Datacenter
AI Servers
High-performance GPU servers for AI workloads.
POWERED BY NVIDIA
Train LLMs faster than ever before.
Purpose-built clusters featuring up to 8x NVIDIA H100 Tensor Core GPUs per node, fully interconnected with NVSwitch and 400Gbps InfiniBand network fabrics.
3.2 Tbps
Bisection Bandwidth
32 petaFLOPS
FP8 Deep Learning
Most Popular
OLADC AI1
Optimized for distributed training
- CPUEPYC 4585PX
- RAM192GB
- Storage2x4TB + 4 GPU