H100 NVL
nvidia · hopper · 94 GB HBM3 · 400W TDP
VRAM
94 GB
BF16 TFLOPS
835
Bandwidth
3938 GB/s
From
$3.09/hr
Spec Sheet
Pricing by Provider
| Provider | On-Demand | Reserved | Spot | Badge |
|---|---|---|---|---|
| coreweave | $4.10/hr | $3.09/hr | - | Cheapest |
| aws | $5.60/hr | $4.20/hr | - |
Compatible Models (278)
Single GPU (228 models)
Multi-GPU (50 models)
Training Capabilities
Estimated GPU count for full fine-tuning (AdamW, BF16) and QLoRA
| Model Size | Full Fine-Tune | QLoRA |
|---|---|---|
| 7B model | 2 GPUs | 1 GPU |
| 13B model | 3 GPUs | 1 GPU |
| 70B model | 15 GPUs | 1 GPU |
Energy Efficiency
Estimated tokens/second per Watt for popular models
Similar GPUs
Methodology Note
Performance estimates for the H100 NVLare based on InferenceBench's roofline performance model with CUDA kernel-level optimization including FlashAttention v2 and PagedAttention. Memory calculations account for model weights (94 GB HBM3 available), KV-cache allocation, and activation memory. Throughput predictions use the H100 NVL's rated 3938 GB/s memory bandwidth and 835 BF16 TFLOPS compute capacity as roofline ceilings, with empirical correction factors per GPU architecture (hopper). See our full methodology.
Frequently Asked Questions
How many AI models can run on H100 NVL?
The H100 NVL can run 278 AI models from our database within a single node. Compatible models range across various parameter sizes depending on the quantization precision (BF16, FP8, INT4). Smaller models fit on a single GPU while larger models may require multi-GPU setups up to 8x H100 NVL.
What is the H100 NVL inference throughput?
The H100 NVL delivers 835 BF16 TFLOPS and 1671 FP8 TFLOPS with 3938 GB/s memory bandwidth. Actual inference throughput (tokens/sec) depends on the model size, precision, and batch size. Use our calculator for model-specific throughput estimates.
How much does H100 NVL cost per hour?
The H100 NVL is available starting from $3.09/hour via coreweave. Prices vary by provider and pricing tier (on-demand, reserved, spot). Compare pricing across all providers in the table above.