Compute · Rack-scale

NVIDIA GB300 NVL72

A fully liquid-cooled rack-scale system — 72 Blackwell Ultra GPUs and 36 Grace CPUs acting as one accelerator, for frontier-model training and agentic AI at the highest scale.

Specifications

One rack, one giant GPU

GPUs per rack72 × Blackwell Ultra
CPUs per rack36 × Grace
GPU interconnectNVLink 5, rack-scale fabric
Cluster interconnect800 Gb/s InfiniBand per GPU (Quantum-X800)
CoolingFully liquid-cooled
AvailabilityBy reservation — contact sales

Interconnect

The network is not the bottleneck

Each GPU gets 800 Gb/s of dedicated InfiniBand bandwidth, so all-reduce traffic across racks keeps pace with the NVLink fabric inside them. Distributed jobs scale the way the reference architecture says they should — we validate it per deployment.

High-speed fabric cabling

Suited workloads

Frontier-model pre-training Large-scale agentic AI Massive MoE serving Multi-trillion-token runs

Start your journey today

Tell us about your workload — an engineer, not a sales bot, will get back to you within one business day.