NVIDIA's official blog post on June 22, 2026, announces that the upcoming Rubin AI infrastructure generation is the...

NVIDIA's official blog post on June 22, 2026, announces that the upcoming Rubin AI infrastructure generation is the world's first to achieve 100% liquid cooling, with coolant running at 45°C and zero water consumption through a closed-loop dry-cooler design source 6. This eliminates fans and reduces cooling energy consumption dramatically—historically up to 40% of a data center's electricity. The blog cites industry estimates that a 50 MW hyperscale facility can save over $4 million annually in cooling costs by adopting such architecture. Motivair (Schneider Electric) CEO Richard Whitmore confirms the partnership and necessity of liquid cooling as chip power densities increased. Separately, NVIDIA's SEC 10-Q filed May 20, 2026, shows revenue of $44B and net income of $18.8B for the quarter, reflecting continued strong demand source 5. While the CUTLASS software releases (sources 1-4) are routine, the liquid cooling breakthrough signals a major advance in AI infrastructure efficiency, likely accelerating adoption by hyperscalers. The low Manifold probability (8%) on Vera Rubin shipping by end of Q2 suggests the announcement is pre-shipment, but the design win with ecosystem partners points to a positive trajectory for NVIDIA's next-generation platform.