Nvidia to host Argonne National Laboratory's largest supercomputer

Nvidia has announced its accelerated computing platform will host Argonne National Laboratory's largest supercomputer, Polaris.

The HPE-developed Polaris is estimated to have 560 nodes with four Nvidia A100 Tensor Core GPUs each. Up to 20 times faster than its predecessor, A100 can be partitioned into seven GPU instances on demand.

With 2,240 Nvidia A100 Tensor Core GPUs, Polaris can reach nearly 1.4 exaflops of theoretical artificial intelligence (AI) performance and approximately 44 petaflops of peak double-precision. Pairing simulation with machine learning, this supercomputer can tackle the most data-intensive AI computing workloads.

According to the US Department of Energy, Polaris will supercharge Argonne Leadership Computing Facility’s (ALCF’s) scientific research and algorithms.

In addition to government agencies, Polaris can be accessed by academic researchers and industry experts, thanks to ALCF's peer-reviewed application and allocation programs.

“Polaris is a powerful platform that will allow our users to enter the era of exascale AI,” stated Michael Papka, director of ALCF.

Papka added, “Harnessing the huge number of Nvidia A100 GPUs will have an immediate impact on our data-intensive and AI HPC workloads, allowing Polaris to tackle some of the world’s most complex scientific problems.”

Lastly, Polaris will allow researchers to upload their workloads to Aurora, the upcoming exascale system from Argonne.

“The era of exascale AI will enable scientific breakthroughs with massive scale to bring incredible benefits for society,” said Ian Buck, vice president and general manager of accelerated computing at Nvidia.

“Nvidia’s GPU-accelerated computing platform provides pioneers like the ALCF breakthrough performance for next-generation supercomputers such as Polaris that let researchers push the boundaries of scientific exploration.”