Huang: 'AGI has arrived'
Jensen Huang declared at GTC that 'AGI has arrived' — a claim framed around practical, enterprise‑grade AI orchestration rather than speculative science fiction. The line was used to emphasize reliability, scale, and business outcomes as the new yardsticks for AI. (startupnews.fyi)
Huang’s comments aired on Lex Fridman Podcast episode #494 released March 23, 2026, coming one week after his GTC keynote in San Jose on March 16, 2026. (podscan.fm) At GTC he unveiled a stack aimed at enterprise orchestration: Vera Rubin rack systems, the DSX “AI Factory” platform, OpenClaw/NemoClaw for long‑running agentic workloads, Groq 3 LPX inference LPUs and NVIDIA’s Nemotron open models. (youtube.com) NVIDIA quantified the opportunity by tying demand to specific architectures, forecasting at least $1 trillion in cumulative orders or AI‑chip revenue through 2027 driven by Blackwell and Rubin systems. (bloomberg.com) The company reported GTC 2026 drew more than 30,000 attendees from over 190 countries across March 16–19, 2026, a turnout NVIDIA used to showcase enterprise buyers and partners for its inference roadmap. (nvidianews.nvidia.com) Commentators noted Huang reframed “AGI” as operational agentic systems judged on reliability, scale, and business outcomes — a definition explored in UCStrategies’ analysis and echoed in coverage emphasizing an “inference inflection.” (ucstrategies.com) Mainstream outlets captured mixed reactions: The Verge reported Huang later walked back nuances of the claim, while Forbes and Mashable ran explanatory pieces tying his language to productization and investor expectations. (theverge.com) Huang’s roadmap stressed hardware‑software co‑design details — NVLink‑72 Rubin modules, BlueField‑4 DPUs, Spectrum‑X switches and co‑packaged optics — all positioned to boost tokens‑per‑second‑per‑watt for large, long‑running inference workloads. (blogs.nvidia.com)
