deep learning networks and any pre- or post processing logic into CUDA, testing the algorithm in MATLAB, and integrating the CUDA code with external applications to run on any modern NVIDIA GPU. When using TensorRT with GPU Coder for a simple traffic sign detection recognition (TSDR) example written in MATLAB, the team noticed 3x higher inference performance with TensorRT running on Titan V GPUs compared with a CPU-only platform. 22
COTS Journal | December 2019
COT’S PICKS Marvin Test Solutions Completes SmartCan™ Deliveries to Israeli Air Force Marvin Test Solutions, Inc. announced that it has successfully completed deliveries of all MTS-3060A SmartCan Universal O-Level Armament Test Sets to the Israeli Air Force (IAF). The rugged MTS-3060A SmartCan™ is the most advanced flightline hand-held test set available, capable of testing all Alternate Mission Equipment (AME) and Aircraft Armament Equipment (AAE) including pylons, launchers, bomb racks, and pods. Addressing an armament test gap that exists across legacy 4th and 5th generation aircraft equipped with Smart weapon technology (MIL-STD-1760), the MTS-3060A enables support for any aircraft and any Smart or legacy armament system. “The IAF has selected the SmartCan from Marvin Test Solutions as a universal flightline armament test set for F-16 and F-15 because of its robust test capabilities,” said Major Dror Hadar from the Israeli Air Force. “With the SmartCan, IAF can replace several obsolete test sets with a single test solution that is smaller, easier to use,
Pixus Develops New OpenVPX Chassis Manager Pixus Technologies, a provider of embed-
and still provides more test capabilities than all the obsolete test sets it replaces.” “We are pleased to support one of the U.S.’ closest allies with a proven universal flightline test solution that greatly improves warfighter effectiveness,” said Major General Stephen T. Sargeant, USAF (Ret.), CEO of Marvin Test Solutions. “Continuing to gain acceptance worldwide, the SmartCan reduces the time and number of maintainers required to perform all armament testing, including the critical MILSTD-1760 functionality test required to ensure successful Smart munitions employment, while reducing logistics and training needs. Furthermore, the SmartCan provides customers the most cybersecure, hand-held O-Level armament test set available.”
tions and armament, the MTS-3060A has the capacity to grow as new and more complex weapon systems are added to the inventory. The SmartCan is currently deployed in over 12 countries on multiple aircraft and utilized by the U.S. Air Force Test Pilot School as a Smart weapons simulator for JDAM and AMRAAM. Marvin Test Solutions, Inc. www.marvintest.com
Incorporating advanced measurement capabilities to support new and Smart muni-
ded computing and enclosure solutions, now offers a chassis management module for OpenVPX that is compliant to the VITA 46.11 specification for system management.
The SHM200 chassis manager for OpenVPX comes in 0.8” or 1.0” pitch sizes. The unit can plug into a standard 3U VPX backplane slot, with versions with just the P0 connector or the P0/ P1 connectors for access to extended I/O. A 6U panel can also be affixed to plug into a 6U slot. The standard panel interface has RS-232, RJ-45, USB and LEDs. Monitoring of at least 6 temperature sensors and 10 fans is available on the SHM200 with custom options available. There is fan PWM/ Tach control, along with 16 digital inputs and outputs. A Level-2 unmanaged 3- port Ethernet hub is also an option. The Pixus SHM200 solution also includes a Web interface to remotely query the system. The user can import images of their boards. By clicking on the image for each slot, the user can drill down to the monitored status of the module. Pixus Technologies pixustechnologies.com
COTS Journal | December 2019
COT’S PICKS SEGGER’s Ozone debugger adds support for custom instructions
SEGGER just released Ozone V3.10, a major update to its full-featured graphical debugger and performance analyzer. The new release includes support for custom instructions, allowing disassembly of application specific instructions added to the CPU core to tailor it for the particular application. Custom instructions are defined in plugin scripts, which can be easily edited by the user of Ozone or the CPU core designer. The software already comes with a plug-in script for the popular RISC-V PULP extension. Custom instructions are used to make embedded systems more efficient. Tailoring
instructions to a specific application increases speed and reduces power consumption. The use of customized instructions, now available for RISC-V and ARM, is becoming more widespread. The new release comes with other new features and enhancements as well, most notably a new unified timeline window. The unified timeline window now shows all time-related values in a single window, so that instruction trace, power consumption, and data values can be viewed at a glance, while at the same time reducing the amount of screen real estate being used. Having this information in the same timeline can reveal important system information, especially the correlation between power, data, and code execution. Ozone works cross-platform for Windows, macOS, and Linux. Like most SEGGER software, Ozone can easily be downloaded and installed without a registration process. Ozone is available for free with a J-Link PLUS, PRO, ULTRA+, or with a J-Trace. Other J-Link models can use Ozone under the terms of SEGGER’s friendly licensing, allowing use for non-commercial purposes as well unlimited evaluation even for the commercial user. SEGGER www.segger.com
Annapolis Micro Systems, has developed a next-generation WILDSTAR™ 100GbE 6E10 6U OpenVPX Switch. It is SOSA™-aligned and VITA 65-compliant The 6E10 Switch provides up to 6.4 Tb/s of switching between backplane slots of multiple channels of 100Gb Ethernet. It also has front panel I/O to connect to external data sources. The 6E10 is a turnkey solution with dense,
flexible I/O. Customers have the option for seven 40/100GbE optical interfaces to VITA 66, or four 40/100GbE optical interfaces to the front panel. The rugged high-performance Switch is readily integrated into any SOSA-aligned 6U VPX system, or deployed within Annapolis’ proven WILD100™ EcoSystem. Annapolis has also developed a 3U VPX Switch that is SOSAaligned, the WILDSTAR 3E10 3U OpenVPX Switch. “Annapolis understood the potential of SOSA early on, and we became a dedicated SOSA contributor and member,” said Noah Donaldson, Annapolis Micro Systems Chief Technology Officer. “We are committed to developing the highest-performing SOSA-aligned products that are first to market.” Annapolis Micro Systems www.annapmicro.com
COTS Journal | December 2019
Curtiss-Wright Announces New NVIDIA® Quadro® Turing™ TU104/106 GPGPU Processor Modules for ISR/EW and AI Applications
further extend Curtiss-Wright’s proven leadership as a supplier of the most advanced computing solutions for embedded ISR applications. To meet demanding rugged military and aerospace specifications, these GPGPU boards feature a chip-down design.
