NVIDIA DGX™ Vera Rubin NVL72

The foundation for gigascale AI training and inference.

End-to-End Deployment Services and Support from AMAX, an NVIDIA DGX Elite Partner.

Request a Quote

NVIDIA DGX Vera Rubin NVL72

NVIDIA DGX™ Vera Rubin NVL72

The foundation for gigascale AI training and inference. NVIDIA DGX™ Vera Rubin NVL72 is a turnkey AI infrastructure platform designed to provide leading edge performance with significantly greater energy efficiency. It offers industry-leading performance per watt and tokens per watt across any AI workload, enabling enterprises to scale intelligence while improving the economics of token generation.

Download Datasheet Request a Quote

DGX Vera Rubin NVL72 Technical Specifications

Specification Details
GPU and CPU
  • 72x NVIDIA Rubin GPUs and 36x NVIDIA Vera CPUs
Total GPU Memory | Bandwidth
  • 20.7 TB HBM4 | Up to 1,580 TB/s
Total Fast Memory
  • 75 TB
Performance
  • NVFP4 Inference: 3,600 PFLOPS
  • NVFP4 Training: 2,520 PFLOPS*
  • FP8/FP6 Training: 1,260 PFLOPS*
NVIDIA NVLink Switch Systems
  • 9x L1 NVIDIA NVLink Switches
NVIDIA NVLink Bandwidth
  • 260 TB/s
Networking
  • 144x OSFP single-port NVIDIA® ConnectX®-9 VPI with 800 Gb/s NVIDIA InfiniBand and Ethernet
  • 18x dual-port NVIDIA® BlueField®-4 VPI with 400 Gb/s NVIDIA InfiniBand and Ethernet
Management Network
  • Host baseboard management controlled (BMC) with RJ45
Storage
  • E1.S
Software
  • NVIDIA Mission Control
  • NVIDIA AI Enterprise
  • NVIDIA DGX OS
  • Supports Ubuntu
Rack Units
  • 48 RU
Enterprise Support
  • Three-year enterprise business-standard support for hardware and software
*Dense specification. Specifications subject to change.
Scroll to Top