Nvidia maps the universe with the world’s most powerful artificial intelligence supercomputer. Nvidia ran a new supercomputer said to be the world’s fastest for artificial intelligence (AI)-based workloads at the National Energy Research Scientific Computer Center (NERSC) in California, USA.
The supercomputer is named after astrophysicist Saul Perlmutter and will initially be tasked with creating the most ambitious three-dimensional map of the universe ever seen using more than 6,000 Nvidia A100 Tensor Core GPUs in-house.
In its current state, Perlmutter can deliver nearly four exaFLOPS of AI performance, making it “the fastest system on the planet for 16- and 32-bit mixed precision math AI uses,” Nvidia says. Performance will be further enhanced as part of the “second phase” with the addition of a second CPU core slice.
When the system is complete, Perlmutter is expected to be among the top five supercomputers in the Top 500 ranking, which evaluates overall performance using the High Performance Linpack (HPL) benchmark.
Notable features of the Nvidia supercomputer
More than 7,000 researchers will run the Perlmutter system, according to Nvidia. Dozens of applications have already been created to advance astrophysics and climate science.
“In one project, the supercomputer will help create 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 up to 5,000 galaxies in a single exposure. ” statement was included.
“Researchers need the speed of Perlmutter’s GPUs to capture dozens of exposures from one night to know where to direct DESI the next night. Preparing a year’s worth of data for publication would have taken weeks or months in previous systems, but Perlmutter should help them accomplish the task in as little as a few days.”
It is hoped that the map of the universe, once complete, will help researchers learn more about dark energy, the force behind the ever-accelerating expansion of the universe. In 2011, the discovery of dark energy earned Saul Permutter the Nobel Prize.
Other projects to run on the Perlmutter machine have similarly ambitious goals. Some of these will use the unique attributes of Tensor Cores in the A100 to simulate interactions between atoms that were previously nearly impossible.
This work could be illuminating in the field of materials science, potentially uncovering ways to develop more efficient batteries, biofuels and the like.