Overview
High-performance 2u rackmount server with 8x NVIDIA H200 SXM GPUs connected via NVLink. Optimized for large-scale LLM training and inference workloads requiring maximum GPU memory bandwidth.
Key Features
- 8x NVIDIA H200 SXM GPUs with NVLink
- High-bandwidth HBM3/HBM3e memory for LLM workloads
- NVSwitch fabric for all-to-all GPU communication
- Dual-socket CPU platform
- Enterprise redundant power and advanced cooling
Ideal For
AI teams requiring enterprise-grade infrastructure for LLM training, fine-tuning, and high-throughput inference.
Feature
900GB/s GPU-to-GPU bandwidth with NVIDIA NVLink⢠and NVSwitchâ˘
Feature
2 x 10Gb/s LAN ports via IntelÂŽ X710-AT2
Feature
4+2 3000W 80 PLUS Titanium redundant power supplies
$339,000.00
Prices may vary. Verify on vendor site.
Quick Specs
- Feature
- 4+2 3000W 80 PLUS Titanium redundant power supplies
Tags
llm-trainingllm-inferencegenerative-ai