New GPGPU modules expand Curtiss-Wright’s family of high performance embedded computing (HPEC) processors for artificial intelligence (AI) applications
“The introduction of these three new embedded AI-engines brings NVIDIA’s industryleading Turing GPU architecture to deployed defense solutions,” said Lynn Bamford, Senior Vice President and General Manager, Defense and Power. “These newest additions to our family of COTS-based modules further strengthens our commitment to being the embedded industry’s proven and trusted HPEC solution provider.”
Curtiss-Wright’s Defense Solutions division announced that, through its Reseller Agreement with WOLF Advanced Technology, it has expanded its family of open architecture high performance embedded computing (HPEC) processors designed for demanding ISR applications with the addition of three new NVIDIA Quadro Turing (TU104/6) GPU/inference engine-based OpenVPX™ modules. Curtiss-Wright also announced the availability of a new AMD Radeon™ (E9171) based XMC graphics engine card. Designed to support computeintensive ISR and EW systems, the fully rugged VPX3-4925, VPX3-4935, and VPX6-4955 modules feature Tensor Cores (288, 384, and 768 respectively) that are ideal for accelerating tensor/matrix computation used for deep learning neural network training and inference used in deployed and artificial intelligence (AI) applications requiring TFLOPS of accelerated processing. These applications include high-performance radar, SIGINT, EO/IR, data fusion ingest, processing and display, and autonomous vehicles. The size, weight, and power (SWaP) optimized VPX3-4925 module, a 3U OpenVPX GPGPU processor, features a NVIDIA Quadro Turing TU106 (RTX3000E) GPU that delivers 6.4 TFLOPS/TIPS performance. It provides 2304 CUDA® cores, 288 Tensor Cores and 36 ray-tracing (RT) cores. For higher performance in SWaP-constrained applications, the 3U VPX3-4935 module, features a NVIDIA Quadro Turing TU104 (RTX5000E) GPU that delivers 11.2 TFLOPS/TIPS. The VPX3-4935’s higher core count includes 3072 CUDA Cores, 384 Tensor Cores, and 48 RT Cores. For more demanding applications, the 6U form factor VPX6-4955 (6144 CUDA cores, 768 Tensor Cores, 96 RT Cores) hosts dual TU104 GPUs for 22 TFLOPS/ TIPS performance. Designed to work in conjunction with NVIDIA TensorRT™, CUDA, the modules’ Turing Tensor Cores add INT8 and INT4 matrix operation while continuing support for high-precision workloads. These state-of-the-art GPGPU modules 26
COTS Journal | December 2019
Complete System-Level HPEC Solutions The VPX3-4925, VPX3-4935, and VPX6-4955 modules are fully interoperable with CurtissWright’s broad family of 3U and 6U system-level C4ISR/EW OpenVPX solutions. SWAP-constrained systems can pair the single GPGPU VPX3-4925/35 modules with the Intel® Xeon®-D processor-based CHAMP-XD1 DSP engine, and together can augment the powerful sensor processing capabilities of our Xilinx® FPGA-based VPX3-534/535 transceiver modules. For higher performance 6U systems, the dual GPGPU VPX6-4955 can pair with our dual Intel Xeon-D processor based CHAMP-XD2 DSP module and powerful CHAMP-FX4 triple-FPGA processer module workhorse, along with many other modules in Curtiss-Wright’s Fabric40™ family of 40 Gbps products. Numerous system configurations can be formed using combinations of these boards. For example, a single CHAMP-XD2 node can be used to control both Turing TU104 GPUs on a VPX6-4955, or each node on a CHAMP-XD2 can be mapped to each individual GPU. Taking this design one step further, a single Xeon-D on the CHAMP-XD2 can control one VPX6-4955 board upstream, while the other Xeon-D
maps to a second VPX6-4955 module downstream. This processing slice, consisting of a CHAMP-XD2 and two VPX6-4955 modules, delivers an impressive 45 TFLOPS of performance that can be connected to the rest of the system via multiple 40 Gbps Ethernet/ InfiniBand™ ports available on the OpenVPX data plane, and through a Curtiss-Wright VPX6-6802 central fabric switch. OpenHPEC™ Accelerator Suite™ Support Curtiss-Wright HPEC modules and systems are supported by the OpenHPEC Accelerator Suite of best-in-class software development tools. These powerful tools enable system developers to develop their software faster. For example, the suite includes the powerful Bright Cluster Manager from Bright