GPUs

Pick a GPU type, choose a size, then review the full commitment before renting or requesting.

GPU type

On-demandBy request

Preview rates are indicative until Pacific confirms your allocation. They are not a live inventory feed.

Preview catalog / Multiple regions / Dedicated nodes

Available now

Term

Sizes of H100 SXM
Size	Nodes	Per node	Availability	Hourly
8× H100 SXM	1 node	192 vCPU / 2 TB	Available	$23.92
16× H100 SXM	2 nodes	192 vCPU / 2 TB	Available	$47.84
32× H100 SXM	4 nodes	192 vCPU / 2 TB	Available	$95.68
64× H100 SXM	8 nodes	192 vCPU / 2 TB	By request	$191.36
128× H100 SXM	16 nodes	192 vCPU / 2 TB	By request	$382.72

Node specification & storage

Allocation: 8× H100 SXM / 1 node
Accelerator memory: 80 GB each
CPU / memory: 192 vCPU / 2 TB
Local storage: 16 TB NVMe
Interconnect: 3.2 Tb/s InfiniBand
Tenancy: Dedicated nodes
Access: SSH access to dedicated bare-metal Linux hosts with NVIDIA drivers and CUDA installed

Local NVMe is erased when the resource ends. Nothing is retained after the rental.

Availability recorded by Pacific 11 Sept 2026, 11:26 UTC. Confirmed when your order is accepted, not a live check.

Loading…