Nvidia announces Vera Rubin AI chip, claims it is 5X more powerful than Blackwell

Nvidia has unveiled its next-generation AI computing platform, called Vera Rubin, at the ongoing Consumer Electronics Show (CES) 2026 in Las Vegas. The announcement was made by Nvidia founder and CEO Jensen Huang, who said the new platform is now in full production.

Rubin is designed to support large-scale artificial intelligence workloads and is aimed at data centres, cloud providers and enterprises building advanced AI systems. Nvidia said the platform is built to lower the cost of AI computing and support faster training and deployment of AI models. Rubin succeeds Nvidia’s Blackwell architecture and is part of the company’s broader push to expand AI use across industries.

What is the Vera Rubin AI platform

Vera Rubin is Nvidia’s first six-chip AI platform developed using an “extreme codesign” approach, where multiple components are designed together. The platform includes Rubin GPUs, Vera CPUs, NVLink 6 networking, Spectrum-X Ethernet Photonics, ConnectX-9 networking cards and BlueField-4 data processing units. Nvidia said this design helps reduce bottlenecks and improve performance for large AI workloads.

The Rubin GPU delivers up to 50 petaflops of inference performance using NVFP4 precision. The Vera CPU is designed to handle data movement and AI agent processing tasks. Nvidia said the platform is built to support large-scale AI training and inference more efficiently.Lower cost and faster AI processingAccording to Nvidia, the Rubin platform can reduce the cost of generating AI tokens to about one-tenth of previous platforms. The company also introduced an AI-focused storage system called Inference Context Memory Storage, which is designed to improve long-context AI processing by increasing speed and efficiency.Nvidia said the Rubin platform will be used alongside its open AI models and software tools to support applications in areas such as autonomous driving, robotics, healthcare and climate research.

Elon Musk on Nvidia’s Rubin AI chip

The announcement was shared by an X user Sawyer Merritt who also shared a video of the chip maker’s next-generation Rubin chips. Responding to the post, Musk said “It will take another 9 months or so before the hardware is operational at scale and the software works well”, highlighting deployment realities beyond raw power claims.