Computing, an NVIDIA partner. Bright Cluster Manager provisions and monitors both the CPU and GPU boards, and includes a fully configurable module environment. The OpenHPEC tool suite includes Bright’s deep learning libraries and tools from both Intel and NVIDIA, including Caffe and TensorFlow. It also provides the Arm® Forge suite, which enables true system level debugging and profiling for both CPU and GPUs, and supports MPI, OpenMP, and ACC. For high-speed, low-latency, peer-to-peer communications, the OpenHPEC Accelerator Suite also includes Dolphin’s PCIe communication library, which hides the complexities of directly programming the system’s PCIe devices. In addition to supporting GPU sharing between the CPUs, the Dolphin library also supports both CPU direct and remote direct memory access (RDMA). Use of the OpenHPEC Accelerator Suite simplifies, speeds, and lowers the cost of ISR application development. These tools deliver the benefits of open standard High Performance Computing (HPC) software to the COTS market to effectively remove risk when developing large scale embedded computer clusters. Resulting from its Reseller Agreement with WOLF Advanced Technology, the VPX3-4925, VPX3-4935, and VPX6-4955 HPEC modules have been pre-validated. They complement CurtissWright’s previously announced family of NVIDIA Pascal™ GPGPU modules by speeding and easing the integration of HPEC solutions into deployed systems. Curtiss-Wright www.curtisswright.com
COT’S PICKS Acceleration Dev Kit Supporting New Software Accelerator Product Line
Solution Targets Multiple Apps Needing Additional High-Speed Extension of FPGAs, Up to 100x Performance of DRAM Random Access Rate MoSys, Inc. announced the availability of its Cheetah Development Kit featuring an integrated MoSys®BE-3 or PHE Device and Virtex UltraScale + VU9P FPGA. The new development kit provides an integrated Xilinx FPGA and a tightly coupled low latency 1Gb MoSys BE-3 or BLAZAR PHE (Programmable HyperSpeed Engine or PHE) that enables rapid implementation of any algorithms that are benefited by offloading them from the FPGA. “The Cheetah card offers users the ability to
offload functions and algorithms into the PHE, which frees up resources in the FPGA that are better suited to higher priority functions,” stated Michael Miller, MoSys CTO. “By enabling designers to add system functionality, users can tailor a solution and provide a 10x to 100x performance improvement over traditional FPGA implementations.”
The Cheetah card is also capable of supporting user applications looking to replace multiple QDR devices with a single MoSys 1Gb high-speed random-access memory (which allows up to 5G accesses into and out of the single device). The card includes a MoSys BE-3 or PHE and supports all Xilinx VCU1525 applications. It is a dual Slot PCIe form factor (Full Height and Length), with the MoSys BE3 or PHE directly attached to the Virtex UltraScale+ SerDes. There are 16 lanes of GTY SERDES connected to the PHE that provide up to 400Gbps full duplex bandwidth. MoSys recently announced its Packet Classification and Graph Memory Platform on its new Software Accelerator Product Line. The Cheetah card is supporting hardware for this platform and the PHE version with 32 RISC cores has been shown at multiple trade shows, executing a high-performance, TCAM-alternative function. MoSys, Inc. mosys.com
Analog Devices Introduces Wideband RF Transceiver to Simplify System Design and Reduce Costs for Base Station Developers Highly Integrated ADRV9026 Transceiver delivers the lowest power, widest frequency range and smallest footprint in a single-chip for both FDD and TDD systems Analog Devices has introduced a new wideband transceiver within its RadioVerse™ design and technology ecosystem. The ADRV9026 is designed to support base station applications including single and multi-standard 3G/4G/5G macrocell base stations, massive MIMO (M-MIMO) and small cell systems. The ADRV9026, ADI’s fourth-generation wideband RF transceiver, offers quad-channel integration with the lowest power, smallest size, common-platform solution. The new software defined transceiver supports both Frequency Division Duplex (FDD) and Time Division Duplex (TDD) standards, simplifying design and reducing system power, size, weight and costs for 3G/4G/5G applications.
