Canada (en)
Canada (fr)
EU / EFTA
United Kingdom
France
India
Quantity must be less than 1,000,000. If you’d like to place a larger order, please reach out to your sales team.
This product has been successfully added to "My Favorites".
This product could not be added to "My Favorites" or is already in "My Favorites".
Add to My Favorites
NVIDIA A100 HGX x4 GPU Air Cooled Baseboard
Mfr Part #:
Copied To Clipboard
Click To Copy
SHI Part #:
Copied To Clipboard
Click To Copy
|
Quick Details
|
|---|
|
NVIDIA A100 HGX x4 GPU Air Cooled Baseboard
- GPU computing processor
- A100 Tensor Core
- 40 GB HBM2
- PCIe 4.0 x16
- for ProLiant XL645d Gen10 Plus, XL645d Gen10 Plus PCIe Accelerator Tray
| Additional Details | |
|---|---|
| Mfr Part #: | R3V35A |
| SHI Part #: | 41944739 |
| Category: | Graphic accelerator cards |
| UNSPSC: | 43201401 |
| Manufacturer: | Hewlett Packard Enterprise |
| Product Type: | Hardware |
| General Information | |
|---|---|
| Category | Graphic accelerator cards |
| Description | NVIDIA A100 HGX x4 GPU Air Cooled Baseboard - GPU computing processor - A100 Tensor Core - 40 GB HBM2 - PCIe 4.0 x16 - for ProLiant XL645d Gen10 Plus, XL645d Gen10 Plus PCIe Accelerator Tray |
| Manufacturer | Hewlett Packard Enterprise |
| UNSPSC | 43201401 |
| Main Specifications | |
| A/V Interface Type | PCI Express 4.0 x16 |
| Header / Brand | NVIDIA |
| Packaged Quantity | 1 |
| Video Memory / Technology | HBM2 |
| Video Output / Graphics Processor | NVIDIA A100 Tensor Core |
| Video Output / Graphics Processor Manufacturer | NVIDIA |
| Compatibility Information | |
| Designed For | HPE ProLiant XL645d Gen10 Plus, XL645d Gen10 Plus PCIe Accelerator Tray |
| General | |
| API Supported | OpenCL, DirectCompute, OpenACC |
| Bus Type | PCI Express 4.0 x16 |
| CUDA Cores | 6912 |
| Device Type | GPU computing processor |
| Features | NVIDIA Tensor Core, Multi-Instance GPU (MIG) technology, NVIDIA Ampere GPU technology, Error Correcting Codes (ECC) Memory, 9.7 Tflops peak double-precision floating point performance, 19.5 Tflops peak single-precision floating point performance, 312 Tflops peak half-precision floating point performance, 432 third-generation Tensor Cores per GPU, dual slot width, Nvidia CUDA technology, full-height full-length (FHFL) |
| Graphics Engine | NVIDIA A100 Tensor Core |
| Memory | |
| Size | 40 GB |
| Technology | HBM2 |
| Miscellaneous | |
| Power Consumption Operational | 250 Watt |
These links will open documents stored on other websites. SHI does not own this content.