What Is The News About?
NVIDIA announced that nine new supercomputers throughout the globe are utilizing NVIDIA Grace HopperTM Superchips to accelerate scientific research and discovery. These systems are driven by artificial intelligence (AI), which is driving a fundamental shift in the high-performance computing market. Effortlessly processing 200 quintillion calculations per second, or 200 exaflops, of AI power is delivered by the combined systems.
In April, the EXA1-HE supercomputer, built on Eviden’s BullSequana XH3000 technology, was delivered to CEA, the French Alternative Energies and Atomic Energy Commission. Eviden is a member of the Atos Group. A new, patented warm-water cooling system is available in the BullSequana XH3000 architecture, and the EXA1-HE comes with 477 Grace Hopper-based compute nodes. Furthermore, systems at the Texas Advanced Computing Center and Los Alamos National Laboratory in the United States, as well as Isambard-AI and Isambard 3 from the University of Bristol in the United Kingdom, are part of an increasing number of NVIDIA Arm-based supercomputers that utilize Grace CPU Superchips and the Grace Hopper platform.
Why Is This News Important?
One of the most efficient supercomputers ever built, the HPE Cray EX2500, is part of the first phase of Isambard-AI. It features 168 NVIDIA GH200 Superchips. The performance will be increased by approximately 32x when the remaining 5,280 NVIDIA Grace Hopper Superchips arrive at the National Composites Centre at the University of Bristol this summer. GPUs built on the NVIDIA Hopper architecture, CPU and Hopper Superchips from NVIDIA, Quantum-2 InfiniBand networking from NVIDIA, and a comprehensive suite of AI and HPC software from NVIDIA make up the company’s accelerated computing platform.
Countries worldwide are investing in domestically owned and hosted data, infrastructure, and workforces to stimulate innovation, which is known as “sovereign AI.” As a result, the urge to create new, more efficient AI-based supercomputers is rising. The GH200 is the brains behind scientific supercomputers all around the world. It combines the Arm-based NVIDIA Grace CPU with the NVIDIA HopperTM GPU architectures, and it uses NVIDIA NVLink®-C2C connection technology. Instead of years, several institutes want to go from installing the system to doing actual science in a matter of months.
Benefits
1. Scientific Acceleration: NVIDIA’s Grace HopperTM Superchips power nine new supercomputers, revolutionizing scientific research with 200 exaflops of AI capability, driving discoveries worldwide.
2. Cutting-edge Cooling: Eviden’s BullSequana XH3000, equipped with patented warm-water cooling, hosts the EXA1-HE supercomputer, enhancing efficiency and sustainability in high-performance computing.
3. Enhanced Performance: The HPE Cray EX2500, with 168 NVIDIA GH200 Superchips, achieves unparalleled efficiency, set to skyrocket with an additional 5,280 Grace Hopper Superchips.
4. Global Advancements: From France to Japan, new supercomputers like EXA1-HE, Helios, and Miyabi, empowered by Grace Hopper technology, spearhead sovereign AI initiatives, fostering innovation worldwide.
Wrapping Up
The following new Grace Hopper-based supercomputers are set to go online: EXA1-HE in France, built by CEA and Eviden; Helios in Poland, by Academic Computer Centre Cyfronet, funded by Hewlett Packard Enterprise (HPE); the Alps at the Swiss National Supercomputing Centre, also funded by HPE; JUPITER at the Jülich Supercomputing Centre, Germany; DeltaAI at the National Center for Supercomputing Applications, UIC; and Miyabi at Japan’s Joint Center for Advanced High-Performance Computing, a collaboration between the University of Tsukuba and the University of Tokyo’s Information Technology Center, UIC, and the University of Tokyo’s Center for Computational Sciences, UIT, and the University of Tokyo’s Center for Information Technology Center.
[To share your insights with us as part of editorial or sponsored content, please write to sghosh@martechseries.com]
Comments are closed.