The RadioVerse design and technology ecosystem offers a one-stop radio design environment focused on simplifying the radio development process for a wide range of markets and applications. The RadioVerse ecosystem includes rapid prototyping platforms, chip-level evaluation systems, simulation tools and development kits as well as a global partnership network that provides multiple levels of design support. Analog Devices www.analog.com
COTS Journal | December 2019
COT’S PICKS TopLine and VPT Components Announce Column Attachment Services for RadHard FPGA and ASIC Devices TopLine announce expanded collaboration and technical cooperation with VPT Components, a MIL-PRF-38535 certified semiconductor manufacturer, to provide CCGA column attachment services to the RadHard community. Dr. Joseph Benedetto, CEO VPT Components, announced his new service at the AFRL RHET meeting held November 5-6, 2019 in Melbourne, FL. TopLine is a California based manufacturer of copper-wrapped Pb80/Sn20 solder columns for
Abaco Announces Flexible, Rugged 4-Channel Gigabit Ethernet Interface XMC to Help Minimize Size, Weight and Power (SWaP) • Offers choice of copper, fiber SFP connectivity • Broad range of operating system support • Simple, cost-effective upgrade for PMC676RC, PMC677RC users Abaco Systems announced the XMC477RC 4-Channel SFP Gigabit Ethernet Interface, further expanding the company’s range of network interface card options. The new XMC is distinguished by its support for four SFP (Small Form Factor Pluggable) modules for maximum flexibility, allowing customers to choose between 1000BASE-X optical or 1000BASE-T copper connectivity. When paired with an Abaco single board computer, the XMC477RC delivers a high degree of functional density that can reduce the number of cards in a chassis – eliminating, for example, the requirement for a dedicated Ethernet switch – that can help minimize the size, weight and 28
COTS Journal | December 2019
Ceramic Column Grid Array (CCGA) packages. “VPT Components is working towards DLA certification with a goal of reaching QML Q-level qualification in early in 2020,” stated Dr. Benedetto. Martin Hart, CEO of TopLine Corporation, remarks, “We are pleased to extend close technical cooperation and support to VPT Components, who serves as an independent contract assembler to provide column attachment services to HiRel aerospace companies. VPT has demonstrated excellent skills and workmanship with attaching TopLine made solder columns to FPGA devices.”
power (SWaP) of a subsystem. The XMC477RC also provides a simple and highly cost-effective form-, fit- and function compatible upgrade for existing users of the PMC676RC or PMC677RC who will benefit from a more configurable product with higher throughput. Fully rugged, with the ability to operate in temperatures between -40°C to +85°C, the XMC477RC is designed for naval/marine-, landand air platforms within defense and aerospace markets including fire control systems, airborne mission computers, radar systems and flight
control. It supports a high speed link to the host via an x4 PCIe™ connection, allowing all front I/O ports to run at full line rate. “Minimizing size, weight and power is be-
Dr. Benedetto added, “There is a need in the RadHard IC community for faster turn-times. VPT developed a CCGA process using TopLine copper wrapped columns with excellent attachment and pull-strength test results. We are working on automating the process to give customers quick turn services at competitive pricing.” TopLine www.TopLin.tv
coming increasingly important to our customers as military platforms become smaller while incorporating increasing functionality – and the XMC477RC can play a key role in helping them achieve that,” said Peter Thompson, Vice President, Product Management at Abaco. “The XMC477RC further broadens our already-extensive product offering, enhancing our ability to provide customers with a complete, single source solution.” Featuring an industry standard Intel® I350AM4 quad port Gigabit Ethernet controller that gives native support for enhanced virtualization features such as VMDq and up to eight virtual machines allocated per port, the XMC477RC is easy to set up and use. It is supported by native drivers for common operating systems including Microsoft Windows®, Linux® and LynxOS®. Abaco can additionally provide drivers for VxWorks® and Solaris™. Abaco Systems www.abaco.com
AdaCore Introduces Support for C++ Embedded Environments
GNAT Pro cross-environment handles standalone C++ and mixed-language embedded software development AdaCore announced that its GNAT Pro product line now supports the development of embedded software written in C++. Projects using C++, either mixed with Ada or standalone, can now benefit from the same high level of quality and support that GNAT Pro Ada customers receive. GNAT Pro C++ handles versions of the language standard up to and including C++17. It is targeted to VxWorks® 7 (ARM, PowerPC, and x86), Embedded Linux (ARM, PowerPC and x86), as well as VxWorks 6.9.x (PowerPC 32 bits), and is hosted on x86 GNU Linux. It is available on both the GNAT Pro Enterprise and GNAT Pro Assurance product lines.
“Embedded software teams have long been moving to multi-language development environments,” said Quentin Ochem, lead of Business Development at AdaCore. “Regardless of whether languages are intertwined in a single application, or used for separate projects, they still require the same high level of support. Our toolchains already support Ada and C for cross-development, as well as Ada, C, and C++ for native development. By adding an embedded component for C++, we can now offer our customers a truly comprehensive solution for a wide array of high integrity requirements.”
