Inspur's AI server just shattered a world speed record

By Harry Menear
The data centre infrastructure firm Inspur just set a world record with its NF5488A5 AI server...

Inspur, a Silicon Valley-based data centre infrastructure, cloud computing services and AI solutions provider, just set a new world record with its flagship NF5488A5 server.

Launched in May of this year, the NF5488A5 is a new AI server powered by eight NVIDIA A100 GPUs fully interconnected with the 3rd generation NVLink plus two of the latest AMD CPUs supporting PCIe4.0.

“NVIDIA A100 Tensor Core GPUs offer customers unmatched acceleration at every scale for AI, data analytics and HPC,” commented Paresh Kharya, Director of Product Management for Accelerated Computing at NVIDIA back in May. “Inspur AI servers, powered by NVIDIA A100 GPUs, will help global users eliminate their computing bottlenecks and dramatically lower their cost, energy consumption, and data centre space requirements.”

The server, which is designed to deliver ultra-high-speed bandwidth while running AI applications - the company lists scenarios like intelligent customer service, financial analysis, smart city, and intelligent language processing - recorded its results using MLPerf, an in dustry-standard AI benchmarking organisation. 

According to Inspur, MLPerf is the most influential industry benchmarking organisation 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, including 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 organisations 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’s NF5488A5 server completed the ResNet50 model training in only 33.37 minutes.

That makes it the fastest single AI server on the market today.


Featured Articles

The importance of properly training AI to avoid bias

Kausik Chaudhuri, CIO at software-enabled services provider Lemongrass, shares his insights into the importance of properly training and implementing AI

Is XaaS the solution for eco-conscious business growth?

With data centres historically significant polluters, Demed L’Her explores sustainable decisions businesses can take to reduce the impact of using XaaS

How data management is key to successful ESG reporting

Levent Ergin, Chief ESG Sustainability Strategist, Informatica, explains how data collection & management are essential to transparent ESG reporting

Data Centre Magazine: Celebrations as we mark 3rd birthday

Data Centres

Equinix and the future of data centre digital transformation

Data Centres

Eastdil bank appoints Rob Walters as data centre chief

Data Centres