Chinese researchers cut GPU wait times with pooling tool to boost campus cluster efficiency
Dubbed gPooling, the pooling solution was able to improve utilization by up to 2x on a cluster powered by Nvidia A100s
News - A100
The latest news on networking, cloud, and AI infrastructure
-
UK's ARIA seeks ideas to cut AI compute costs by 1,000x
Research project backed by £42m government pot
-
Gcore launches generative AI cluster with Nvidia GPUs
Cluster-based on Nvidia A100 and H100 Tensor Core GPUs
-
Petrobras deploys first AI supercomputer
Tatu system features 224 GPUs from Nvidia
-
Nvidia creates pared back H100 GPU for export to China, called H800
Worse data transfer rates to get past sanctions
-
AMD EPYC, VMware power dense new Oracle Cloud instances
Announcement underscores the diverse array of processor architectures on which OCI runs and the rapid rise of AMD's EPYC data center chips