Nvidia unveiled its next-generation Rubin computing architecture at CES 2026, positioning it as the most advanced platform for AI workloads in the company’s history. The architecture has already entered full-scale production, with further capacity expansion expected in the second half of the year.
First announced in 2024, Rubin continues Nvidia’s rapid hardware innovation cycle and is set to replace the Blackwell platform. The new chips are expected to be adopted by leading cloud providers and Nvidia partners, including OpenAI, Anthropic, and Amazon Web Services. Rubin will also serve as the foundation for flagship supercomputers, such as HPE’s Blue Lion and the upcoming Doudna system at Lawrence Berkeley National Laboratory.
The platform integrates six specialized chips designed to operate as a unified system. At its core is the Rubin GPU, paired with the new Vera CPU optimized for agentic reasoning, along with upgrades to BlueField networking and NVLink interconnects aimed at eliminating bottlenecks in data transfer and system communication. According to the company’s internal benchmarks, Rubin delivers up to 3.5x faster model training and up to 5x faster inference compared with Blackwell, reaching performance levels of up to 50 petaflops while significantly improving energy efficiency.
Beyond the Rubin architecture, Nvidia made several other major announcements at CES 2026. The company introduced the Alpamayo family of open models for autonomous driving and broader AI adoption in transportation, as well as Nvidia Cosmos, a platform for physical AI and robotic simulation.
Don't waste another minute manually transferring leads from Facebook to other systems. SaveMyLeads is a simple and effective tool that will allow you to automate this process so that you don't have to spend time on the routine. Try SaveMyLeads features, make sure that this tool will relieve your employees and after 5 minutes of settings your business will start working faster.