Loading the catalog…
NVIDIA A800 40GB ACTIVE (Amax)
Blurred values are subscriber-only. Sign in to start a free trial — 2 full pages a day — or subscribe for unlimited access.
| SKU | NVIDIA A800 40GB ACTIVE |
|---|
| Product Name |
|---|
| Tagline | |
|---|---|
| Architecture | |
| Tensor Core generation | |
| Key feature - FP64 | |
| Key feature - training/inference | |
| NVLink generation | |
| NVLink scaling | |
| NVLink bidirectional bandwidth |
| Memory type (key features) |
|---|
| Memory capacity (key features) |
|---|
| Memory bandwidth (key features) |
|---|
| Multi-Instance GPU (MIG) |
|---|
| MIG partitioning (key features) |
|---|
| Software subscription |
|---|
| GPU Memory |
|---|
| Memory Interface |
|---|
| Memory Bandwidth |
|---|
| CUDA Cores |
|---|
| Tensor Cores |
|---|
| Double-Precision Performance |
|---|
| Single-Precision Performance |
|---|
| Peak Tensor Performance |
|---|
| Multi-Instance GPU |
|---|
| NVIDIA NVLink |
|---|
| NVLink Bandwidth |
|---|
| Graphics Bus |
|---|
| Power Consumption |
|---|
| Thermal |
|---|
| Form Factor |
|---|
| Display Capability |
|---|
| Display ports footnote |
|---|
| Relative performance - AI Training (BERT-Large) |
|---|
| Relative performance - AI Inference (BERT-Large) |
|---|
| Relative performance - HPC (GTC) |
|---|
| Relative performance - HPC (LAMMPS) |
|---|
| Copyright |
|---|
| Document ID |
|---|
| Document date |
|---|
