Compact Modular RFSoC Solutions

The boundary-pushing team at Annapolis Micro Systems has taken a pioneering approach to delivering powerful RFSoC functionality. Their WILD FMC+ RFSoC Cards are compact modular solutions that operate standalone. If your application is very processing-intensive they can be paired with a WILDSTAR 3U or 6U VPX or PCIe Baseboard.

The concept behind Annapolis' RFSoC Cards is simple: to provide full RFSoC functionality in a small, flexible package that has solved the most demanding circuitry, signal integrity, and packaging challenges. No small feat.

Xilinx’s Zynq® UltraScale+™ RF System-on-Chip (RFSoC) combines multiple elements that were formerly discrete operations. In defense and military applications like C4ISR and radar, Size, Weight, Power and Cost (SWaP-C) considerations are crucial. Applications like UAVs, backpacks, handheld devices, and custom-integrated applications are increasingly subject to smaller spaces and power sources.

Following an intensive design and manufacturing process Annapolis' RFSoC cards provide turnkey solutions ready for immediate deployment within any standard OpenVPX or PCIe system, or within Annapolis’ proven 100Gb Ethernet WILD100™ EcoSystem. They are also compatible with SOSA™-aligned Boards and Systems.

If you require high-performance RFSoC-enabled data digitisation, processing or storage in a compact, rugged environment, these WILD RFSoC Cards are for you. Contact us to have an engineer recommend the best approach for your application.

©2020 Annapolis Micro Systems, Inc.

SOSA - Sensor Open System Architecture for C4ISR Applications

If you haven’t heard about SOSA, you’re probably not the only one, but it’s big news for Command, Control, Communications, Computers, Intelligence, Surveillance and Reconnaissance (C4ISR) systems.

SOSA, or the Sensor Open System Architecture, is a collaboration between government and industry to establish guidelines for C4ISR systems. The United States is currently spearheading SOSA development, but the intention is to encourage the use of SOSA in both government and commercial markets across the world.

Using OpenVPX as its framework, SOSA will ensure improved SWaP-C optimisation, extensive reconfiguration options and rapid technology upgrades within a wide range of applications. The fundamental objective is to give designers maximum flexibility in the initial selection of subsystems that provide sensor data collection, processing, exploitation, communication, and related functions over the full life cycle of the C4ISR system.

Annapolis Micro Systems was an early adopter of the modular open standards model for COTS product development, and is a dedicated SOSA contributor and member.


The 3XBM from Annapolis is SOSA-aligned and VITA 65-compliant.

These boards provide either one Kintex® UltraScale™ XCKU115 or Virtex® UltraScale+™ XCVU5P/XCVU7P FPGA with up to 10 GB of DDR4 DRAM, providing up to 40 GB/s of DRAM bandwidth. WM3XBM - Find Out More

A WFMC+ mezzanine I/O site with stacking support is included, alongside an on-board Zynq Quad ARM CPU and 1Gb Ethernet Switch. High-bandwidth backplane connectivity is enabled by MULTIGIG RT3 interconnects, which deliver 100Gb per Fat Pipe. The new high-density RT3 boosts VPX backplane speeds to a remarkable 25 Gbaud, without sacrificing signal integrity.

For more information please contact the Sarsen Technology Engineering Team who can answer questions about SOSA, Annapolis and any of our other high performance embedded hardware.

Brand New Small Form Factor (SFF) System Features 10 Gb/s Optical Links for High-Bandwidth Embedded Computing

The latest 3U VPX embedded system from Extreme Engineering Solutions (X-ES) sets a new standard for high-bandwidth data acquisition and signal processing in harsh environments.

With multiple high-speed optical links, an Intel® Xeon® D-1500 processor, and a Xilinx Kintex® UltraScale™ FPGA, the XPand6215 offers uncompromising performance and reliability in a SWaP-optimised two-slot system.

