The GPX QN6-42E2-8NVLINK GPU-accelerated server is a 4U, 8-GPU system designed to support highly parallel workloads. This server provides extreme GPU density, ideal for rack-scale, highly parallel workloads. A mix of NVMe and SSD storage devices provide balanced on-board data storage that avoids bottlenecks. Users can expect strong performance from this system in HPC and AI applications.
With support for 8 HGX A100 80GB GPUs, AMD® EPYC® 7003 series Processors, 4TB of DRAM, and ten hot-swappable NVMe/SAS/SATA drives, the GPX QN6-42E2-8NVLINK is the ideal high-performance GPU-accelerated server for highly parallel HPC and AI workloads.
Ideal GPU Optimized Server Use Cases
- AI / Deep Learning Training
- AI Inference
- High-Performance Computing