Nvidia has announced that its GH200 Grace Hopper Superchip is now in full production and will power systems that run complex AI programs. The Nvidia Grace, Nvidia Hopper, and Nvidia Ada Lovelace architectures, among others, are expected to help meet the increasing demand for generative AI. The GH200-powered systems will be made available to global hyperscalers and supercomputing centres in Europe and the US. Over 400 configurations based on Nvidia’s latest CPU and GPU architectures have been created to enable enterprises to build and deploy generative AI applications that leverage their unique proprietary data.
Nvidia CEO Jensen Huang recently revealed new systems, partners, and additional details surrounding the GH200 Grace Hopper Superchip, which is manufactured by combining the Arm-based Nvidia Grace CPU and Hopper GPU architectures. The chip uses Nvidia NVLink-C2C interconnect technology, delivering up to 900GB/s total bandwidth, which is seven times higher than standard PCIe Gen5 lanes found in traditional accelerated systems.
Nvidia AI Enterprise, the software layer of the Nvidia AI platform, offers more than 100 frameworks, pretrained models and development tools to streamline development and deployment of production AI, including generative AI, computer vision, and speech AI. Nvidia GH200-powered systems are expected to be available later this year and Nvidia DGX GH200 supercomputers by the end of the year.
Nvidia is an American multinational corporation that designs graphics processing units for gaming and professional markets, as well as system on a chip units for the mobile computing and automotive market.
Jensen Huang is a Taiwanese-American billionaire businessman; he is the co-founder and CEO of Nvidia, a multinational technology company.
