At CES 2026, Nvidia introduced a major AI infrastructure update featuring six new chips built into its Rubin platform, aimed at accelerating both training and inference for advanced AI workloads.
Alongside the hardware, Nvidia launched a broad set of open generative AI models including expanded Nemotron agent-building models and World Foundation Models for robotics and synthetic data creation to support reasoning, simulation and physical AI applications.
These combined offerings emphasize enabling developers and enterprises to build more capable, open AI systems that go beyond traditional chat interfaces into embodied and real-world workflows.
.jpg)
