Inspur Announces General Availability of Servers with Qualified Xilinx Alveo Accelerator Cards
Inspur has released its NF5280M5 and NF5468M5 servers with qualified Xilinx Alveo U200 and U250 accelerator cards to the general market, following the successful testing and certification of such accelerator cards on the servers.
NF5280M5 and NF5468M5, two of Inspur’s leading server products, are designed for AI application optimization and AI cloud. They can adapt to user demands and provide computing power with high-performance and high-reliability for common AI application scenarios including deep learning, offline training, voice recognition and video analysis. Xilinx Alveo U200 and U250 are designed for data centers with AI workloads, providing low latency, up to 90x performance increase over CPU, and 1/3 the total cost of ownership for common workloads. The accelerators can adapt to changing acceleration requirements and algorithmic standards, accelerating any workload without hardware changes, thus providing an application insurance policy for the future.
Read More: The Artificial Intelligence Week
“FPGA technology is becoming more and more popular among AI users due to its customizability, low latency and outstanding power efficiency,” said Liu Jun, General Manager of AI&HPC, Inspur. “Qualified on two mainstream Inspur servers, the Alveo U200 and U250 will provide users with more advanced AI computing acceleration solutions, helping them flexibly migrate workloads and algorithms. Inspur will further cooperate with Xilinx to provide more FPGA solutions and build upon its large and growing software ecosystem, to better meet global AI users’ demand for higher computing power.”
“Inspur is the world’s leading provider of AI computing power, and Xilinx is very pleased to have this opportunity to cooperate with them,” said Freddy Engineer, corporate vice president, data center sales, at Xilinx, Inc. “We are also grateful for Inspur’s interest and recognition of Alveo U200 and Alveo U250. Working together, we will provide powerful computing acceleration for workloads including video transcoding, financial risk management, database search and analytics, security, medical AI applications, improving users’ work efficiency and reducing operating costs.”