The NVIDIA® A40 accelerates the most demanding visual computing workloads from the data center, combining the latest NVIDIA Ampere architecture RT Cores, Tensor Cores, and CUDA® Cores with 48 GB of graphics memory. From powerful virtual workstations accessible from anywhere to dedicated render nodes, NVIDIA A40 brings nextgeneration NVIDIA RTX™ technology to the data center for the most advanced professional visualization workloads.
| Architecture | Ampere |
| Process Size | 8nm |
| Transistors | 28.3 Billion |
| Die Size | 628.4 mm |
| CUDA Cores | 10752 |
| RT Cores | 84 |
| Gen 2 | 336 |
| Tensor Cores | 2-Way Low Profile, 2-Slot |
| Gen 3 | 112 GB/s Bidirectional |
| NVLink | 48 GB GDDR6 ECC |
| NVLink Interconnect | 384-bit |
| GPU Memory | 696 GB/sec |
| Memory Interface | 3x DisplayPort 1.4, Off by Default |
| Memory Bandwidth | Passive |
| Display Connectors | NVIDIA® Virtual PC, NVIDIA RTX Virtual Workstation (vWS), NVIDIA Virtual Compute Server (vCS, no MIG) |
| Thermal Solution | 1 GB, 2 GB, 3 GB, 4 GB, 6 GB, 8 GB, 12 GB, 16 GB, 24 GB, 48 GB |
| vGPU Support | PCIE Express 4.0 x16 |
| vGPU Profiles Supported | CEC 1712 |
| System Interface | Level 3 |
| Secure and Measured Boot Hardware Root of Trust | 8-pin CPU |
| NEBS Ready | 300 W |
| Power Connector | |
| Maximum Power Consumption |