Nvidia showcased data center skills at this year’s European supercomputer event
ISC Group’s flagship supercomputing event in Germany, ISC High Performance 2022, Returned to physical form this year for the first time in two years at the Congress Center in Hamburg, Germany. The event, which runs through tomorrow, has attracted more than 120 exhibitors At a special address there, Nvidia Corporation sketched out a daring stroke Its vision Accelerated computing using its hardware and software-driven high-performance computing (HPC) solution.
Nvidia has unveiled the new Venado supercomputer at Los Alamos National Laboratory (LANL), which will provide 10 exoflop computer power for artificial intelligence (AI) research and development aimed at physics and renewable energy. The company claims that its Arm v9-based Grace Hopper “Superchips” will run Venado three times faster than the powering company’s last-generation graphics processing units (GPUs), the company said. What’s more, the Grace Central Processing Unit (CPU) can provide twice as much efficiency per watt as traditional CPUs.
Make the best of a bad situation
Nvidia famously failed to acquire Arm earlier this year. Regulators in the three countries have rejected a $ 40 billion deal, citing potential competitive impact in the semiconductor industry. Softbank has opted to spin off the arm later this year instead of an initial public offering (IPO). It will maintain a majority stake in Arm, but the news has attracted numerous investors inside and outside the semiconductor business, including Qualcomm and competitors.
While the ultimate future of the Arm may still be in question, Nvidia believes that the Arm unlocks new opportunities for accelerated computing. When Nvidia founder and CEO Jensen Huang fell through the arm deal, he made it clear that his company’s arm cooperation would continue regardless. Nvidia rolled out Grace Hopper Designs at its own GTC event a month later.
At last week’s Computex 2022 event in Taipei, Nvidia showcased the first manufacturing data center solutions built around Grace Hopper, named after a Navy admiral and computer scientist whose groundbreaking work led to machine-independent programming languages such as COBOL. At the same event, Nvidia introduced the Liquid-Cooled A100 GPUs, designed to reduce data center power consumption by up to 30% using more efficient cooling.
Creating fusion energy with omniverse
While Facebook maker Meta beats the drum of infrastructural collaborations to enliven its metavers vision, Nvidia is in a very different place. The equivalent of Nvidia’s Metaverse, which it calls Omniverse, is a real thing. Omniverse is a collaborative 3D environment used in production for artists, engineers and research scientists. Ericsson has used Omniverse to create digital twin city simulations, to aid in the study of its 5G small cell density before physical placement. The system helped Ericsson alleviate signal interference and propagation problems, which were not possible before.
At the ISC this week, Nvidia explained that Omniverse is a platform for researchers in the UK to create real-time simulations of an operating fusion power plant. Fusion energy may not yet exist in the real world, but researchers hope it will save enough time for the research and development needed to create a workable solution, once the difficult real-world physics and engineering work is done.
Accelerate computing research around a new version of Quantum Computing and Nvidia’s imaging software Holscan that has rounded out Nvdia’s ISC 2022 announcements aimed at HPC Edge settings.
