posted by Grad about 2 months ago

$3.00 / gpu / hour - 64 H100s available from 04/01/2024 to 04/01/2025 (US, California)

Unverified
Interconnect: Infiniband 3.2 Tb
Cores per node: 64
RAM per node: 94 GB
NVME storage per node: 8000 GB
Location: US, California
Cloud provider: 10Gbit Ethernet
Cluster interface: K8s
Minimium bookable GPUs: 8
Minimium bookable duration: 12 weeks
Additional details
H100 servers physically located in the Bay Area. Interconnected with 3.2tbps Infiniband, 1tb RAM, AMD EPYC 32 core CPUs, and 8TB of NVME storage per server.
contact seller

seller email domain is gradient.ai