High-Speed, Fibre Optic Communications
The XPand6215 is a commercial-off-the-shelf (COTS) rugged chassis hosting two conduction-cooled 3U VPX modules. One slot contains the XPedite2570, a reconfigurable fibre-optic I/O module based on the Xilinx Kintex® UltraScale™ family of FPGAs.

The XPedite2570 supplies 12 protocol-independent optical transceivers operating at up to 10.3125 Gb/s, along with 8 GB of DDR4 ECC SDRAM in two channels capable of up to 38 GB/s aggregate bandwidth. The transceivers utilise 50/125 μm multi-mode fibre with MT connectors and are qualified over the full -40°C to +85°C industrial temperature range for reliable performance in extreme environments.

These optical links are paired with a user-programmable Xilinx Kintex® UltraScale™ FPGA, which provides a blend of performance and cost-effectiveness that is ideal for packet processing and DSP-intensive processing in embedded applications.

Unparalleled, Multi-Core Processing Power
The other slot of the XPand6215 contains the XPedite7670, a 3U VPX Single Board Computer based on the Intel® Xeon® D-1500 family of processors (formerly Broadwell-DE). Available with up to 16 Xeon®-class cores in a single, power-efficient System-on-Chip (SoC) package, this processor puts the XPand6215 in the upper echelon of embedded system performance. Xeon® D processors with up to 12 cores offer native extended temperature support with an operating temperature range of -40⁰C to +85⁰C.

The XPedite7670 also provides up to 16 GB of DDR4-2133 ECC SDRAM, while an XPort3305 10 Gigabit Ethernet XMC module installed on the XPedite7670 provides a rugged 10GBASE-SR Ethernet port.

Industry-Leading Functional Density in a Small Form Factor (SFF) Enclosure
The XPand6215 makes the most of every pound and square inch, packing maximum functionality into a streamlined two-slot design. Its integrated 28 VDC power supply eliminates the need for an additional power supply slot. It also includes a memory bay for a removable 2.5 in. solid-state drive (SSD), adding the option for up to 1 TB of storage. Fully loaded, the system weighs approximately 4kg, making it ideal for cutting-edge military and industrial applications that require small, lightweight solutions.

Qualified to the MIL-STD-461E/F, MIL-STD-810, and MIL-STD-704F standards, the XPand6215 has passed a battery of tests to ensure reliable operation in even the most challenging environments.

The XPand6200 Series of rugged, embedded systems are available in a diverse range of configurations, making them uniquely-suited to a variety of processing-intensive applications.

Please contact Sarsen Technology to discuss your requirements.

Extreme Engineering Releases 3U VPX Xilinx Kintex UltraScale FPGA Module

Extreme Engineering Solutions (X-ES) introduces the XPedite2570, a rugged FPGA processing module with a high-speed optical front-end interface. Equipped with an integrated Xilinx Kintex® UltraScale™ FPGA, the XPedite2570 optimises both cost and performance for high-bandwidth embedded computing applications.

XPedite2570 Datasheet

Xilinx Kintex® UltraScale™ FPGA Delivers Great Performance Without Great Cost
The XPedite2570 is a high-performance, conduction-cooled 3U VPX FPGA processing module based on the Xilinx Kintex® UltraScale™ XCKU115 FPGA.

The Xilinx Kintex® UltraScale™ family of FPGAs provides the best price/performance/watt at 20nm, as well as the highest signal processing bandwidth for a mid-range device.

Ideal for programs that need to optimise both cost and capability, the Kintex® UltraScale™ XCKU115 packs over 1.4 M system logic cells, 5,520 DSP slices, and 75.9 Mb of block RAM into a cost-effective package. It also provides a straightforward upgrade path for legacy systems, offering up to 40% power savings compared to the previous generation of FPGAs.


