NVIDIA contributes Vera Rubin rack designs to open MGX specification
NVIDIA submitted Vera Rubin rack and compute-tray designs to the Open Compute Project through its open MGX architecture for AI factories.
Event / data period: 2026-10-12
Published: · Updated:
The updated MGX 1.2 specification covers blind-mate liquid cooling, interoperable rack interconnects and a liquid-cooled 5,000-amp busbar. NVIDIA says the design supports rack deployments of up to 250 kilowatts.
The company said it was contributing an open design for Vera Rubin computing racks as part of the third-generation MGX effort in OCP’s Open DC Initiative.
Data
Scope of the announced design
NVIDIA says the modular rack includes configurable I/O bays, fanless liquid-cooled trays and standardized power and cooling interfaces. The specification is an industry design contribution; it does not by itself indicate shipments or customer deployment.
| Maximum announced rack power | 0.25 |
Methodology
Summarizes features of NVIDIA’s design contribution to OCP; it is not the OCP technical specification or an independent performance test. The source gives an upper design limit of 250 kW; the table converts it using 1 MW = 1,000 kW, so 250 ÷ 1,000 = 0.25 MW.
Updated:
Subjects and places
Subject proposals
- Artificial intelligenceSuggested
- Subject suggested from the title or summary: AIblogs.nvidia.com ↗
Geographic references
Only source-provided geographic references are shown. Opening an external map is optional. Optional story maps use OpenStreetMap. Opening a map sends your IP address and this site’s address to the map provider; location data remains readable as text without loading the map.
- United States · Place role: Coverage · Place type: Country or area · Location precision: RegionalEvidence
- Explicit preserved editorial country/region; no city or national centroid inferredblogs.nvidia.com ↗