What to Expect from NVIDIA GTC 2026 and the Next Wave of AI

NVIDIA GTC 2026 will be the biggest AI event of the year. It starts in just a few days and is the biggest event in the world for developers, researchers, and business leaders to learn about accelerated computing and AI. The conference will take place in San Jose, California, from March 16 to 19, 2026. It comes at a crucial time when AI is moving from being an experimental tool to being an essential part of infrastructure that powers everything from robotics to data centers and agentic systems.

Attendees and viewers around the world can expect big news about the next wave of AI, including updates on the Vera Rubin architecture, AI factories, physical AI, and breakthroughs in inference. This is because NVIDIA CEO Jensen Huang will be giving his famous keynote speech. If you’re going to NVIDIA GTC 2026 in person or online, here’s everything you need to know: the dates, where to watch it live, what to expect, unconfirmed rumours, and NVIDIA’s long-term plan to be the best in the AI era.

When and Where Is NVIDIA GTC 2026?

The NVIDIA GTC 2026 conference will be held in downtown San Jose, California, from March 16 to 19, 2026. This is the center of Silicon Valley. The main conference and exhibits are at the San Jose McEnery Convention Center (150 West San Carlos Street). There are also many other places in the city where sessions, hackathons, and networking can happen.

The main event, Jensen Huang’s keynote speech, will take place on Monday, March 16, from 11 a.m. to 1 p.m. PT at the SAP Center. Get there early (or tune in) for the GTC Live 2026 Pregame Show at 8 a.m. PT. It will have special guests talking about real-world AI advances.

There will be more than 30,000 people from more than 190 countries at the event, which will have more than 700 sessions, hands-on labs, exhibits, certifications, and networking events. There are no more full conference passes available, but there are still Exhibits Only passes available. Key content is also free to access online.

Where to Watch NVIDIA GTC 2026 Live (Keynote and More)

You don’t need a ticket or registration to catch the biggest moments. NVIDIA streams the entire Jensen Huang keynote and pregame show free and live on the official site:

Primary Livestream: nvidia.com/gtc/keynote— no registration required. It will also be available on-demand afterward.
NVIDIA’s main site (nvidia.com) will host the stream directly.
YouTube: NVIDIA’s official channel often mirrors or archives keynotes for easy replay (search “NVIDIA GTC 2026 Keynote”).

The GTC Developer Community Livestream on March 18 will also have real-time show-floor demos, interviews, and behind-the-scenes content. You can watch it on NVIDIA channels through evt.to links. To interact live, use the NVIDIA app or social media hashtags like #GTC or #ShowMeGTC.

Certain passes allow you to attend keynote watch parties in person. To get the full conference experience (sessions, labs, and exhibits), you need to sign up at nvidia.com/gtc.

Tip: Set a reminder for March 16 at 8 a.m. PT. The pregame will get you excited with CEOs from Perplexity, LangChain, Mistral, and others, and then Huang will give his main speech.

What to Expect at NVIDIA GTC 2026: Key Themes and Highlights

Jensen Huang’s keynote will cover the whole AI stack, from chips and software to models and real-world uses. He will focus on “buildouts measured in gigawatts.” Expect news that makes AI a necessary part of infrastructure, like the internet or electricity.

Core themes include:

Physical AI and Robotics — Advances in simulation, autonomous vehicles, and humanoid robots using open models.
Agentic AI and Reasoning Systems — Always-on AI agents (spotlight on projects like OpenClaw for tasks like coding or scheduling).
AI Factories and Inference — Scaling data centers into intelligence engines with high-performance training and low-latency inference.
Open Frontier Models — Rapid progress in customizable, trustworthy AI (highlighted in a special panel moderated by Huang on March 18).
CUDA and Developer Tools — Updates on GPU-accelerated computing, new features, and libraries for building production AI apps.

