Compute · Large-memory Hopper
NVIDIA HGX H200
141 GB of HBM3e per GPU — serve and train large models without quantization compromises, on a mature Hopper platform.
Specifications
Memory first
| GPUs per node | 8 × H200 (SXM) |
| GPU memory | 141 GB HBM3e per GPU |
| Memory bandwidth | 4.8 TB/s per GPU |
| Node interconnect | NVLink 4, NVSwitch |
| Cluster interconnect | 400 Gb/s InfiniBand per GPU (Quantum-2) |
| Availability | On-demand & reserved |
Interconnect
400 Gb/s per GPU, non-blocking
Quantum-2 InfiniBand with GPUDirect RDMA keeps multi-node H200 clusters scaling linearly — the network never becomes the distributed-training bottleneck.
Suited workloads
Large-context inference
LLM training & fine-tuning
Memory-bound models
RAG at scale
Start your journey today
Tell us about your workload — an engineer, not a sales bot, will get back to you within one business day.