Statements (30)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:graphics_card
|
| gptkbp:announced |
2022
|
| gptkbp:architecture |
Data center GPU
|
| gptkbp:designedBy |
gptkb:NVIDIA
|
| https://www.w3.org/2000/01/rdf-schema#label |
Hopper GPU
|
| gptkbp:marketedAs |
Supercomputing
Cloud computing Enterprise |
| gptkbp:memoryBusWidth |
up to 3 TB/s (H100 SXM)
|
| gptkbp:namedAfter |
gptkb:Grace_Hopper
|
| gptkbp:notableProduct |
gptkb:NVIDIA_GH200_Grace_Hopper_Superchip
gptkb:NVIDIA_H100_PCIe gptkb:NVIDIA_H100_SXM |
| gptkbp:predecessor |
Ampere GPU
|
| gptkbp:processNode |
gptkb:TSMC_4N
|
| gptkbp:product |
gptkb:NVIDIA_H100
|
| gptkbp:releaseDate |
March 2022
|
| gptkbp:successor |
Blackwell GPU
|
| gptkbp:supports |
gptkb:NVLink
gptkb:HBM3_memory gptkb:PCIe_Gen5 gptkb:Tensor_Cores FP16 precision FP8 precision |
| gptkbp:transistorCount |
80 billion (H100)
|
| gptkbp:usedFor |
AI inference
AI training High Performance Computing |
| gptkbp:bfsParent |
gptkb:Nvidia_GH200_Grace_Hopper_Superchip
|
| gptkbp:bfsLayer |
7
|