Nvidia CTO Spotlights Nebius for Fast Node Repair, European GB300 NVL72 Deployment, and Serverless Push

Nvidia ($NVDA) Chief Technology Officer Michael Kagan spotlighted AI cloud provider Nebius ($NBIS), noting that the company's infrastructure developments are "worth paying attention to." Kagan pointed to Nebius's 15-minute automatic node repair capabilities, its distinction as the first operator in Europe with running GB300 NVL72 systems, and its early-access roadmap for Nvidia's next-generation Vera Rubin architecture.
The endorsement highlights the growing strategic importance of specialized "neocloud" providers to Nvidia. As hyperscale cloud giants—including Meta ($META), Amazon ($AMZN), Microsoft ($MSFT), and Alphabet ($GOOGL)—accelerate internal development of proprietary custom silicon, dedicated AI infrastructure providers purchasing across the full Nvidia hardware and networking stack represent an increasingly vital commercial base.
Kagan also emphasized Nebius's "genuine bet on serverless inference," underscoring how the operator is transitioning beyond bare-metal GPU capacity rental toward higher-margin, software-integrated platform services designed to improve customer retention.