Statements (25)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
GPU server platform
|
| gptkbp:architecture |
gptkb:Ampere
|
| gptkbp:coreCount |
Yes
|
| gptkbp:formFactor |
Server board
|
| https://www.w3.org/2000/01/rdf-schema#label |
NVIDIA HGX A100
|
| gptkbp:interconnectBandwidth |
600 GB/s (NVSwitch)
|
| gptkbp:manufacturer |
gptkb:NVIDIA
|
| gptkbp:maxGPUs |
8
|
| gptkbp:maxMemoryPerGPU |
80GB
|
| gptkbp:memoryType |
gptkb:HBM2
|
| gptkbp:NVLinkSupport |
Yes
|
| gptkbp:NVSwitchSupport |
Yes
|
| gptkbp:powerSource |
Up to 400W per GPU
|
| gptkbp:releaseYear |
2020
|
| gptkbp:supportedBy |
gptkb:NVIDIA_A100
gptkb:NVIDIA_A40 |
| gptkbp:supportsMultiGPU |
Yes
|
| gptkbp:supportsPCIeVersion |
Yes
|
| gptkbp:targetMarket |
gptkb:artificial_intelligence
gptkb:cloud_service gptkb:HPC |
| gptkbp:usedIn |
gptkb:NVIDIA_DGX_A100
OEM server systems |
| gptkbp:bfsParent |
gptkb:NVIDIA_A100
|
| gptkbp:bfsLayer |
7
|