Nvidia GTC 2025: AI Goes Big, Robots Get Smarter, and GPUs Rule the World

The world has been waiting for Nvidia's GTC 2025 conference (GPU Technology Conference) held in San Jose, California with a bundle of anticipation. This year, the giant chipmaker showcased the company's latest advancements in artificial intelligence (AI) and computing technologies. CEO Jensen Huang delivered a keynote address highlighting several significant announcements.​

Huang introduced its upcoming AI chip architectures, Blackwell Ultra, Vera Rubin, and Vera Rubin Ultra. Blackwell Ultra is slated for release later this year, offering enhanced memory capacity. Vera Rubin is expected in late 2026, featuring increased processing speeds, with Vera Rubin Ultra following in 2027.

We spoke to Liam Galin of data analytics company SQream, a first timer at the annual conference, who thinks this is probably the most important conference when it comes to AI.

“It's the combination of theoretical and practical AI and that is something that I really, really liked,” he says.

This year’s event is expected to attract approximately 25,000 in-person attendees and a virtual participation of 300,000 attendees. As always, the event is attracting a diverse audience, including developers and AI engineers, business executives and investors, academia and researchers, automotive and robotics industry experts, healthcare and scientific computing specialists, and media and gaming enthusiasts.

“It's a great place to get to know the industry,” says Galin. “You see the ecosystem, you see the partners, you see customers,” he gushes.

Tech industry leaders, CEOs, and venture capitalists are present at the conference to explore investment opportunities and partnerships.

“I'm looking at it from a business perspective, but the beauty of this conference is that there are many different opportunities for people from the academia, for programmers and for business people to meet their peers. It's very unique in the sense that it covers all the layers and all the fronts of the AI world,” he says.

He points out that software developers working on AI, deep learning, and high-performance computing are of course attending the conference to learn about new tools, SDKs, and frameworks, and have a lot to gain.

“I think that one of the key benefits of the conference, especially for developers, is the endless amount of workshops where they really touch the basics and the advanced capabilities that are relevant for different types of industries, different types of applications, and different type of technologies.

“So, developers gain a lot in this conference and that's why in terms of demographics, you see a lot of developers coming to this conference. They come to this conference to absorb what is new and what is next in the world of GPUs and AI.”

Galin has been in the AI industry for eight years and is happy to see that “AI got to the stage of practicality.” The company he works for uses GPUs to make advanced data queries more cost effective for companies.

“You see how you can really harness AI and not just generate funny pictures, but harness AI for business values and addressing business pains. This conference covers it all, starting from the very minor details of technology,” he explains.

The conference has been highlighting Nvidia's focus on robotics too, introducing "Blue," a robot developed in collaboration with Disney Research and Google DeepMind. This initiative aims to advance robotics technologies and includes the new physics engine, Newton, designed to enhance simulation capabilities.

Additionally, Nvidia unveiled Isaac GR00T N1, an open-source model for developing humanoid robots, and the evolving Cosmos AI model for creating synthetic training data, aiming to lower costs and improve efficiency in AI and robotics training. ​

Nvidia announced Dynamo, an open-source inference software system designed to scale AI models effectively. The company also revealed a partnership with General Motors to integrate custom AI systems into self-driving cars, reflecting Nvidia's strategic move into the automotive sector. ​

“If you look at the roadmap of Nvidia, with both its hardware and software suite, the company is covering so many industries and so many directions in the future of, I would say, even mankind,” he says.

For software developers, GTC remains essential, offering insights into cutting-edge AI frameworks, GPU advancements, and open-source tools that shape the future of AI development. As Nvidia continues pushing technological boundaries, GTC serves as a crucial hub for learning, networking, and driving the next wave of AI-driven innovation.