Optimised for High-Bandwidth Applications
The XPedite2570 is a reconfigurable FPGA resource designed to meet the demands of high-bandwidth applications such as packet processing, signal processing, and DSP-intensive applications. It features 12 rugged, protocol-independent fiber optic transceivers operating at up to 10.3125 Gb/s, along with 8 GB of DDR4 ECC SDRAM in two independent channels capable of up to 38 GB/s aggregate bandwidth. The XPedite2570 has several options for high-performance backplane I/O, including a x8 Gen3 PCI Express interface, dual GTH transceivers with a maximum data rate of 16.375 Gb/s, and up to 44 LVDS transceivers for user I/O.

X-ES’ board support package for the XPedite2570 includes an FPGA Development Kit for Xilinx Vivado complete with a reference design, IP Integrator IP blocks, HDL, test benches, and Linux drivers, so developers can jump right into application development without having to work through hardware integration issues.

Build a Rugged COTS System with Compatible X-ES Products
For lower development costs and faster deployment, pair the XPedite2570 with a 3U VPX single board computer (SBC) and install them together in the XPand6215, a Commercial-Off-the-Shelf (COTS) rugged system designed to meet the rigorous standards of MIL-STD-704F and MIL-STD-461. The XPand6215 can support several embedded single board computers from the XPedite767x product line, such as the XPedite7670. This family of high-performance SBCs is based on the Intel® Xeon® D processor, which provides up to 16 Xeon®-class cores in a single, power-efficient System-on-Chip (SoC) package.

More about Extreme Engineering Solutions (X-ES)

BittWare Announces SmartNIC Shell for Building FPGA-powered 100G NICs

SmartNIC Shell Supports DPDK, Xilinx SDNet, P4 Programming, User Customisations, and Timestamping

BittWare has announced SmartNIC Shell, a suite of IP modules for building 100G network interface controllers (NICs) using FPGAs for hardware packet processing.

Users can fully customise Shell components, including a Match/Action pipeline with features including Xilinx SDNet support and the P4 network programming language. For host application interaction the Arkville DPDK IP is integrated for kernel bypass. SmartNIC Shell allows teams to avoid the time-consuming process of building core NIC functions into an FPGA, allowing them to focus resources on their own application.

Read more - Full Press Release

“We built the SmartNIC Shell because we noticed that many of our customers spend half their development time creating a networking shell,” said Craig Lund, Vice-President Network Products, BittWare. “All of that development time should be going into their own product’s unique value instead.”


SmartNIC Shell Top Features

Quickly Build 100G NICs	Focus your attention on your unique application, instead of re-inventing a NIC.
Match/Action Pipeline	Standardised kernel bypass for host interaction over PCIe. SmartNIC Shell provides DPDK offload to interact with host applications.
TimeServo Timestamping	Precision time stamping including 1588-compatible clock adjustments. Uses TimeServo IP from Atomic Rules
Xilinx UltraScale+ FPGA	Large, powerful FPGAs with ample room for user IP. Selection of BittWare boards including traditional low-profile NIC size (XUPPL4) to 3/4-length boards with additional logic and memory options (XUPP3R).

For more information please contact the Sarsen FPGA team on +44 1672 511166 or send us an email - info@sarsen.net

BittWare Acquires nCk Research and Forms Network Products Group

BittWare has announced the acquisition of nCk Research, a network solutions provider with a portfolio of FPGA intellectual property (IP) designed for BittWare’s PCIe boards.

nCk’s staff joined BittWare as employees under a newly-formed Network Products group. Craig Lund, formerly with CSPI-Myricom, will lead BittWare Network Products as V.P. and General Manager.

www.bittware.net


The new group’s first product, based on nCk’s groundbreaking IP, was recently announced at the RSA Conference in San Francisco. The StreamSleuth is a network packet processing appliance, providing Xilinx UltraScale+ FPGA-based accelerated monitoring, filtering, and routing at 100 GbE line rate.