“Support for C++ is critical for modern development practices and next generation system development,” said Michel Chabroux, Senior Director of Product Management at Wind River. “Wind River is pleased to be collaborating with AdaCore to ensure the industry has access to best-in-class solutions for high integrity requirements, as well as provide our joint customers with a competitive advantage.” AdaCore www.adacore.com
Jessie Glockner, Director of Marketing and Communications, AdaCore. GNAT Pro’s build environment is based on the multilingual GCC code generation technology, to which AdaCore has contributed and among whose official maintainers are AdaCore personnel. GNAT Pro provides advanced interfacing capabilities with Ada and C++, such as binding generation to and from both languages, mixed-language exception propagation and catching, and cross-language inheritance and dispatching. The GNAT Pro Assurance edition provides unique benefits for pure-C++ developers, including long-term support, known problem analysis, and safety-critical fixes provided by the same experts who developed the product. It also improves multiple platform portability, providing consistent technology across native and embedded environments. COTS Journal | December 2019
COT’S PICKS OSS Introduces Industry’s First Solution Incorporating NVIDIA GPUs with NVLink and Marvell ThunderX2 Arm-based Servers OSS, Marvell and NVIDIA Demonstrate Ideal Solution for AI on the Fly® Applications One Stop Systems, Inc. announced availability of its new SCA8000 compute acceleration expansion platform for use with Marvell® ThunderX2® Arm® based servers. The SCA8000 includes eight NVIDIA V100 Tensor Core GPUs interconnected with NVIDIA NVLink providing 300GB/s of peer-to-peer GPU bandwidth. The SCA8000 enables up to 64GB/s of uplink bandwidth between the GPUs and Arm based servers using up to four PCI Express Gen3 x16 host interconnects. In July, NVIDIA announced it was making available to the Arm ecosystem its full stack of AI and HPC software that accelerates more than 600 HPC applications and all AI frameworks. The stack includes all NVIDIA CUDA-X AI and HPC libraries, GPU-accelerated AI frameworks and software development tools. The SCA8000 is the industry’s only NVIDIA NVLink based solution available that leverages this NVIDIA CUDA support for the Arm processor architecture. The resulting solution allows unparalleled scalability of multiGPU compute performance with power optimized, high-performance host processing ideal for AI on the Fly® edge applications. AI on the Fly edge appliances based on OSS technology bring datacenter AI capabilities to edge applications in military, aerospace, medical and autonomous driving. OSS’ flexible solutions provide the high bandwidth and low latency data transport needed to dramatically scale AI performance at the edge. The SCA8000 is a 3U 8-way SXM2 expansion chassis designed to be installed in a standard 19 inch rack. It supports eight passively cooled SXM2 V100 GPU modules and has an additional 4 PCIe x16 slots to support options, such as 100/200Gb Ethernet, Infiniband or other IO and data acquisition cards. It can connect to an Arm based host server with up to four PCI-SIG PCIe 30
COTS Journal | December 2019
Cable 3.0 compliant links at a distance of up to 100m. “This new compute acceleration expansion platform demonstrates how OSS continues to be the leader in delivering the latest in high performance technology for mission critical applications,” said Steve Cooper, CEO of OSS. “We have worked closely with NVIDIA and Marvell to develop this industry-first Arm solution with NVIDIA NVLink capabilities, leading the market in enabling Arm and CUDA solutions and supporting our AI on the Fly initiatives.” Paresh Kharya, director of product management of Accelerated Computing at NVIDIA, commented: “NVIDIA is working with the Arm ecosystem to provide the highperformance computing industry a new path to build extremely energy-efficient, AI-enabled supercomputers. The OSS expansion platform accelerated by NVIDIA GPUs with NVLink and NVIDIA’s CUDA-accelerated software provides the HPC community an easy to deploy Armbased GPU-accelerated solution.”
Gopal Hedge, vice president and general manager of Server Processor Business Unit of Marvell Semiconductor, said: “Our collaboration with NVIDIA and OSS marks continued advancement of the ecosystem for ThunderX2, the most widely supported Arm-based server processor. ThunderX2 with NVIDIA GPUs and OSS’s NVLink based platform showcases exciting new capabilities to meet the requirements for performance and scale of AI applications.” The Supercomputing 2019 demonstration includes a 2U server reference platform with two 32-core Arm-based Marvell ThunderX2 processors running at up to 2.5GHz in Turbo mode and supporting up to 4TB of system memory. The ThunderX2 processors provide industry-leading performance per power and memory bandwidth and high core count with hyper-threading. The server is connected to the SCA8000 populated with eight NVIDIA V100 Tensor Core GPUs interconnected with NVIDIA NVLink. ThunderX2 based servers are available from OEMs such as HPE, Bull-Atos, Cray, and GigaByte. One Stop Systems, Inc. www.onestopsystems.com
New COM Express Type 10 Carrier Hosts Four AcroPack® mPCIe I/O Modules for a Mix of Processing Functions on a Mini-ITX Platform GNAT Pro cross-environment handles standalone C++ and mixed-language embedded software development
Acromag’s newest carrier board interfaces four I/O module slots to an Intel® Atom® COM Express, CPU module on a Mini-ITX format for a variety of data acquisition and control applications. Acromag broadens their family of rugged AcroPack mini-PCIe I/O module carrier cards with the new Model ACEX4041 which also hosts a COM Express® Type 10 processor module. The Mini-ITX form factor board interfaces four plugin AcroPack I/O modules to an Intel Atom® E3950 (Apollo Lake) CPU module with provisions for both M.2 and SATA SSD data storage plus a variety of peripherals on-board. Ports for two RJ45 Gigabit Ethernet, two USB 3.0, two RS232, one Mini-DisplayPort, and audio connections are also included. The four AcroPack I/O slots provide a reliable PCIe mini card interface for a mix of analog, digital, serial, FPGA, avionics and other I/O functions. Four 68-pin VHDCI connectors securely route the field I/O to the AcroPack modules without any loose internal cables. AcroPack I/O modules support conduction-cooling and are rated for -40 to 85°C operation. Power is sourced from a 10-36V DC and/or ATX supply with redundant, auto-switch power capability. A development system package is offered to simplify evaluation and proof-of-concept engineering efforts in a lab.
