2 min Analytics

Nvidia makes supercomputing easy with new AI technology

Nvidia makes supercomputing easy with new AI technology

Latest version of SuperPod makes supercomputing available to enterprises.

Nvidia opened its virtual GPU Technology Conference on Monday by announcing NVIDIA DGX A100. The company claims that this new infrastructure offering will allow customers to build a supercomputer optimized for deep learning in a very short period of time, often just a few weeks.

“Traditional supercomputers can take years to plan and deploy, but the turnkey NVIDIA DGX SuperPOD Solution for Enterprise helps customers begin their AI transformation today,” said Charlie Boyle, vice president and general manager of DGX systems at NVIDIA.

The NVIDIA DGX A100 solution will be in multiple global markets by year’s end

SuperPODs are available in cluster sizes ranging from 20 to 140 individual NVIDIA DGX A100 systems. They are now shipping and expected to be installed in Korea, the U.K., Sweden and India before the end of the year.

“State-of-the-art conversational AI, recommender systems and computer vision workloads rapidly exceed the capabilities of traditional infrastructure,” said Boyle. “And our new solution gives customers a fast track to the world’s most advanced, scalable AI infrastructure and NVIDIA expertise.”

Nvidia’s graphics cards can be found in many of the world’s fastest supercomputers. With SuperPod Solution for Enterprise, Nvidia hopes to increase its chip sales by expanding its market base.

The company hopes to realise this expansion by making its supercomputing technology available at the enterprise level.

How it works

SuperPod makes it possible to cobble together a multipetaflop system in a matter of weeks. Customers simply need to use Nvidia compute appliances and networking gear from its Mellanox subsidiary.

The basic building block of a SuperPod supercomputer is Nvidia’s DGX A100 appliance. The appliance pairs eight of the chipmaker’s top-of-the-line A100 data center graphics processing units with two central processing units, plus a terabyte of memory.

SuperPod allows customers to link between 20 and 140 DGX A100 appliances in a network. SuperPod effectuated the connection via Mellanox’s HDR InfinityBand switches. This process creates a supercomputer with 100 to 700 petaflops of processing power, Nvidia claims. One petaflop equals a million billion computing operations per second.

The DGX SuperPOD Solution for Enterprise is available from select NVIDIA partners worldwide.