ULL Algorithmic Trading
From complex algorithmic trading to market-making services
Overview
The AMD Alveo™ UL3422 accelerator offers ultra-low latency (ULL) trading in a slim form factor for cost-effective deployment.
Purpose-Built for ULL
Less than 3 ns transceiver latency2 for high performance, deterministic trade execution.
Slim Form Factor
Cost-effective deployment for broad market adoption and deployment in exchanges world-wide.
Ease of Development
Fast path to trade through reference designs and ecosystem solutions.
Key Features
Purpose-Built for Ultra-Low Latency (ULL) Trade Execution
Powered by Purpose-Built FPGA
The AMD Alveo™ UL3422 is the world’s fastest trading accelerator providing superior ultra-low latency1 giving traders the edge they need for faster decision-making. The card is powered by the AMD Virtex™ UltraScale+™ VU2P FPGA built for electronic trading and features a breakthrough transceiver architecture to achieve less than 3 ns latency for world-class trade execution.2
Slim Form Factor for Cost-Effective Deployment in Diverse Servers in Any Exchange
The AMD Alveo UL3422 accelerator comes in a slim form factor for broad deployment in diverse server configurations, including servers from Hypertec for immediate deployment. By using purpose-built HFT infrastructure, trading firms can cost-effectively optimize rack space co-located at market exchanges4.
Ease of Development & Fast Path to Trade
FPGA Design Tools and Ecosystem Solutions
Featuring 780K LUTs of FPGA fabric and 1,680 DSP slices of compute, the Alveo UL3422 accelerator card is built to accelerate custom trading algorithms in hardware, where traders can tailor their design to evolving trading algorithms and new trade regulations. The accelerator is supported by traditional RTL development flows using the Vivado™ Design Suite.
Special licensing is required to enable the targeted Virtex UltraScale+ device. Developers can request access to the Alveo UL3422 Secure Site for licensing and access to additional technical documentation. Reference designs to evaluate latency and performance, as well as test other features of the card, are available at the GitHub Repository.
Additionally, AMD provides the ability to evaluate performance with the open source PyTorch-based framework (FINN) for developers of low latency, AI-enabled trading strategies. The card is also integrated with partner solution ecosystem partner solutions such as Xelera Silva and Exegy nxFramework for rapid implementation of trading algorithms.
Fintech Applications
Competitive Advantage in Capital Markets
Proprietary trading firms, hedge funds, market makers, brokerages, and data vendors can deploy the Alveo UL3422 accelerator for ULL algorithmic trading, offering world-record performance1, pre-trade risk management, market data delivery, and more. The convergence of hardware acceleration, FPGA flexibility, and low latency networking ensures high performance and determinism across a breadth of use cases.
Get Started
Get started today with the Alveo UL3422 accelerator card. Available from AMD and authorized distributors.
Alveo UL3422 Accelerator Card
The AMD Alveo™ UL3422 is an ultra-low latency accelerator in a slim form-factor for cost-effective server deployment¹ in exchanges world-wide. It is powered by an AMD Virtex UltraScale+ FPGA purpose-built for electronic trading. The FPGA features a breakthrough transceiver architecture to achieve less than 3 ns latency2 for world-record trade execution1, delivering up to 7X lower latency vs. previous AMD FPGA technology3.
Resources
Alveo UL3422 Accelerator Product Brief
The Alveo UL3422 accelerator card combines ultra-low latency networking with adaptable hardware to accelerate trading strategies at nanosecond speed.
Contact Us
Sign up to receive the latest product news or get in touch with an AMD sales representative.
Footnotes
- The 2024 AMD world record for latency is based on 3rd party testing commissioned by AMD and Exegy, by Strategic Technology Analysis Center, LLC (STAC®) in April 2024, using the STAC-T0 benchmark to test the AMD Alveo UL3524 accelerator card powered by the AMD Virtex Ultrascale+ VU2P FPGA, running on the Exegy nxFramework and Exegy nxTCP-UDP-10g-ULL IP Core, in a Dell PowerEdge R7525 server with AMD EPYC 7313 processors. See https://stacresearch.com/news/AMD240422 for the full STAC report. AMD holds the previous world record for latency (2020): https://www.stacresearch.com/news/XLX200514. Stated results for the Alveo UL3524 accelerator have been extrapolated to the AMD Alveo UL3422 card, based on identical silicon and product features. (ALV-20).
- Testing conducted by AMD as of 8/16/23, using Vivado™ Design Suite 2023.1 running on Vivado Lab (Hardware Manager) 2023.1, and the GTF Latency Benchmark Design on the Alveo UL3524 accelerator card, configured to enable GTF transceivers in internal near-end loopback mode. Benchmark was run 1,000 times with 250 frames per test, where PCS logic (physical medium attachment) of the transceiver passes data ‘as-is’ to FPGA fabric to measure latency. Stated latency result based on GTF transceiver “RAW Mode.” Configuration differences or variance in testing protocol will yield different results. Actual results will vary (ALV-10).
- Based on a simulated comparison by AMD using the Synopsis VCS 2019.06-SP2 ultra-low latency Virtex UltraScale+ VU2P GTF transceivers and GTY transceivers in February 2022 (ALV-15).
- Cost reduction of the AMD Alveo UL3422 accelerator, as compared to the previous generation Alveo UL3534 accelerator. Pricing is based on MSRP as of September 2024 (ALV-21).
Footnotes
- The 2024 AMD world record for latency is based on 3rd party testing commissioned by AMD and Exegy, by Strategic Technology Analysis Center, LLC (STAC®) in April 2024, using the STAC-T0 benchmark to test the AMD Alveo UL3524 accelerator card powered by the AMD Virtex Ultrascale+ VU2P FPGA, running on the Exegy nxFramework and Exegy nxTCP-UDP-10g-ULL IP Core, in a Dell PowerEdge R7525 server with AMD EPYC 7313 processors. See https://stacresearch.com/news/AMD240422 for the full STAC report. AMD holds the previous world record for latency (2020): https://www.stacresearch.com/news/XLX200514. Stated results for the Alveo UL3524 accelerator have been extrapolated to the AMD Alveo UL3422 card, based on identical silicon and product features. (ALV-20).
- Testing conducted by AMD as of 8/16/23, using Vivado™ Design Suite 2023.1 running on Vivado Lab (Hardware Manager) 2023.1, and the GTF Latency Benchmark Design on the Alveo UL3524 accelerator card, configured to enable GTF transceivers in internal near-end loopback mode. Benchmark was run 1,000 times with 250 frames per test, where PCS logic (physical medium attachment) of the transceiver passes data ‘as-is’ to FPGA fabric to measure latency. Stated latency result based on GTF transceiver “RAW Mode.” Configuration differences or variance in testing protocol will yield different results. Actual results will vary (ALV-10).
- Based on a simulated comparison by AMD using the Synopsis VCS 2019.06-SP2 ultra-low latency Virtex UltraScale+ VU2P GTF transceivers and GTY transceivers in February 2022 (ALV-15).
- Cost reduction of the AMD Alveo UL3422 accelerator, as compared to the previous generation Alveo UL3534 accelerator. Pricing is based on MSRP as of September 2024 (ALV-21).