NVIDIA has announced a significant shift in the high-performance computing industry towards AI-powered systems. The company revealed that nine new supercomputers worldwide are utilizing NVIDIA Grace Hopper™ Superchips to expedite scientific research and discovery. Collectively, these systems deliver 200 exaflops, equivalent to 200 quintillion calculations per second, of energy-efficient AI processing power.

Among the new Grace Hopper-based supercomputers coming online are EXA1-HE in France, Helios in Poland, Alps in Switzerland, JUPITER in Germany, DeltaAI in Illinois, USA, and Miyabi in Japan. These machines are expected to contribute significantly to various sectors of scientific research.

In April, CEA (the French Alternative Energies and Atomic Energy Commission) and Eviden (an Atos Group company) announced the delivery of the EXA1-HE supercomputer. This machine is based on Eviden’s BullSequana XH3000 technology and features a new patented warm-water cooling system. The EXA1-HE is equipped with 477 compute nodes based on Grace Hopper.

Ian Buck, vice president of hyperscale and HPC at NVIDIA stated that "AI is accelerating research into climate change, speeding drug discovery and leading to breakthroughs in dozens of other fields." He further noted that "NVIDIA Grace Hopper-powered systems are becoming an essential part of HPC for their ability to transform industries while driving better energy efficiency."

Additionally, Isambard-AI and Isambard 3 from the University of Bristol in the U.K., along with systems at the Los Alamos National Laboratory and the Texas Advanced Computing Center in the U.S., have joined a growing wave of NVIDIA Arm-based supercomputers using Grace CPU Superchips and the Grace Hopper platform.

The construction of new, more efficient AI-based supercomputers is gaining momentum as countries globally recognize sovereign AI's strategic and cultural importance. Sovereign AI involves investing in domestically owned and hosted data, infrastructure, and workforces to stimulate innovation.

NVIDIA's GH200, which combines the Arm-based NVIDIA Grace CPU and NVIDIA Hopper™ GPU architectures using NVIDIA NVLink®-C2C interconnect technology, is powering scientific supercomputing centers worldwide. Many of these centers plan to transition from system installation to real science in months rather than years.

Isambard-AI phase one comprises an HPE Cray EX2500 supercomputer with 168 NVIDIA GH200 Superchips. When the remaining 5,280 NVIDIA Grace Hopper Superchips arrive at the University of Bristol’s National Composites Centre this summer, performance is expected to increase by approximately 32x.

Simon McIntosh-Smith, professor of high-performance computing at the University of Bristol said that "Isambard-AI positions the U.K. as a global leader in AI, and will help foster open science innovation both domestically and internationally." He also noted that working with NVIDIA has allowed them to deliver phase one of the project in record time.

NVIDIA’s accelerated computing platform consists of NVIDIA Hopper architecture-based GPUs, NVIDIA Grace CPU Superchips, NVIDIA Grace Hopper Superchips, NVIDIA Quantum-2 InfiniBand networking and a full suite of NVIDIA AI and HPC software.