Designed for defense, industrial, and mobile embedded computing systems, the ACEX4041 is ideal for small form factor (SFF) test & measurement, data acquisition & control, communication, avionics, simulation, and signal processing applications. It is well-suited for systems with size, weight, power and cost (SWaP-C) limitations. More than 25 AcroPack modules are available to perform A/D, D/A, discrete I/O, RS232/485, MILSTD-1553, CAN bus, Gigabit Ethernet, and other functions. Third-party mPCIe modules are supported for use in the AcroPack slots. The COM Express Mini Size Type 10 slot can also host Intel Core™ i7 processors and other compatible computer-on-module processors on the market. “This versatile SFF carrier card brings powerful modular embedded computing and flexible I/O signal processing capabilities to a standard Mini ITX platform that is easy to integrate within a custom system or mount in a rugged enclosure,” asserts Robert Greenfield, Acromag’s Business Development Manager. AcroPack mezzanine modules improve on the mini PCI Express architecture by adding a down-facing 100-pin connector that securely routes the I/O through a carrier card to its integrated external connectors thus eliminating internal cables and metalwork modifications. More than 25 models are available for data acquisition, signal processing, test & measurement, command/control, and network communication applications. Software tools support embedded applications running on Linux®, Windows®, or VxWorks® operating systems.
Acromag Inc www.acromag.com
COTS Journal | December 2019
COT’S PICKS Pentek Announces 6.4 GHz A/D and D/A Jade 3U VPX Module for Wideband Defense, Radar and Communication Applications • Utilizes state-of-art high-speed Texas Instruments A/Ds and D/As • Matching 6.4 GHz A/D and D/A sampling for wideband signals • 3U VPX with PCIe Gen.3x8, Optical I/O and RF I/O backplane options • Jade architecture uses Xilinx Kintex UltraScale FPGA • Navigator Design Suite supports Xilinx Graphical Vivado IP Integrator Pentek, Inc., introduced the newest member of the Jade™ family of high-performance data converter 3U VPX modules based on the Xilinx Kintex Ultrascale FPGA. The Model 54141A is a dual channel analog-to-digital and digital-to-analog converter with sample rates up to 6.4 GHz. Programmable DDCs (digital downconverters) and DUCs (digital upconverters) support connections to IF or RF signals.
For flexibility across different I/O requirements, the board can be optioned to support VITA 67.3C. This provides analog signal routing through the VPX backplane to replace front panel connectors for RF In/Out, Sample/Reference clocks and Gate/Trigger/ Sync/PPS signals. This option is often required in ruggedized deployments and for simplifying unit field replacements and upgrades. A/D Stage and Digital Downconverter The Model 54141A uses the Texas Instruments ADC12DJ3200 12-bit A/D converter with an input bandwidth of 6 GHz, which operates in singlechannel interleaved mode with a sampling rate of 6.4 GHz or in dual-channel mode with a sampling rate of 3.2 GHz. Digital Upconverter and D/A Stage A Texas Instruments DAC38RF82 D/A with
“The 54141A combines our popular 7114A XMC module with a 3U VPX carrier, yielding a powerful, forward-looking package for very wideband communications or radar applications that require advanced I/O resources,” said Rodger Hosking, vice-president of Pentek, Inc. Not only does the Model 54141A comply with the VITA 65.0 3U VPX specification, it also offers flexible analog and digital interface options for the VPX P2 backplane connector to meet systemspecific requirements. In addition to its PCIe Gen.3x8 interface on the VPX P1 connector, it is possible to add up to 8 more gigabit serial lanes connected directly to the FPGA for supporting user-installed protocols. The Model 54141A supports the emerging VITA 66.5 optical interconnect standard by providing four optical duplex lanes to a mating spring-loaded backplane connector. With the installation of a serial protocol like 10 or 40 Gigabit Ethernet in the FPGA, the interface enables high-bandwidth digital communications between boards or chassis independent of the PCIe interface. 32
COTS Journal | December 2019
DUC accepts a baseband real or complex data stream from the FPGA and delivers it to the interpolation, upconversion and dual D/A stages for output signals up to 4 GHz. The two 6.4 GHz 14-bit D/As pair well with the dual input channels while delivering more than twice the output performance of previous generations of Pentek products. Performance IP Cores The Model 54141A factory-installed functions include two A/D acquisition modules and a D/A waveform generation IP module. In addition, IP modules for DDR4 SDRAM memories, a controller for all data clocking and synchronization functions, a test signal generator and a PCIe Gen.3x8 interface complete the factory-installed functions. An optional VITA 49.2 data transport protocol IP module conveys digitized signal metadata for signal acquisition and
processing elements in communication, radar or storage systems. System integrators get to market with less time and risk, because the Model 54141A can provide a complete turnkey acquisition/generation solution, often without the need to develop FPGA IP. The Jade Architecture The Pentek Jade Architecture is based on the Xilinx Kintex UltraScale FPGA, which raises the digital signal processing (DSP) performance by over 50% with equally impressive reductions in cost, power dissipation and weight. As the central feature of the Jade Architecture, the FPGA has access to all data and control paths, enabling factory-installed functions including data multiplexing, channel selection, data packing, gating, triggering and memory control. A 5 GB bank of DDR4 SDRAM is available to the FPGA for custom applications. The x8 PCIe Gen 3 link can sustain 6.4 GHz data transfers to system memory. Eight additional gigabit serial lanes and LVDS general purpose I/O lines are available for custom solutions. Navigator Design Suite for Streamlined IP Development Pentek’s Navigator™ Design Suite consists of two components: Navigator FDK (FPGA Design Kit) for integrating custom IP into Pentek sourced designs and Navigator BSP (Board Support Package) for creating host applications. Users can work efficiently at the API level for software development and with an intuitive graphical interface for IP design. The tools reduce the development time and cost associated with complex designs. The Navigator BSP is available for Windows and Linux operating systems. Unlike others in the industry, Pentek still provides application support to customers at no cost. Pre-Configured SPARK System Ready for Immediate Use With a Pentek 8266 SPARK® PC, 8264 SPARK 6U VPX, or 8267 SPARK 3U VPX development system, work can begin immediately on applications. A SPARK system saves engineers time and expense associated with building and testing a development system and ensures optimum performance of Pentek boards. Pentek, Inc. www.pentek.com