“Building on the recent success of our horizontally-focused FPGA Platform business and the acquisition of nCk, it is huge step for BittWare to launch a vertically-focused network solutions group,” said Jeff Milrod, President & CEO of BittWare. “The addition of Craig’s leadership and extensive history of successfully driving new technology into emerging markets makes me confident that we will be able to quickly combine our world-class FPGA platforms with our innovative IP to deliver compelling network products to a market desperate for 100 GbE solutions.”

nCk Research had initially chosen BittWare’s FPGA solutions for prototyping their 100 Gbit/s IP. The unique approach and features of the IP so impressed BittWare that a collaboration was formed to accelerate the product development time. This grew into a need for closer access to BittWare’s software and hardware engineering teams, resulting in the acquisition in the second half of 2016.

Craig Lund joined BittWare in January, 2017 as V.P. and General Manager of the newly-formed Network Products group. Craig’s extensive knowledge of the capabilities—and limitations—of today’s high-speed network hardware and software plus his decades of experience leading teams in product development ensures the products BittWare creates are a match for the real-world needs of sophisticated, high-end network users.

Intel® Xeon® D 3U VPX Single Board Computers with Integrated Security FPGAs Now Available

September 2016

Extreme Engineering Solutions (X-ES) has announced two new 3U VPX embedded single board computers based on the Intel® Xeon® D-1500 processor family. These powerful boards support up to 16 Xeon®-class cores in a single, power-efficient System-on-Chip (SoC) package with native extended temperature support on four, eight, and twelve core-count SKUs.

XPedite7674

The XPedite7674 is an Intel® Xeon® D-1500 processor-based 3U VPX single board computer that can support up to 16 core-count SKUs with native extended temperature support on up to 12 core-count SKUs.

With an integrated, user-configurable, Xilinx Kintex® UltraScale™ FPGA module delivering enhanced performance and security for a wide range of embedded computing applications, the board supports a variety of I/O to the VPX connectors, including dual 10 Gigabit Ethernet (10GBASE-KR), SATA port capable of 6 Gb/s, USB 2.0, and RS-232/422/485 serial ports.

The FPGA also provides GPIO, 1000BASE-X Gigabit Ethernet, and High-Speed Serial ports to the VPX connectors. The XMC site supports a x8 PCI Express Gen3-capable port and a SATA port, as well as XMC P16 I/O, mapping P1w9-X12d per VITA 46.9.

XPedite7676

The XPedite7676 is a secure, Intel® Xeon® D-1500 processor-based 3U VPX single board computer
that can support up to 16 core-count SKUs with native extended temperature support on up to 12 core-count SKUs.

The XPedite7676 integrates SecureCOTS™ technology with a Microsemi SmartFusion®2 security System-on-Chip (SoC) for hosting custom functions to protect data from being modified or observed and provides an ideal solution when stringent security capabilities are required.

Dual 10GBASE-KR 10 Gigabit Ethernet, 10/100/1000BASE-T Gigabit Ethernet, USB 2.0, SATA ports capable of 6 Gb/s, RS-232/422/485 serial, dual x4 PCI Express Gen3, and GPIO are all routed to the backplane VPX connectors.



Both the XPedite7674 and the XPedite7676 support the Wind River VxWorks Board Support Package (BSP), coreboot firmware powered by Intel® FSP to enable rapid and reliable sub-2.5 second boot times, and the X-ES Enterprise Linux (XEL) BSP.

Join us at the next FPGA Network meet-up on Tuesday 11th October in Swindon!

FPGA experts who care about tools and methodologies can’t afford to miss the next meeting of the NMI FPGA Network... You will find this meeting especially compelling if you are an architect, designer, verification specialist, project manager or engineering manager.

We know how important design tools and methodologies are in creating better FPGA designs and FPGA systems-based, this is why we are excited to be sponsoring this event and would be delighted to see you there too!

On the day we will contribute to the debate with a discussion on ‘FPGA Platform Development Kit Enables Fast TTM’ delivered by Chad Hamilton, VP of Intellectual Property (IP), Software and Support at BittWare.

