Intel Accelerator
INTEL ACCELERATOR
NVIDIA · Rubin

NVIDIA Vera Rubin NVL72

Next-generation Vera Rubin rack-scale platform. Facility power figures must come from the current official NVIDIA specification — none is applied by default.

Demo data

Rack-scale specification

GPU count
72
GPU
NVIDIA Rubin GPU
CPU count
36
CPU
NVIDIA Vera CPU
Cooling architecture
Liquid cooled (rack-scale)
Rack architecture
NVL72 — 72-GPU NVLink domain, single rack
Networking family
NVIDIA scale-up NVLink + scale-out InfiniBand / Spectrum-X (confirm generation)
Last updated
2026-08

Cooling options

  • Direct liquid coolingREFERENCE ARCHITECTURE

    Rack-scale liquid cooling is the stated architecture.

  • Hybrid liquid + airENGINEERING REVIEW REQUIRED

    Residual air load cannot be derived until the official facility specification is on file.

  • Air coolingNOT APPROPRIATE BASED ON AVAILABLE DATA

    No air-cooled facility configuration documented.

Approximate module sizing

ModuleDesign rack powerApprox. racksBasis
2.5 MWSOURCE REQUIREDOfficial value required
5 MWSOURCE REQUIREDOfficial value required
10 MWSOURCE REQUIREDOfficial value required
20 MWSOURCE REQUIREDOfficial value required

Facility power requirement

FACILITY POWER REQUIREMENT — SOURCE REQUIRED
NOT AVAILABLE · SOURCE REQUIRED
Source: No current official facility-power figure on file
Source date:

No value is assumed. Upload an official manufacturer document in the RFP workspace and the extracted value, with its source, is applied to your project only.

What this determines

  • Rack quantity and power distribution cannot be calculated until an official rack power value is supplied.
  • Upload an official manufacturer document to apply a sourced value to this project.

Take action

INTEL ACCELERATOR. All figures shown are demo data for structural demonstration. Nothing is presented as verified market intelligence.