Nvidia CEO Jensen Huang spoke at GTC 2025 about Nvidia’s big plans. Huang unveiled the next wave of artificial intelligence (AI) and accelerated computing, from the Blackwell GPU architecture to agentic AI, AI factories, and robotics breakthroughs. The Sync newsletter has a summary.
Nvidia makes GPUs that handle graphics, math, and AI tasks fast. They launched a new GPU called Blackwell Ultra and unvelied Rubin, a chip three times stronger than Blackwell. They also teased Feynman, a chip for even bigger tasks.
Nvidia wants to lead in AI. Huang described AI factories that create intelligence in the form of tokens that turn into code, music, or new designs. The factories need lots of power, like a nuclear plant gives. They run digital twins, which are virtual copies of real things. Companies use these to test ideas fast.
Nvidia also showed Groot N1, a model for human-like robots.
Huang discussed Nvidia software for science and engineering. This software works with Nvidia's chips. It helps people in medicine or physics solve problems quicker. Huang said AI is ready to help in big ways. It can write code or design drugs before making them real.
Building a smart world
Nvidia aims to put AI everywhere. Their chips keep getting faster. Blackwell Ultra has more memory for big tasks. Rubin uses a tiny 3nm design for extra power. Feynman will push limits even further. Nvidia also launched DGX Spark, a small AI computer. It fits on a desk and costs $3,000. DGX Station is bigger and stronger. These computers help people build AI at home.
Digital twins play a key role. Nvidia’s Omniverse platform runs copies of things like wind turbines or factories. Engineers test designs without building first. This saves time and money. AI factories power these twins. They also host smart robots that learn and move. Groot N1 shows how robots could work with us.
Huang wants Nvidia to shape AI’s future. They mix hardware and software for all industries. Chips alone aren’t enough. They offer CUDA, a tool for coding on GPUs. This keeps developers using Nvidia’s stuff. The vision is bold. AI could change how we live and work. Huang believes it’s coming, even if some hype fades.
