Data center · accelerator
NVIDIA A40
An Ampere server GPU with 48 GB GDDR6 ECC, RT cores and Tensor cores for graphics and AI. It supports two-GPU NVLink configurations and virtual workstation deployments.
Technical specifications
| Architecture | NVIDIA Ampere |
|---|---|
| GPU memory | 48 GB GDDR6 ECC |
| Memory bandwidth | 696 GB/s |
| Host interface | PCIe Gen4 |
| Maximum board power | 300 W |
| Cooling / size | Passive; 4.4 x 10.5 inches, dual-slot |
| Power connector | 8-pin CPU |
| NVLink | 112.5 GB/s bidirectional; 2 GPUs |
| Display outputs | 3 x DisplayPort 1.4, disabled by default |
Applications
- Ray-traced rendering, 3D visualization and simulation.
- Training and fine-tuning neural networks.
- LLM and other neural-network inference.
- Video encoding, decoding and AI video processing.
Compatibility
- Verify the exact server model, slot clearance, power cabling and cooling with the OEM. A PCIe slot alone does not establish compatibility. Passive cooling requires forced airflow from the server.
- The 8-pin CPU power connector must not be treated as an interchangeable PCIe 8-pin connector. Display outputs are disabled in the default virtualization configuration and need enabling with management tools.
- Memory belongs to individual GPUs. NVLink does not automatically turn it into one GPU's local memory; using multiple GPUs requires software that distributes the model or workload.
- Check driver, CUDA/framework and operating-system support for the intended workload. Virtualization and enterprise software may require separate licenses and a supported host configuration.
Manufacturer sources
Specifications checked against the listed NVIDIA sources: .
Request a quote
Each quote is prepared individually for the selected variant, quantity and destination. The RFQ covers condition, configuration, source, warranty, availability and delivery timing.