## Nvidia Vera Rubin NVL72: The Future of AI Inference During his CES 2026 keynote, Nvidia CEO Jensen Huang introduced Vera Rubin NVL72, the company’s next-generation rack-scale AI accelerator. This supercomputer promises a significant leap in inference performance, with a 5x increase compared to the previous generation, Grace Blackwell NVL72. Another crucial aspect is cost reduction: Vera Rubin NVL72 aims for a 10x lower cost per token, making AI inference more accessible and efficient. Shipping of this new system is expected in the second half of 2026. Nvidia continues to invest heavily in developing cutting-edge hardware solutions for artificial intelligence, solidifying its position as a leader in the industry. Innovation is critical to meeting the growing demand for computing power needed for increasingly complex AI models.