Supermicro expands AI product line by supporting NVIDIA H200

(Global TMT November 14, 2023) Supermicro, Inc., a provider of overall IT solutions for artificial intelligence, cloud technology, storage and 5G/edge, provides support for the new NVIDIA HGX H200 with built-in H200 Tensor Core GPU , to further expand its artificial intelligence product line. Supermicro covers 8U and 4U general-purpose GPU systems, ready to apply HGX H200 8-GPU and 4-GPU. It has HBM3e memory with nearly 2x the capacity and 1.4x higher bandwidth compared to the NVIDIA H100 Tensor Core GPU. In addition, Supermicro NVIDIA MGX systems have the richest product portfolio and are compatible with the upcoming NVIDIA Grace Hopper super chip equipped with HBM3e memory.

Supermicro 4U 8GPU H200

Utilizing the latest liquid cooling technology, Supermicro has launched a liquid-cooled 4U server based on the NVIDIA HGX H100 8-GPU system, achieving the highest server density in the industry. As the industry's most compact high-performance GPU server, it can help data centers reduce floor space requirements, save energy costs, and provide top-notch single-rack high-performance artificial intelligence training capabilities. By employing the highest density GPU systems and advanced liquid cooling solutions, enterprises can reduce total cost of ownership

The above is the detailed content of Supermicro expands AI product line by supporting NVIDIA H200. For more information, please follow other related articles on the PHP Chinese website!