In addition to the keynote speech, there will be more than 700 technical sessions, more than 70 hands-on labs, researcher posters, hackathons, certifications, and live demos from more than 240 startups. The multi-gigawatt Vera Rubin deployment with Thinking Machines Lab is an example of a partnership that will be shown, showing real-world AI infrastructure at a large scale.

This year’s GTC focuses on putting AI to use in everyday life, turning it from a buzzword into a part of everyday life in fields like healthcare, energy, entertainment, and science.

Unconfirmed News and Rumors: What Analysts and Leaks Are Predicting

While NVIDIA keeps major reveals under wraps until the keynote, analysts and reports point to several high-impact updates:

Vera Rubin Platform Deep Dives — Already in full production and slated for H2 2026 availability. Expect detailed specs on HBM4 memory, up to 5x performance gains over Blackwell in dense floating-point and inference workloads, plus massive NVL72/NVL144/NVL576 rack configurations. Rubin Ultra (2027) and related systems could get early showcases.
New “World-Surprising” Inference Chips — Jensen Huang has teased “several new chips the world has never seen before.” Rumors point to an inference-focused processor incorporating technology from recent partnerships (including Groq LPU elements post-acquisition talks), targeting 10x lower inference costs and specialized agentic workloads.
Feynman Architecture Tease — The post-Rubin generation (targeted for 2028) may get preliminary looks, including 1.6nm process tech, backside power delivery, and massive interconnect improvements for long-context reasoning and agentic AI.
Silicon Photonics, Networking, and Power Innovations — CPO (co-packaged optics), advanced NVLink, and energy-efficient designs to handle gigawatt-scale AI factories while addressing power and cooling challenges.
Expanded Software and Ecosystem — Deeper integration for agentic systems, sovereign AI, and hybrid quantum-HPC workflows.

These are based on roadmaps shared in prior events (e.g., CES 2026 and GTC 2025) and analyst predictions — nothing is officially confirmed until Huang takes the stage. Past GTCs have delivered surprise announcements, so expect at least one “leather jacket moment” that shifts industry expectations.

The Future of NVIDIA: Leadership in the Next Wave of AI

NVIDIA isn’t just going along with the AI trend, it’s setting the standard for it. The company has a clear multi-year roadmap that keeps it ahead of competitors making custom chips. It updates its architecture every year (Blackwell → Rubin → Rubin Ultra → Feynman).
Important parts of NVIDIA’s future are:

Full-Stack Dominance: From GPUs and CPUs (Grace/Vera) to networking (NVLink, BlueField), software (the CUDA ecosystem), and apps (Omniverse for simulation and DGX systems).
AI as Infrastructure—Huang sees AI factories as the new backbone of industry, and Rubin systems are already making it possible to train frontier models at gigawatt scale.
Going beyond data centers, new sources of income will come from physical AI (robots, autonomous everything), agentic systems, and tools for consumers and businesses.
Open Innovation and Ecosystem: Working with hyperscalers, startups, and governments (sovereign AI) makes sure that everyone uses it, while CUDA stays the standard.

There are problems with power use, competition, and geopolitics, but NVIDIA’s integrated platform and quick updates put it far ahead of the rest. Analysts think that systems from the Rubin era will cut the need for training GPUs by up to 75% and inference costs by a lot, which will speed up the spread of AI in all industries.

GTC 2026 is more than just a conference; it’s a plan for how NVIDIA will run the next decade of AI.

Conclusion: Don’t Miss the Next Wave of AI

NVIDIA GTC 2026 will be a defining event that shows how AI and faster computing are changing the world. Set a reminder for March 16 at 8 a.m. PT, go to nvidia.com/gtc/keynote/ for the free livestream, and get ready for tech breakthroughs that will change the world for years to come.

This is where the future of AI starts, whether you’re a developer working on models, a business leader planning AI infrastructure, or an investor keeping an eye on the next big leap. Stay tuned for the leather jacket, which will bring with it the next wave of new ideas. For more articles like this, make sure to also read: Your AI Slop Bores Me: The Website Where You Become the AI