NVIDIA H100 NVL 94GB
Performance
Peak theoretical throughput by precision type
| Precision | Dense | 2:4 Sparse |
|---|---|---|
FP64 64-bit floating point | 30TFLOPS | Structured sparsity is a tensor-core feature; this vector precision has no sparse form |
FP64 Tensor 64-bit floating point with Tensor Cores | 60TFLOPS | Structured sparsity is a tensor-core feature; this vector precision has no sparse form |
FP32 32-bit floating point | 60TFLOPS | Structured sparsity is a tensor-core feature; this vector precision has no sparse form |
TF32 TensorFloat-32 | The vendor publishes only the 2:4 sparse figure for this precision | 835TFLOPS |
BF16 Brain Float 16 | The vendor publishes only the 2:4 sparse figure for this precision | 1,671TFLOPS |
FP16 16-bit floating point | The vendor publishes only the 2:4 sparse figure for this precision | 1,671TFLOPS |
FP8 8-bit floating point | The vendor publishes only the 2:4 sparse figure for this precision | 3,341TFLOPS |
FP6 | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
FP4 | No verified data available | The part supports 2:4 sparsity, but the vendor publishes no sparse figure for this precision |
INT8 8-bit integer | The vendor publishes only the 2:4 sparse figure for this precision | 3,341TOPS |
Dense peak divided by accelerator TDP. Board power only: excludes host CPUs, networking, cooling and facility overhead. TDP for this part: 400 W.
Specifications
Architecture
Hopper
Form Factor
PCIe
Launch Year
2023
Process Node
4nm
Memory
94 GB HBM3
Bandwidth
3,900 GB/s
TDP
400 W
Max power (Flopper estimate)
~460 W est. Flopper estimate: 400 W TDP x 1.15. The vendor publishes no maximum board power for this part.
Spec Confidence
Official
Full Specifications
| Memory | |
|---|---|
| Memory | 94 GB |
| Memory Type | HBM3 |
| Bandwidth | 3.9 TB/s |
| Interface Width | No verified data available |
| Power & Thermal | |
| TDP | 400 W |
| Max power (Flopper estimate) | ~460 W est. |
| Enterprise Features | |
| Sparsity | Yes |
| General | |
| Form Factor | PCIe |
| Architecture | Hopper |
| Process Node | 4nm |
| Launch Year | 2023 |
Datasheet & Resources
Flopper Datasheet
NVIDIA H100 NVL 94GB specifications, generated from our database. Printable.
Vendor product page
NVIDIA documentation for the H100 NVL 94GB
NVIDIA Developer Documentation
Technical specs, programming guides
Data Provenance
Every figure traced to a source
Primary Source
- Document
- —
- Publisher
- NVIDIA
- Published
- No verified data available
Data Quality
- Spec confidence
- Official
- Clock basis
- Boost
- Core precisions with figures
- 7 of 9
- Normalization
- All values in TFLOPS
Rent this GPU
$2.59 / GPU-hr
Lowest of 6 live listings across 3 providers· median $5.08
Median of each provider's cheapest on-demand listing, so a provider with many configurations counts once.
Price history and every provider on the pricing page. Some links are affiliate links.
Similar GPUs
Frequently Asked Questions
How many TFLOPS does the NVIDIA H100 have?
The NVIDIA H100 delivers 60 TFLOPS FP32, 1,671 TFLOPS FP16 (2:4 sparse) and 3,341 TFLOPS FP8 (2:4 sparse) at peak.
What is the power consumption of the NVIDIA H100?
The NVIDIA H100 has a TDP (Thermal Design Power) rating of 400 watts.
How much memory does the NVIDIA H100 have?
The NVIDIA H100 is equipped with 94 GB of memory with 3,900 GB/s of memory bandwidth.
What architecture is the NVIDIA H100 based on?
The NVIDIA H100 is based on the Hopper architecture, launched in 2023.
Stay Updated on GPU Releases
Get notified when new GPUs are added or specifications are updated.
No spam, unsubscribe anytime.