Perlmutter, the world's fastest supercomputer at the National Energy Research Scientific Computing Center (NERSC), has been turned on.
One of the supercomputer's initial initiatives will be to create the largest 3D map of the universe, using the 6159 NVIDIA A100 Tensor Core GPUs.
According to NVIDA, Perlmutter currently produces over four exaFLOPS of AI performance, making it "the fastest system in the planet on 16- and 32-bit mixed-precision math AI needs." With the addition of a second tranche of CPU cores in "phase two," its performance will be boosted even more.
Nvidia Corporation is a multinational technology company based in Santa Clara, California.
SANTA CLARA, CA - MAY 10: A sign is posted in front of the Nvidia headquarters on May 10, 2018 in Santa Clara, California. Nvidia Corporation will report first quarter earnings today after the closing bell.
Perlmutter to Process Data From Dark Energy Spectroscopic Instrument
The Perlmutter will use data from Dark Energy Spectroscopic Instrument (DESI) to create the world's largest 3D map of the Universe. DESI is a deep-space imaging equipment that can catch more than 5,000 galaxies in a single image.
According to Nvidia, researchers can interpret photos from the previous night using Nvidia's A100 Tensor Core GPUs, so they know where to point DESI the next time.
Perlmutter is much faster than previous systems.
"Preparing a year's worth of the data for publication would take weeks or months on prior systems, but Perlmutter should help them accomplish the task in as little as a few days," NVIDIA said in a blog posted on their website.
The main goal of the 3D map is to "shine light on dark energy" which is thought to be responsible for the universe's continuing expansion. The supercomputer is named after a scientist named Saul Perlmutter. He won the Nobel Prize in Physics in 2011 for his work on dark energy.
DYK #Perlmutter will quadruple the computational power currently @NERSC?
The @HPE_Cray EX supercomputer—containing @nvidia A100 Tensor Core GPUs + 3rd Gen @AMD EPYC CPUs— will be among the world's fastest for simulation, analysis, and #AI.
More: https://t.co/XeAcwkqIZJ | #HPC pic.twitter.com/b3hypWd7XI— NERSC (@NERSC) May 28, 2021
Dark energy is a hypothetical form of negative energy which is the opposite of gravity. It is believed to account for the accelerated period of expansion of the universe.
Scientists also want to harness Perlmutter's computer prowess to analyze atomic interactions to develop better biofuels and batteries. Traditional supercomputers, according to Nvidia, are "barely capable" of processing the math required to produce atom simulations in a few nanoseconds.
ALSO READ: Preventing AI Takeover: Microsoft, IBM, Nvidia Join Campaign to Protect Us From AI Threats
Scientists want to achieve just that with Nvidia's latest supercomputer.
Nvidia said the Perlmutter system will be used by over 7,000 researchers, and tens of apps have already been prepared for astronomy and climate science.
"In one project, the supercomputer will help assemble the largest 3D map of the visible universe to date. It will process data from DESI, a kind of cosmic camera that can capture as many as 5,000 galaxies in a single exposure," Dion Harris, Nvidia HPC & AI Product Marketing Lead, explained in a TechRadar report.
Is Software Essential in Perlmutter?
Other initiatives planned for the Perlmutter machine have equally lofty objectives. Tom's Hardware said a handful of them will use the Tensor Cores in the A100's unique properties to model atom interactions that were previously difficult to simulate.
Perlmutter's supercomputer has both cutting-edge hardware and cutting-edge software.
Nvidia has been concentrating on AI for a long time. With Perlmutter on board, the business believes "AI supercomputing" is now here. AI is expected to play a significant role in scientific advancements, according to scientists.
Perlmutter supports "OpenMP and other popular programming models in the NVIDIA HPC SDK the system employs," according to the blog post.
This, in combination with RAPIDS-an open-source data science code for GPUs will greatly accelerate scientific data discovery.
RELATED ARTICLE: Robots Finally Confirm Why Fish Swim in School
Check out more news and information on Technology on Science Times.
