H100

Rent NVIDIA H100 GPU

H100 GPU: 80 GB HBM3, Hopper architecture. Explore workload requirements and check rental availability in GlobalAtlas AI Console.

Choose a configuration in Console to review its resources and estimated cost.

What to plan for

The reference is H100 SXM 80 GB. H100 NVL uses a different memory configuration; an offer must identify the exact variant.

Size weights, runtime buffers and data separately. The host also needs CPU, RAM, disk and a suitable region. No rental price or availability is implied by this hardware specification.

What this GPU can be used for

AI training, inference and accelerated computing. Plan within the 80 GB memory budget after accounting for runtime overhead. The exact software stack determines which workload is suitable.

Popular deployment scenarios

Hardware specifications — not a performance benchmark
GPUVRAMMemoryArchitecturePrice
B200180 GBHBM3eBlackwellCheck price
H200141 GBHBM3eHopperCheck price
H10080 GBHBM3HopperCheck price
A10080 GBHBM2eAmpereCheck price

Check availability: H100

Check the current price and configuration in Console. Your choice is saved through registration. The current lookup uses mock adapters; no real capacity is reserved.

Before you deploy

Is this GPU available right now?

The hardware page is permanent. Console checks the current adapter offers; if none match, it shows alternatives and lets you retry.

Does VRAM determine application performance?

No. Model, precision, context, software and batching matter. Measure your workload before committing to a configuration.

Related products and guides

Specification sources

Hardware and model documentation; these sources do not confirm rental availability.

Official documentation 1 →