NVIDIA has launched its next-generation of AI supercomputer chips that can doubtless play a big position in future breakthroughs in deep studying and enormous language fashions (LLMs) like OpenAI's GPT-4, the corporate introduced. The expertise represents a major leap over the past era and is poised for use in knowledge facilities and supercomputers — engaged on duties like climate and local weather prediction, drug discovery, quantum computing and extra.
The important thing product is the HGX H200 GPU primarily based on NVIDIA's "Hopper" structure, a alternative for the favored H100 GPU. It's the corporate's first chip to make use of HBM3e reminiscence that's sooner and has extra capability, thus making it higher fitted to massive language fashions. "With HBM3e, the NVIDIA H200 delivers 141GB of reminiscence at 4.8 terabytes per second, almost double the capability and a pair of.4x extra bandwidth in contrast with its predecessor, the NVIDIA A100," the corporate wrote.
When it comes to advantages for AI, NVIDIA says the HGX H200 doubles inference pace on Llama 2, a 70 billion-parameter LLM, in comparison with the H100. It'll be out there in 4- and 8-way configurations which might be suitable with each the software program and {hardware} in H100 techniques. It'll work in each kind of information middle, (on-premises, cloud, hybrid-cloud and edge), and be deployed by Amazon Net Providers, Google Cloud, Microsoft Azure and Oracle Cloud Infrastructure, amongst others. It's set to reach in Q2 2024.
NVIDIA's different key product is the GH200 Grace Hopper "superchip" that marries the HGX H200 GPU and Arm-based NVIDIA Grace CPU utilizing the corporate's NVLink-C2C interlink. It's designed for supercomputers to permit "scientists and researchers to deal with the world’s most difficult issues by accelerating advanced AI and HPC purposes operating terabytes of information," NVIDIA wrote.
The GH200 will probably be utilized in "40+ AI supercomputers throughout world analysis facilities, system makers and cloud suppliers," the corporate stated, together with from Dell, Eviden, Hewlett Packard Enterprise (HPE), Lenovo, QCT and Supermicro. Notable amongst these is HPE's Cray EX2500 supercomputers that can use quad GH200s, scaling as much as tens of hundreds of Grace Hopper Superchip nodes.
Maybe the most important Grace Hopper supercomputer will probably be JUPITER, positioned on the Jülich facility in Germany, which is able to change into the "world’s strongest AI system" when it's put in in 2024. It makes use of a liquid-cooled structure, "with a booster module comprising near 24,000 NVIDIA GH200 Superchips interconnected with the NVIDIA Quantum-2 InfiniBand networking platform," based on NVIDIA.
NVIDIA says JUPITER will assist help scientific breakthroughs in numerous areas, together with local weather and climate prediction, producing high-resolution local weather and climate simulations with interactive visualization. It'll even be employed for drug discovery, quantum computing and industrial engineering. Many of those areas use customized NVIDIA software program options that ease growth but in addition make supercomputing teams reliant on NVIDIA {hardware}.
The brand new applied sciences will probably be key for NVIDIA, which now makes most of its income from the AI and knowledge middle segments. Final quarter the corporate noticed a document $10.32 billion in income in that space alone (out of $13.51 billion whole income), up 171 p.c from a yr in the past. It little question hopes the brand new GPU and superchip will assist proceed that development. Simply final week, NVIDIA broke its personal AI coaching benchmark document utilizing older H100 expertise, so its new tech ought to assist it prolong that lead over rivals in a sector it already dominates.
This text initially appeared on Engadget at https://www.engadget.com/nvidia-announces-its-next-generation-of-ai-supercomputer-chips-140004095.html?src=rss
Source link
