G6401
8-GPU AI Training Server
8-GPU 2-AMD CPU air-cooled training platform for LLM and vision workloads.
- 8x NVIDIA Blackwell GPUs
- GPU
- 24x DDR5
- MEMORY
- 8 x 419 TFLOPS FP16 Tensor
- COMPUTE
| GPU | 8x NVIDIA Blackwell GPUs |
|---|---|
| MEMORY | 24x DDR5 |
| COMPUTE | 8 x 419 TFLOPS FP16 Tensor |
8-GPU air-cooled AI training server with 24x DDR5 memory and PCIe 5.0 host interfaces, delivering 104.8 TFLOPS FP16 dense compute per node.
8x NVIDIA Blackwell GPUs
Dense FP16 training throughput per rack.
24x DDR5
Memory bandwidth and host I/O for large model checkpoints.
Air-cooled 6U
Standard datacenter air handling, no liquid loop required.