Chad will join the discussion along with FPGA specialists from other Industry leaders.

Confirmed Speakers 
Synopsys: Flexible debug and visibility techniques to enhance all FPGA design and deployment cycles
BittWare: FPGA platform development kit enables fast TTM
Mentor Graphics: Staying competitive by evolving your FPGAverification methodologies
Bitvis AS: A game changer for VHDL verification: advanced VHDL verification – made simple for anyone
Telexsus: FPGA real-time debug with vastly increased operational capture time – live demo
Cadence: 100M gate designs in FPGAs – fact or fiction?
Altera now part of Intel: Zen and the art of high-speed design
Xilinx: Vivado HLx Design Methodology
ITDev: Static code analysis using Blue Pearl software

See full details (pdf)

We are also very pleased to have Manuele Papais from DAVE Embedded Systems with us, showcasing their Xilinx Zynq based hardware and talking to engineers about Asymmetric Multi Processing on Cortex-A CPU: system integrity and reliability on Linux and RTOS cores.

Find out more and book your place here.

We look forward to seeing you at the FPGA Network Event ‘Design Tools and Methodologies’!

BittWare Announces OpenPOWER CAPI DevKit for Xilinx FPGA-Enabled Accelerator Cards

FPGA platform supplier provides CAPI enabled FPGA accelerator cards for POWER8 processors.

April 2016

BittWare, an industry-leading board supplier for over 25 years, announced today an OpenPOWER CAPI Developer’s Kit for its Xilinx FPGA-enabled accelerator cards. Providing a fast, efficient way to connect the Xilinx All Programmable FPGA to a CAPI-enabled IBM POWER8 system, the developer’s kit includes the FPGA accelerator card, IBM Power Service Layer (PSL) IP to provide the connection to the POWER8 chip, CAPI host support library, and an example CAPI design.

The FPGA accelerator card included in the kit is BittWare’s XUSP3S, which is a ¾-length PCIe board featuring the Xilinx Virtex® UltraScale™ VU095, four QSFPs for 4x 100GbE, and flexible memory configurations with up to 64 GBytes of memory and support for Hybrid Memory Cube (HMC).

Most FPGA accelerators use PCIe to connect to the host processor unit. IBM has developed CAPI (Coherent Accelerator Processor Interface) as an alternative that takes some of the overhead and complexity out of the I/O subsystem by providing a direct interface between the host and the FPGA. Developers can run their specific FPGA solutions as a peer to the IBM POWER8 cores, enabling higher system performance with less programming complexity.

“BittWare is pleased to work with Xilinx and IBM to offer CAPI enabled FPGA acceleration platforms for POWER processors,” said Ron Huizen, VP of Systems and Solutions at BittWare. “Our XUSP3S board is a compelling option for data center applications, and with the addition of CAPI will now allow our customers to develop their applications even more quickly and efficiently.”

“CAPI on POWER8 systems provides a high-performance solution for implementing customer-specific, computation-heavy algorithms on an FPGA,” said Sumit Gupta, VP, High Performance Computing & Data Analytics at IBM. “The IBM solution enables higher system performance with a much smaller programming investment, allowing hybrid computing to be successful across a much broader range of applications. BittWare’s CAPI support brings this innovation to an even wider range of customers.”

“The demand for FPGA acceleration in data centers is increasing, and with that the need for innovative, high performance accelerator solutions,” said Curtis Pulley, data center business manager at Xilinx. “The CAPI-enabled Xilinx FPGA-based offerings from BittWare are a great addition to the OpenPOWER ecosystem and extend our efforts to deliver industry-leading solutions for IBM POWER systems.”

Availability
BittWare’s OpenPOWER CAPI Developer’s Kit will be available in early Q2 2016. Contact Sarsen Technology for details.



