Nvidia GTC 2026 Keynote: Inference Chip, NemoClaw Rumored
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Jensen Huang's Nvidia GTC 2026 keynote is scheduled for 11 a.m. PT Monday at the SAP Center in San Jose, with the four-day conference running March 16–19 and the livestream available on the event's website
- NemoClaw, Nvidia's rumored open-source platform for enterprise AI agents, would let businesses build and deploy AI agents that carry out multistep tasks autonomously — positioning Nvidia against similar offerings from OpenAI, as first reported by Wired
- Nvidia's expected new chip targets AI inference, where competition from custom silicon built by Google, Amazon, and others is fast intensifying — a different battleground from the AI training market, where Nvidia already holds an estimated 80% share
- Nvidia reportedly paid $20 billion late last year to license technology from Groq, with Groq founder Jonathan Ross, president Sunny Madra, and other team members agreeing to join Nvidia to help advance and scale the licensed tech
- Kevin Cook, a senior equity strategist at Zacks Investment Research, told TechCrunch that attendees should expect to learn how Nvidia plans to use the Groq relationship as part of its broader inference push
Why it matters: Nvidia's GTC keynote reveals its next phase: with 80% of AI training locked up, the real battleground is inference, where a $20 billion Groq licensing deal and rumored new chip show Nvidia spending big to fend off Google and Amazon's custom silicon. NemoClaw extends the fight to AI agents.
