NVIDIA · Hopper · 2022
NVIDIA H100. 80 GB HBM3, up to 989 TFLOPS FP16 tensor.
The Hopper workhorse for production training and inference: 80 GB of HBM3, up to 8 GPUs in one node.
H100 on demand
from$4.29/hr per GPU
- Availability
- Available now · 4 regions
- By region and size
- $4.29 to $7.79/hr
- Billing
- By the second on most GPUs, a few by the minute
- 8+ GPUs or a month
- Launch any size yourself, and prepay 1 to 12 months when you launch an instance. Custom quotes on request.
The machine
80 GB of HBM3. At up to 3.4 TB/s.
- Memory
- 80 GB
- HBM3
- Bandwidth
- up to 3.4 TB/s
- Memory to compute
- FP16 tensor
- up to 989
- TFLOPS
- Per node
- 8 GPUs
- up to 700 W each
Workloads
What teams run on the H100. Each one opens the launch form, sized.
Full specs
NVIDIA H100 specifications.
Memory
- Memory
- 80 GB
- Type
- HBM3
- Bandwidth
- up to 3.4 TB/s
Compute
- FP32
- up to 67 TFLOPS
- FP16 tensor
- up to 989 TFLOPS
- INT8
- up to 2400 TOPS
- Tensor cores
- up to 528
- CUDA cores
- up to 16,896
Power and form
- TDP
- up to 700 W
- Architecture
- Hopper
- Released
- 2022
Node
- GPUs per node
- 8
- Maker
- NVIDIA
Alternatives
Alternatives to the H100. Same account, different price and performance.
| GPU | Memory · FP16 tensor | Stands out | Price gap | From |
|---|---|---|---|---|
| NVIDIA H100This GPU | 80 GB HBM3 · up to 989 TFLOPS | $4.29/hr | ||
| NVIDIA H200 | 141 GB HBM3e · 989 TFLOPS | +61 GB memory | +8% | $4.63/hr |
| AMD MI300X | 192 GB HBM3 · 1307.4 TFLOPS | +112 GB memory | Reserve | |
| NVIDIA L40S | 48 GB GDDR6 · 362 TFLOPS | -32 GB memory | -73% | $1.14/hr |
| NVIDIA A100 80GB | 80 GB HBM2e · 312 TFLOPS | -68% FP16 tensor | -60% | $1.73/hr |
Questions
Renting the H100. What buyers ask first.
Other GPUs
Browse every GPU by architecture.
NVIDIA Blackwell
NVIDIA Hopper
NVIDIA Ampere
NVIDIA Ada Lovelace
NVIDIA Volta