Share
Dimapur, May 30 (EMN): US-based tech major Nvidia and the National Energy Research Scientific Computing Center (NERSC) has announced that Perlmutter, an AI supercomputer, will help build the largest-ever 3D map of the visible universe.
Named after Nobel Prize-winning astrophysicist Saul Perlmutter, the supercomputer is touted to be the "fastest system on the planet on the 16- and 32-bit mixed-precision math AI uses."
Perlmutter is designed to deliver nearly four exaflops of AI performance for more than 7,000 researchers. AMD has confirmed that Phase 1 of the Perlmutter system consists of 1,536 nodes, each with one AMD EPYC 7763 processor and four NVIDIA NVlink-connected A100 Tensor Core GPUs.
Phase 2 will add another 3,072 CPU-only nodes, each with two AMD EPYC 7763 processors and 512 GB of memory per node.
Congratulations to @BerkeleyLab and @NERSC on today’s launch of the “Perlmutter” supercomputer. Looking forward to the groundbreaking research that it will enable with @AMDServer 7003 Series processors! pic.twitter.com/9eS00PRLNc
— Lisa Su (@LisaSu) May 27, 2021
JUST IN: Our Lab has deployed its #nextgen , named in honor of the Lab's -winning astrophysicist Saul Perlmutter. aims to bolster US scientific research!
— Berkeley Lab (@BerkeleyLab)
The Perlmutter supercomputer will process data collected by the Dark Energy Spectroscopic Instrument to help guide scientists on their quest to map the universe in a precise 3D model and to better understand dark energy by accurately measuring the expansion of the universe.
"In one project, the supercomputer will help assemble the largest 3D map of the visible universe to date. It will process data from the Dark Energy Spectroscopic Instrument (DESI), a kind of cosmic camera that can capture as many as 5,000 galaxies in a single exposure," Nvidia global HPC/AI product marketing lead Dion Harris wrote in a blog post.
Besides helping scientists map the universe, Perlmutter will also help researchers create better batteries and biofuels.
"In the past, it was impossible to do fully atomistic simulations of big systems like battery interfaces, but now scientists plan to use Perlmutter to do just that,” wrote Brandon Cook, an applications performance specialist at NERSC.