ALPHA DATA’S LATEST FPGA ACCELERATOR BOARD IS FIRST TO OFFER HBM-ENABLED FPGAs IN SERVER-FRIENDLY, LOW PROFILE FORM FACTOR

Alpha Data’s ADM-PCIE-9H3 accelerator board, with Xilinx VU33P Virtex® Ultrascale+™ FPGA, offers 200Gb/s front IO bandwidth, PCIe Gen 4 capability, 460GB/s on-die memory, and SlimSAS connector for 200Gb/s low latency rear IO (OpenCAPI compliant)

12th December 2018: Golden, Colorado & Edinburgh, Scotland

The ADM-PCIE-9H3 Image

Alpha Data, a leading supplier of high-performance FPGA solutions, announces the release of the ADM-PCIE-9H3, the latest in a family of products for accelerating Datacenter and HPC workloads. The ADM-PCIE-9H3 is a low-profile (half-length, half-height), PCI Express® add-in card featuring an Ultrascale+™ VU33P FPGA, which includes 8GB on substrate High Bandwidth Memory (HBM Gen2). This platform provides exceptional memory Read/Write performance at 460GB/s while reducing the overall power consumption of the board by eliminating the need for external SDRAM.

A key feature of the ADM-PCIE-9H3 is the low-profile form factor, which allows for integration into even the smallest of server chassis. Network or Storage-Attached applications will benefit from the flexibility of the front IO, which can be adapted to different physical interfaces. The 8-lane QSFP-DD cage supports 2x100GbE, 8x32G Fibre Channel or other line-rate and protocol combinations.

“The integration of FPGA and HBM allows memory-bound applications to fully benefit from the power-efficient performance of FPGA acceleration. The Alpha Data ‘9H3 couples 200G IO to the compute capacity of a Xilinx HBM FPGA, and packages it in a low-profile card. This flexible and high-performance combination will enable low-power acceleration at line rates in the next generation of Datacenter deployments.”

David Miller, Managing Director

The ADM-PCIE-9H3 includes on-board system monitoring which can provide real-time temperature, voltage and current measurements independent of the FPGA. It also controls programmable clocks, logs events and elapsed time and allows remote recovery of the FPGA configuration flash memory. The system monitor can be accessed directly through the USB interface via the front panel, a serial connection to the FPGA or through the SMBus interface on the card's PCI Express® edge connector. IPMI can also be used to communicate with the system monitor, allowing the ADM-PCIE-9H3 to be remotely managed and deployed with confidence.


For More Details

Further product details and ordering options for Alpha Data’s FPGA accelerator cards please visit.

https://www.alpha-data.com/ph3


About Alpha Data

Established in 1993, Alpha Data is a world leader in Xilinx FPGA based plug-in acceleration boards for Data Center, high-performance computing, and rugged embedded computing applications. Alpha Data’s high-reliability hardware platforms are ideal for development as well as full-scale production deployment - shaping the future in video processing, machine learning, and network acceleration.




Data Space 2019 - Glasgow, Scotland

24th Jan-25th Jan 2019

Data Space 2019 Logo

Alpha Data is exhibiting at Data Space 2019.

Contact: Alpha Data for further information.

Conference Details available on the Data Space 2019 website.

Embedded World 2019 - Nuremberg, Germany

26th Feb-28th Feb 2019

Embedded World 2019 Logo

Alpha Data is exhibiting at Embedded World 2019 (Hall 3, 3-100).

Contact: Alpha Data for further information.

Conference Details available on the Embedded World 2019 website.

23 Oct 2018

ALPHA DATA ANNOUNCES 12X100G NETWORK ACCELERATOR BOARD, FEATURING SAMTEC TWINAX FLYOVER™ SYSTEMS AND XILINX Ultrascale+™ FPGA

20 Jun 2018

WHITE PAPER: BREAKING MEMORY BANDWIDTH BARRIERS USING HIGH BANDWIDTH MEMORY FPGA

17 Apr 2018

ALPHA DATA RELEASE ADM-PCIE-9H7 DATA CENTER BOARD WITH XILINX VIRTEX Ultrascale+™ HBM FPGA

29 Mar 2018

OPENPOWER CAPI SNAP TEAM ADDS SUPPORT FOR THE ADM-PCIE-8K5

27 Feb 2018

ALPHA DATA RELEASES MPSOC EMBEDDED SYSTEM FPGA BOARD AT EMBEDDED WORLD

13 Nov 2017

OPENCAPI AT SUPERCOMPUTING 2017 (SC17)




©2019 Alpha Data Parallel Systems - All rights reserved