The Miyabi supercomputer in Japan is using NVIDIA Grace Hopper superchips to power its advanced high-performance computing (HPC) initiatives.
Jointly by the Center for Computational Sciences at the University of Tsukuba and the Information Technology Center at the University of Tokyo, the supercomputer harnesses the Grace Hopper superchips which can collectively deliver 200 exaflops of energy-efficient AI processing power.
It is among several new Grace Hopper-based supercomputers across the world that have come online to herald a new era of AI supercomputing.
Developed by CEA and Eviden in France, the EXA1-HE supercomputer is based on Eviden’s BullSequana XH3000 technology. It features a patented warm-water cooling system and is equipped with 477 compute nodes powered by Grace Hopper.
Located at the Academic Computer Centre Cyfronet in Poland, Helios is another Grace Hopper-based supercomputer. It promises high-performance computing capabilities for scientific endeavors.
The Swiss National Supercomputing Centre hosts the Alps supercomputer, which leverages Grace Hopper technology to advance research across various domains.
In Germany, the Jülich Supercomputing Centre houses the Jupiter supercomputer that leverages the power of Grace Hopper for computational simulations and data analysis.
At the National Center for Supercomputing Applications in the US, DeltaAI contributes to AI-driven research using Grace Hopper superchips.
Isambard-AI and Isambard 3 from the University of Bristol (UK) and systems at the Los Alamos National Laboratory and the Texas Advanced Computing Center (US) contribute to the growing wave of NVIDIA Arm-based supercomputers using Grace CPU superchips and the Grace Hopper platform.
“AI is accelerating research into climate change, speeding drug discovery and leading to breakthroughs in dozens of other fields. NVIDIA Grace Hopper-powered systems are becoming an essential part of HPC for their ability to transform industries while driving better energy efficiency,” said Ian Buck, Vice President of Hyperscale and HPC at NVIDIA.
Entelechy Asia 