Specifications
GPUs
8× NVIDIA H200
GPU Memory
1,128GB HBM3e total
NVSwitch
4×
Networking
ConnectX-7 400Gb/s IB
CPU
2× Intel Xeon
System Memory
2TB DDR5
DPU
BlueField-3
Application
NVIDIA DGX H200 delivers 8× H200 Hopper GPUs with 1,128GB of fast HBM3e for LLM training and inference at scale. QS Compute supplies the turnkey system with NVLink fabric, storage, and rack integration.
Request a Quote — NVIDIA DGX H200
QS Compute — Global B2B AI computing hardware supply, 15-day sample lead time.
Get Your Quote →