Interactive 3D Model
Loading viewer...
Drag to rotate ┬╖ Scroll to zoom
ARCHITECTURE
Hopper
SERIES
H100
Memory
80GB HBM3
Bandwidth
3.35 TB/s
TDP
700W
In StockH100┬╖Best for: AI Training
NVIDIA H100 Tensor Core GPU
Hopper
|Data Center GPUsOverview
The NVIDIA H100 Tensor Core GPU delivers exceptional AI training and inference performance with its Hopper architecture. Featuring the Transformer Engine for LLM optimization and FP8 Tensor Cores for accelerated computing, the H100 is the gold standard for enterprise AI deployments.
Key Features
- Transformer Engine
- FP8 Tensor Cores
- NVLink 4.0
- MIG
Ideal Use Cases
ai traininghpc
Full Specifications
Complete technical specifications for the H100.
Memory
- Memory
- 80GB HBM3
- Memory Bandwidth
- 3.35 TB/s
Compute
- Tensor Cores
- 528
- CUDA Cores
- 18432
- FP8 TFLOPS
- 1979
- FP16 TFLOPS
- 989
- TF32 TFLOPS
- 494.5
- FP64 TFLOPS
- 60
Interconnect
- Interconnect
- NVLink 4.0, 900GB/s
Physical
- TDP
- 700W
- Form Factor
- SXM5
- Cooling
- Liquid Cooled
Manufacturing
- Launch Date
- Q1 2024
- Manufacturing Process
- 4nm