X
Tech

Nvidia, AMD show high performance computing gains

The new TOP500 supercomputers list​ dropped on Monday and both Nvidia and AMD are touting momentum in the HPC market.
Written by Natalie Gagliordi, Contributor

The new TOP500 supercomputers list dropped on Monday and both Nvidia and AMD are touting gains in their high-performance computing technology. Nvidia said its GPU accelerated systems are powering eight of the 10 fastest supercomputers, while AMD claims that its EPYC processors are powering four of the 50 highest-performance supercomputing systems in the world. 

Nvidia's supercomputing momentum is tied to the A100, the firm's 8th generation GPU design and its first based on Ampere architecture. Nvidia said the A100 represents the biggest generational leap ever for one of its GPUs. Designed for data centers, the multi-instance GPU is optimized for HPC and inference, delivering 20x speed improvements over Volta, with more than 54 billion transistors and third-generation Tensor Cores.

Nvidia said the latest DGX system, the DGX A100, includes new InfiniBand interconnect technology from Mellanox, which Nvidia bought for $7 billion in 2019. Mellanox's InfiniBand switches were integrated with the A100 to create Selene, which has debuted in the No. 7 spot on the TOP500 list. Designed in less than a month, Nvidia said Selene provides more than 1 exaflop of AI processing, and claims that it's highly energy-efficient and the fastest industrial supercomputer in the US. 

"Combined with Mellanox our position is stronger than ever," said Nvidia product marketing director Paresh Kharya in a press briefing. "If you look at the top 500 list, the reason why Nvidia is so successful in supercomputing is because scientific computing has changed. We've entered a new era, one that has expanded beyond traditional modeling and simulation workloads to include AI, data analytics, edge screening, and big data visualization."

Meanwhile, AMD continues to tout its momentum in the high-performance computing market. AMD said its EPYC processors are powering four of the 50 highest-performance supercomputing systems in the world. AMD's processors are also part of new HPC installations at Indiana University, Purdue University and CERN, as well as HPC cloud instances from Amazon Web Services, Google and Oracle Cloud. 

The four EPYC-powered supercomputers on the latest TOP500 listing come from French firm Atos. The systems include Joliot-Curie, Belenos, Mahti, and Uninett Sigma2.

AMD's EPYC processors are also part of Frontier and El Capitan, two of the world's fastest exascale supercomputers still in development. El Capitan, expected to come online in 2023, will be used to protect the US nuclear stockpile. The Frontier supercomputer is set to arrive at the Lawrence Livermore National Laboratory in 2021. 

"Our AMD EPYC CPUs, Radeon Instinct accelerators and open software programming environment are helping to advance the industry towards exascale-class computing, and we are proud to strengthen the global HPC ecosystem through our support of the top supercomputing clusters and cloud computing environments," said Forrest Norrod, SVP and GM of AMD's data center and embedded systems group. 

Editorial standards