Advanced Micro Devices ( AMD Quick Quote AMD - Free Report) is expanding its FPGA portfolio with the launch of the Alveo UL3524 accelerator card, designed for ultra-low latency electronic trading applications. The latest state-of-the-art FPGA solution will help AMD expand its footprint in the fintech domain. Powered by a custom 16 nanometer Virtex UltraScale+ FPGA, the Alveo UL3524 accelerator card delivers seven times lower latency compared with the previous generation FPGA technology. This helps it achieve less than three nanoseconds of FPGA transceiver latency for superfast trading. This superlative performance is achieved due to a novel transceiver architecture with hardened, optimized network connectivity cores. Alveo UL3524 accelerator card features 64 ultra-low latency transceivers, 780K LUTs of FPGA fabric, and 1,680 DSP slices of compute that accelerate custom trading algorithms in hardware. In terms of software, AMD is offering developers access to the open-sourced and community-supported FINN development framework. FINN uses PyTorch and neural network quantization techniques that help developers reduce the size of AI models while retaining accuracy and compiling to hardware IP. AMD’s strong partner base is expected to drive the adoption of the new FPGA solutions. Alpha Data, Exegy, and Hypertec are currently offering solutions leveraging the Alveo UL3524 accelerator card and Virtex UltraScale+ FPGA.
Alpha Data’s ADA-R9100 rack-mount appliance is supported by Virtex UltraScale+ FPGA. Exegy is supporting the Alveo UL3524 card with its nxFramework, a software and hardware development environment customized for creating and maintaining ultra-low latency FPGA applications within the financial industry.
Meanwhile, Hypertec has optimized its ORION HF X410R-G6 High Frequency Server for the Alveo UL3524, with a custom cooling system to deploy in a 1U server form factor. AMD Benefits From Strong Portfolio & Partner Base
AMD shares have returned 51.4%, outperforming the Zacks Computer & Technology sector’s growth of 30.8% on a year-to-date basis. The outperformance can be attributed to a robust product portfolio and expanding partner base.
AMD continues to strengthen its footprint in the enterprise data center arena by leveraging the power of fourth-generation EPYC CPUs and Pensando data processing units (DPUs). At VMware Explore 2023, AMD demonstrated that the combination of a fourth-generation EPYC 9654 CPU and a Pensando DPU delivers a staggering 3.3 times boost in Redis application performance and a 1.75 times surge in aggregate network throughput compared to a fourth-generation EPYC system equipped with standard NICs. Moreover, servers with 2P fourth-generation EPYC 9654 CPUs use up to 35% fewer servers in a setup running 2000 virtual machines (VMs) compared with 2P Xeon 8490H-based servers from Intel ( INTC Quick Quote INTC - Free Report) . The combination of EPYC CPUs, Pensando DPUs and adaptive computing products is helping AMD to solidify its position as a go-to solution for enhanced virtualized environments. The company’s ongoing collaboration with VMware is helping customers become more efficient and agile in their digital transformation journey. AMD, along with its partners, continues to offer solutions that enable greater data center consolidation. The company is benefiting from the impressive performance of EPYC CPUs, delivering a 1.7 times performance increase in combination with 2P EPYC 9654 CPUs compared to 2P Intel Xeon 8490H-based servers running VMmark. Moreover, AMD’s partnerships with the likes of Dell Technologies ( DELL Quick Quote DELL - Free Report) , Microsoft ( MSFT Quick Quote MSFT - Free Report) , Amazon Web Services (“AWS”), Alibaba and Oracle have been key catalysts. Cloud providers like Microsoft, AWS, Alibaba and Oracle deployed Genoa in the second quarter of 2023. Dell is leveraging AMD EPYC fourth-generation CPUs in its latest PowerEdge C6615 server. In combination with OpenManage Enterprise software, Dell servers enable cloud service providers to intelligently monitor their systems and deliver more efficient compute services. Microsoft cloud arm Azure announced the first Genoa-X HPC instances that offer more than five times higher performance in technical computing workloads compared to their prior generation. Currently, there are more than 670 AMD-powered cloud instances publicly available, and this number is expected to grow 30% to 900 by the end of 2023, primarily due to the adoption of Genoa. Moreover, the availability of Bergamo is crucial, as several server providers like Dell, HPE, Lenovo and Supermicro are set to launch their new Bergamo-based platforms in the current quarter. This is expected to boost AMD, which currently has a Zacks Rank #3 (Hold), over the long term. You can see the complete list of today’s Zacks #1 Rank (Strong Buy) stocks here.