NVIDIA's Robotics & AI Infrastructure Blitz at GTC Berlin
NVIDIA announced a sweeping set of AI and robotics initiatives at GTC Berlin, including a modular AI factory (DSX POD) that slashes deployment from 6 months to 14 days, major humanoid robotics platform expansions, and a 300%+ packaging capacity deal with AMKOR.
What we know
Jensen Huang revealed the NVIDIA DSX POD, a modular AI factory built on the Vera Rubin platform that cuts time-to-deployment from 6 months to as little as 14 days, and shrinks token production startup from 4–5 months to roughly 1 month. A 1-gigawatt AI data center costs $50B and generates $40B/year in revenue. On robotics, NVIDIA expanded its Isaac platform with new humanoid robot hardware, AI training datasets, and simulation tools — Huang declared the 'ChatGPT moment for robotics is fast approaching.' NVIDIA also signed agreements with AMKOR for 300%+ expansion in advanced packaging capacity for next-gen AI platforms.
By the Global Eye 24 desk, with AI assistance. Updated as the story develops. Spot an error? Write to corrections@globaleye24.com.