Nvidia CEO Jensen Huang Outlines Transition to AI Factories and New Vera Rubin Hardware
At the GTC Conference in Berlin, Nvidia CEO Jensen Huang detailed the company's strategic shift toward 'AI factories,' centered on new comprehensive server racks. These systems integrate 72 next-generation Vera Rubin GPUs with 256 custom language processing units and the company's own Vera family CPUs. This integrated hardware approach aims to capture more value from AI inference, which is essential for running large-scale models efficiently. Nvidia expects significant growth in monetization per unit of power, with sales opportunities per gigawatt of compute projected to rise from $18 billion with Hopper to $40 billion with the Vera Rubin family. Huang emphasized the high stakes of this infrastructure build-out, noting that a 1-gigawatt AI factory could generate approximately $40 billion in annual revenue, making rapid deployment critical for operators to avoid massive potential losses.
Tickers: $NVDA
Topics: Artificial Intelligence Data Centers GPU Product Launch
Read the full article at Motley Fool
Part of the story U.S. Intelligence Investigated Potential Chinese Espionage Plot Targeting Nvidia CEO (3 articles)