Data center · SXM4
NVIDIA A100 SXM4 80 GB
An 80 GB Ampere SXM4 module for HGX-based AI and HPC systems. It uses HBM2e memory and baseboard NVLink connections rather than a PCIe card's bridge layout.
Technical specifications
| Architecture | NVIDIA Ampere |
|---|---|
| GPU memory | 80 GB HBM2e |
| Memory bandwidth | 2039 GB/s |
| Module format | SXM4 |
| NVLink bandwidth per GPU | 600 GB/s |
| Standard TDP | 400 W |
| MIG | Up to 7 instances |
Applications
- Training and fine-tuning neural networks.
- LLM and other neural-network inference.
- Scientific simulations and GPU-accelerated data analysis.
Compatibility
- This is an SXM module, not a PCIe expansion card. It needs a matching GPU baseboard, firmware, power delivery and cooling validated by the server manufacturer. SXM generations are not interchangeable.
- 400 W applies to the standard configuration. NVIDIA lists up to 500 W for the HGX A100 80 GB custom thermal solution (CTS) SKU; do not assume that limit for every SXM4 system.
- 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.