Inspired by the demands of deep learning and analytics, NVIDIA DGX™ Systems are the essential instruments for AI research built on the new NVIDIA Ampere Architecture.
The third generation DGX™ A100 provides universal infrastructure for AI, with unmatched capabilities for training, inference and big data analytics.
- 8 x NVIDIA A100 Tensor Core GPUs with 80GB, or 8 x NVIIDA A100 Tensor Core GPUs with 40GB for either 320GB or 640GB combined GPU memory
- 5 petaFLOPS of performance in 6RU
- System memory of either 1TB (320GB model) or 2TB (640GB model)
- 6x NVLink® Switches with 4.8TB/s bandwidth
- 8 x Mellanox ConnectX-6 200Gb/s NICs
The NVIDIA DGX Station™ A100 brings powerful updates to the popular DGX Station™ while retaining the plug and play, workstation form factor.
- 4 x NVIDIA A100 Tensor Core GPUs with 80GB, or 4 x NVIIDA A100 Tensor Core GPUs with 40GB
- 2.5 petaFLOPS AI, 5 petaOPS INT8
- Single AMD 7742, 64 Cores, 2.25GHz (base) – 3.4GHz (max boost)
- 512GB DDR4