Supercomputing '23: NVIDIA High-Performance Chips Power AI Workloads
Briefly

NVIDIA announced expanded availability of the NVIDIA GH200 Grace Hopper Superchip for high-performance computing and HGX H200 Systems and Cloud Instances for AI training.
NVIDIA HGX H200 features the following: NVIDIA H200 Tensor Core GPU for generative AI and high-performance computing workloads that require massive amounts of memory.
Amazon Web Services, Google Cloud, Microsoft Azure, CoreWeave, Lambda, Vultr and Oracle Cloud Infrastructure will offer H200-based instances in 2024.
Read at TechRepublic
[
add
]
[
|
|
]