Military Grade “NO-DRONE” CounterUAS Radar Detection System Released for Airport, Facility and Event Anti Drone Protection
In response to recent threats of rogue and errant drones launched against airports and other critical facilities, infrastructure and events, 34 North Drones has partnered with IDS North America to offer and install the NO-DRONE radar counter drone system for all government and civilian clients in need of protection. 34 North Drones - Military Grade NODRONE Counter Drone-UAS Radar System for protection of airports, critical infrastructure, facilities and events available for permanent installation or rental for events, rallies, concerts, sporting events or any high risk gathering 34 North Drones Airport NO-DRONE Countermeasures system for facility and airspace protection against drones The NO-DRONE radar detection system evolved from a proven and existing military battlefield solution originally tuned to recognize small objects such as incoming mortar, artillery and rocket fire with the ability to detect, locate and track both fixed wing and rotorcraft UAVs, Class I category (mini or small) and larger. The system provides full 360° coverage, day or night, as well as in adverse weather conditions with the ability to detect dozens and, depending on configuration, hundreds of targets simultaneously.
The system can be upgraded with an optional “slew to cue” EO/IR turret and an RF detector which can be fully integrated into the system to provide substantive enhancements to drone tracking and identification capabilities. The military, governments and certain federal agencies are authorized to use various optional drone mitigation technologies that can be integrated with the NO-DRONE system such as EMP, powerful multiband jamming, GPS spoofing or live fire systems to disable, redirect or destroy threats. Airports are known as extraordinarily difficult environments in which to integrate and operate electronic equipment without interference with critical aircraft guidance and navigation systems. The NO-DRONE counter drone radar system utilizes a suite of EMC/EMI simulation tools for predictive assessment of possible interference between various radiating and receiving elements around the airport which models any potential interference with and from any existing NAVAIDS and COMMS equipment. These specialized tools promote safety of operations and maximize radar coverage. Comprehensive on-site technical support and training is provided to ensure successful installation and seamless ongoing operations. The NO-DRONE system will also be available for rental with trained operators in a mobile platform for temporary use at facilities and events across the USA, and later, internationally for drone mitigation where permanent installation is not practical or needed. This includes VIP protection, outdoor concerts, sporting events, rallies and political events or any large gatherings of persons which are potentially high-risk environments. In partnership with local authorities, FBI and DHS, a protective bubble can be established around locations to mitigate the unauthorized entry by aerial threats. Extensive demonstrations and tests at the China Lake Naval Air Weapon Station and internationally have been conducted. The system has been installed at various airports and prisons and has been used on protective details for the Pope. The NO-DRONE system is now being released through 34 North Drones for all government and civilian protection applications. 34 North Drones www.34northdrones.com
COTS Journal | December 2019
COT’S PICKS Eurotech Announces the DynaCOR 4035 Rugged NAS HPEC System Powered by Intel® Xeon® D2100 CPUs, the new DynaCOR 40-35 delivers up to 123TB of scalable NVMe storage and 6GByte/s ultra-high speed network data transfer rates Eurotech, has added the DynaCOR 40-35 to its High Performance Embedded Computing (HPEC) product portfolio. The DynaCOR 40-35 is a liquidcooled, customizable NAS HPEC system designed for extreme storage capacity and high speed data transfer rates in data-intensive applications at the Edge. The DynaCOR 40-35 brings unprecedented network storage capabilities to harsh and mobile field environments: it can reach 123TB of raw capacity supporting up to 16x 7.68TB NVMe drives. Additionally, two 100 Gigabit Ethernet ports allow
Astronics Announces a New Family of Rugged COTS Avionics Control and Communication Devices The ‘NG’ Family is a New Generation of Ballard Boxes for Avionics Interfacing and Computing that are Secure, Right-Sized for the Application, and Ready to Deploy Astronics Corporation is launching a new family of flexible, compact, 64-bit avionics I/O computing platforms for problem-solving and tech insertion in areas such as mission computing, distributed control and avionics data conversion. Astronics’ NG Family of rugged Ballard Boxes provide three levels of avionics I/O capability for computing, controlling, and converting operations on military and civil platforms. (Graphic: Business Wire) Astronics’ NG Family of rugged Ballard Boxes provide three levels of avionics I/O capability for computing, controlling, and converting operations on military and civil platforms. (Graphic: Business Wire) The NG Family consists of three distinct series of commercial off-the-shelf (COTS) products with varying capability for use on fixed wing, UAV, rotocraft and ground vehicles. These highly-flexible, low SWaP (Size, Weight, and Power) systems include multi-protocol avionics I/O coupled with a high-throughput Ethernet backbone for inter-device communication and compatibility 34
COTS Journal | December 2019
