NVIDIA Vera Rubin NVL72

Manufacturer specification reference

Architecture

Vera CPU / Rubin GPUNVIDIA Vera Rubin NVL72 (external source) · NVIDIA Vera Rubin datasheet (external source) · From Training to Production, NVIDIA and CoreWeave Close the Loop on Agentic AI (external source) · Inside NVIDIA Rubin GPU Architecture (external source)

Category

Complete rack-scale referenceNVIDIA Vera Rubin NVL72 (external source) · NVIDIA Vera Rubin datasheet (external source) · From Training to Production, NVIDIA and CoreWeave Close the Loop on Agentic AI (external source) · Inside NVIDIA Rubin GPU Architecture (external source)

Accelerator memory

20,700 GBManufacturer-rounded 20.7 TB aggregate GPU HBM, converted to decimal GB; Vera CPU memory excluded.NVIDIA Vera Rubin NVL72 (external source)

Memory type

HBM4NVIDIA Vera Rubin NVL72 (external source) · NVIDIA Vera Rubin datasheet (external source) · From Training to Production, NVIDIA and CoreWeave Close the Loop on Agentic AI (external source) · Inside NVIDIA Rubin GPU Architecture (external source)

Memory bandwidth

1,400 TB/sPublished rack aggregate GPU memory bandwidth; not per-GPU or NVLink bandwidth.NVIDIA Vera Rubin NVL72 (external source)

Power

Not verified for this recordNo total measured rack input power verified. Power-shelf nameplate capacity and benchmark power assumptions must not be reported as operating rack draw.NVIDIA Vera Rubin NVL72 (external source) · NVIDIA Vera Rubin datasheet (external source) · From Training to Production, NVIDIA and CoreWeave Close the Loop on Agentic AI (external source) · Inside NVIDIA Rubin GPU Architecture (external source)

Precision formats

NVFP4, FP8, FP6, FP16, BF16, TF32, FP32, FP64NVIDIA Vera Rubin NVL72 (external source)

Cooling / integration

Liquid-cooled rack; the September-revision datasheet describes a fully liquid-cooled warm-water design.NVIDIA Vera Rubin NVL72 (external source) · NVIDIA Vera Rubin datasheet (external source) · From Training to Production, NVIDIA and CoreWeave Close the Loop on Agentic AI (external source) · Inside NVIDIA Rubin GPU Architecture (external source)

Configuration boundaries

• The verified rack is NVL72. NVL4 and HGX NVL8 are separate configurations and are not treated as the same rack.
• Current rack table lists 19.2 TB/s per GPU; the architecture article states up to 22 TB/s. This record uses the explicitly published rack aggregate, without blending these boundaries.
• Manufacturer sparse inference and dense training FLOPS must remain separate.

Published arithmetic throughput

Deployment considerations

• Use current rack-level documents rather than extrapolating from an early Rubin GPU architectural maximum.
• Verify facility supply, coolant quality and temperatures, commissioning criteria and software release support with the OEM.

Software ecosystem

NVIDIA AI Enterprise · NVIDIA NIM · NVIDIA Mission Control. Confirm the exact model, runtime release, server integration and vendor support matrix before evaluating a deployment.