Newsroom
Monday, August 17, 2020

Inspur NF5488A5 Breaks AI Server Performance Record in Latest MLPerf Benchmarks

Posted to Business Wire.

NF5488A5 completed the ResNet50 model training in only 33.37 minutes, ranking first in single server performance of Resnet50 benchmark.

San Jose, California, Aug. 5 – MLPerf, an industry-standard AI benchmark, released its latest results. Inspur NF5488A5 server has set a new AI performance record in the most renowned Resnet50 training task, topping the list of all single server performance.

MLPerf is the most influential industry benchmarking organization in the field of AI around the world. Established in May 2018, MLPerf is supported and participated in by a number of industry giants and academic institutions, among which are Amazon, Baidu, Facebook, Google, Harvard University, Intel, NVIDIA, Microsoft, Alibaba, Inspur, and Stanford University. The MLPerf 0.7 training benchmark included 8 tasks focusing on typical deep learning scenarios like image classification, object detection, reinforcement learning, recommendation, and translation. A total of 9 organizations participated in the training benchmark and submitted results, namely Google, NVIDIA, Intel, Alibaba, Tencent, Inspur, Dell, Fujitsu, and SIAT.

Resnet50 is the world’s most widely-accepted standard for evaluating the performance of AI computing systems and AI chips. In the Resnet50 training task of this benchmark, Inspur used NF5488A5 server equipped with 8 NVIDIA A100 GPUs and 2 AMD EPYC 7742 processors. By completing the ResNet50 model training in only 33.37 minutes, the server ranked first in single server performance.

Launched in this May, NF5488A5 is a new AI server powered by 8 NVIDIA A100 GPUs fully interconnected with the 3rd generation NVLink plus 2 of the latest AMD CPUs supporting PCIe4.0. It provides users with ultimate AI performance and ultra-high-speed bandwidth while offering powerful computing support for various AI scenarios such as intelligent customer service, financial analysis, smart city, and intelligent language processing.