for sustained data transfer rates of up to 6GBytes/ second. The system also supports advanced RAID configurations to provide extra performance and data protection. “With the DynaCOR 40-35, the Eurotech HPEC portfolio reaches a new level for rugged high-performance storage systems” commented Pierfrancesco Zuccato, Eurotech’s Senior Product Manager “This rugged HPEC NAS is an answer to the increasing demand for hyperconverged infrastructures in the field. Automotive applications, such as Level 5 Autonomous Driving, is only one example of use cases that require data-center class performance and capability in an operating environment that can be extremely harsh”. The DynaCOR family of high performance computing and storage systems are fanless and ventless for use in a broad range of harsh with a variety of avionics systems. The NG Family meets the demands of DoD open system business and technical strategies such as MOSA (Modular Open Systems Approach) and OMS (Open Mission Systems). The NG Family shares similar design features and each variant is optimized for different levels of deployment: • NG3 Series Avionics I/O Computers,available for order now, feature the highest level of capability and I/O, along with 64-bit processing and audio/video, for the most demanding applications. • NG2 Series Avionics I/O Controllers,available in 2020, feature the same processing power as the NG3, but in a smaller size with a corresponding reduction of I/O, when lower SWaP is critical. • NG1 Series Avionics I/O Converters, available in 2020, are extremely compact and optimized for converting avionics protocol data to Ethernet for interfacing to the NG3/NG2 or other avionics systems. “The NG Family builds on the success of our highly-popular and ground-breaking AB2000 and AB3000 avionics I/O computer products,” said Jon Neal, President of Astronics Ballard Technology. “The NG Family provides our large installed base of users with a direct upgrade path to 64-bit processing, robust security features and higher performance, extending the useful life of their programs for many years to come. The NG Family will also appeal to new users as flexible, problem-solving platforms for intelligence, surveillance, and reconnaissance (C4ISR) programs.”
environments and demanding Edge applications. The DynaCOR 40-35 compact design – only 17.7 x 19.2 x 49.5 cm - enables installation in app-lications where space is at a premium without requiring ventilation thanks to Eurotech’s advanced liquid cooling technology. The system is pre-certified for shock/vibration and other harsh environmental and electrical certifications in a broad range of mobile and vehicular applications. It can be simply integrated in traditional, hybrid and electric vehicles thanks to an automotive isolated power supply (12VDC and 48VDC). It can also be customized to meet application-specific certifications through Eurotech Professional Services. Eurotech www.eurotech.com
The NG3 Series is currently ready for customer evaluation for in-progress and upcoming programs. Key features include: • Latest Generation Intel Atom® E3950 64-bit quad core processor to run newer 64-bit operating systems that provide higher capability for today’s cybersecurity requirements. • Rugged, maximum-density, modular architecture provides flexible I/O combinations from the factory, including MIL-STD-1553, ARINC 429, 708, 717, discrete I/O, and Ethernet. These units arrive pre-configured, pre-validated and ready for quick deployment. • Cyber-security enablement consists of built-in tools to help customers build proprietary security solutions, including secure storage and operation, security SSD features and flexible write-protect capability. Two solid-state drive (SSD) media are user-removable and include sanitize discretes to enable users to declassify NG3 if necessary. • Easy system migration for current AB3000 customers looking for low-risk technical refresh and insertions. This drop-in, form-fit replacement shares the same programming interface, size, and mounting configuration, with many variants being pin-to-pin compatible. • Mini PCIe expansion slot allows I/O expansion while maintaining the ability to keep the same chassis envelope and size. This provides a low cost, low risk method to integrate additional I/O. Astronics Corporation www.astronics.com
Annapolis Micro Systems ......................................
29 ........................................ www.annapmicro.com
Avalex Technologies .............................................. 22
Behlman Electronics ............................................
Fairview Microwave ................................................ 27 ............................. www.fairviewmicrowave.com Kingston Technology ............................................. IBC ............................................. www.kingston.com MilesTex ................................................................
33 ............................................. www.milestexs.com
Neonode ................................................................ 14/23 . .............................................. www.neonode.com New Wave DV .........................................................
21 ......................................... www.ewwavedve.com
OSS ........................................................................ IFC ................................. www.onestopsystems.com Pasternack ............................................................
25 ......................................... www.pasternack.com
31 .............................................. www.percepic.com
5 ................................................ www.pentek.com
PICO Electronics, Inc .............................................
11 ................................... www.picoelectronics.com
4 .............................................. www.sealevel.com
Vicor Cororation ..................................................... 1 9/IBC ............. www.vicorpower.com/defense-aero.com COTS Journal (ISSN#1526-4653) is published monthly at; 3180 Sitio Sendero, Carlsbad, CA. 92009. Periodicals Class postage paid at San Clemente and additional mailing offices. POSTMASTER: Send address changes to COTS Journal, 3180 Sitio Sendero, Carlsbad, CA. 92009.
Ultra-High Voltage Bus Converter Provides 98% Efficiency This unique K=1/16 fixed ratio bus converter sets the industry benchmark for efficiency and power density. The thermally adept VIA package simplifies customer cooling approaches in addition to providing integrated PMBus control, EMI filtering, and transient protection. These strong abilities make it ideally suited to military applications. Evaluate it today!