NVIDIA’s GPU Technology Conference (GTC) returns to San Jose on March 16–19, 2026, positioning itself as a full‑stack AI and accelerated computing event—from infrastructure and developer tools to open models, agentic systems, robotics, and physical AI.
If you’re building with AI (or supporting teams that are), GTC is one of the few conferences that consistently blends (1) platform announcements, (2) deep technical sessions and hands‑on labs, and (3) face‑to‑face access to practitioners across the ecosystem.
Quick links (official)
- Main event page: https://www.nvidia.com/gtc/
- Registration / pricing: https://www.nvidia.com/gtc/pricing/
- Session catalog: https://www.nvidia.com/gtc/session-catalog/
- Keynote info: https://www.nvidia.com/gtc/keynote/
Featured speakers and key moments
Jensen Huang keynote (live + livestream)
NVIDIA says CEO Jensen Huang will deliver the keynote from SAP Center on Monday, March 16 at 11 a.m. PT. The company describes it as covering advancements across the AI stack—from accelerated compute and “AI factories” to open models, agentic systems, and physical AI. NVIDIA also notes the keynote will be livestreamed and available on demand, and that online viewing doesn’t require registration.
GTC Live keynote pregame show
NVIDIA also highlights a pregame show focused on accelerated computing beyond AI and the “five‑layer” stack behind large infrastructure buildouts. NVIDIA’s newsroom announcement called out notable speakers including:
- Aravind Srinivas (CEO, Perplexity)
- Harrison Chase (CEO, LangChain)
- Deepak Pathak (CEO, Skild AI)
- Daniel Nadler (CEO, OpenEvidence)
- Arthur Mensch (CEO, Mistral AI)
“Open frontier models” discussion
NVIDIA also highlighted a discussion hosted by Huang on the state of the art in open frontier models, with leaders from organizations including a16z, AI2, Black Forest Labs, Cursor, Reflection AI, and Thinking Machines Lab (among others).
Why attend
- Builder‑level technical depth: From CUDA/CUDA‑X and libraries to inference optimization and large‑scale training/inference patterns.
- Hands‑on training + certification: NVIDIA is advertising full‑day workshops, hands‑on labs, and onsite certification opportunities across AI and accelerated computing domains.
- Coverage across the AI stack: Agentic AI, open models, AI infrastructure (“AI factories”), robotics/physical AI—useful both for hands‑on implementers and platform decision‑makers.
- Network effects: Meetups, expert sessions, and ecosystem conversations with developers, researchers, startups, and enterprise practitioners.
Tips for getting the most out of GTC
- Use the session catalog early and build a schedule around the 2–3 tracks you care about most.
- Mix strategy + tactics: Pair the keynote/panels with at least one lab or workshop so you leave with something immediately actionable.
- Bring your real constraints: Latency, cost, evals, fine‑tuning, GPU utilization—GTC is a good place to find the one session (or person) who’s solved your exact problem.
Registration
You can register via NVIDIA’s pricing page: https://www.nvidia.com/gtc/pricing/. NVIDIA’s event page also mentions a 20% discount code (GTC26-20) for a conference pass—confirm current eligibility and terms during checkout.