About OpenPOWER

BittWare is a member of the OpenPOWER Foundation, an open development community based on the POWER microprocessor architecture. The Foundation includes a growing roster of technology organizations working collaboratively to build advanced server, networking, storage, and acceleration technology as well as industry-leading open source software aimed at delivering more choice, control, and flexibility to developers of next-generation, hyperscale, and cloud data centers. The group makes POWER hardware and software available to open development for the first time, as well as making POWER intellectual property licensable to others, greatly expanding the ecosystem of innovators on the platform.

To learn more about OpenPOWER and to view the complete list of current members, go to www.openpowerfoundation.org.

About BittWare, Inc.

For over 25 years, BittWare has designed and deployed high-end network processing, signal processing, and high performance computing platforms that significantly reduce technology risk and time-to revenue for our OEM customers. Our products utilize the latest FPGA technologies on industry-standard COTS form factors, including PCIe, and are available either as boards or in systems. We also provide a great deal of value-add with IP, software, support, and services. When customer requirements make it difficult to use industry-standard boards, BittWare can provide modified solutions and/or licensed designs. For more information on BittWare and its innovative FPGA platform solutions, visit www.bittware.com/xilinx.

Learn more: Find your solution... Built on BittWare.

BittWare and nCk Research Collaborate to Provide an FPGA-based 100G Network Packet Broker with PCAP Filtering

As one of the market leading manufacturers in their field, new announcements from BittWare are always exciting...

The latest from the BittWare team in Concord, NH, is a collaboration with nCk Research - an innovative supplier of FPGA IP. This dream team have worked together to come up with a 100G network packet broker with PCAP filtering, which is implemented on BittWare's new PCIe board featuring Xilinx's UltraScale Virtex FPGA.

BittWare XUSP3S - Xilinx UltraScale 3/4-Length PCIe Board with Quad QSFP, DDR4, QDR-IV, and QDR-II+

PCAP processing expressions are used by many packet broker systems, but the point-to-point fabric from nCk Research is the first in the industry to optimise user inputs before loading resource settings into registers in the FPGA. This optimisation allows more rules to be packed into a smaller space and increases packet processing efficiency. The packet broker IP also provides configurable support for RMON1, line rate policies, aggregation, decimation, and de-duplication.

The XUSP3S is backward compatible with 10G/40G, and with nCk's IP supports up to 4x 100G ports and/or up to 16x 10/25G ports on each board. It can also be configured to implement any combination in between, allowing users to deploy with some 10G today and easily upgrade to all 100G in the field.


"We're pleased to be able to work with nCk to add innovative and comprehensive application IP to our already well established FPGA platforms," said Ron Huizen, Vice President of Systems & Solutions at BittWare.
"Their advanced filter IP component lets programmers use standard PCAP syntax to filter packets at full 100G line rate, and they can modify the filters on the fly- without having to reconfigure the FPGA. This, combined with nCk's other packet processing IP running on the most advanced FPGAs in the world, creates a compelling solution."


"FPGAs are the ideal choice for 100G at full line rate, and we're pleased to be able to partner with BittWare to build on their best-in-class FPGA platforms," said Matthew Chabot, President of nCk Research. "Their XUSP3S UltraScale board with support for quad 100G is a perfect fit."

Learn more:
10/25/40/100G Ethernet Filter IP Core for FPGA 
10/25/40/100G Packet Broker with PCAP Filtering 



Terabox
Up to 8 XUSP3S application cards can be installed in a BittWare TeraBox 4U rackmount system to handle up to 32x 100G ports and/or up to 128x 10/25G ports, allowing users to build virtually any packet processing solution with nCk network IP libraries; coupled with an optional PCIe DMA or NIC interface, it becomes possibly the most feature rich packet processing appliance in the industry.

Packet Broker with PCAP Filtering IP & Appliance Pricing and Availability
This configuration is available for purchase now, please contact Sarsen Technology for more information.