Overview
- Nvidia released an open-source family of AI models built to manage and optimize quantum hardware, aiming to speed progress toward practical quantum computing.
- The company introduced the Vera Rubin platform with seven new chips, including the Vera CPU and Rubin GPU, designed to cut inference token costs by up to tenfold.
- Recent results showed fiscal fourth-quarter revenue of $68.1 billion with $62.3 billion from data centers as agentic and multimodal AI drove demand, reinforcing Nvidia’s lead in AI infrastructure.
- Competition intensified as Cerebras filed to go public and was reported to land large deals with OpenAI and AWS, including a $20 billion pact tied to 750 megawatts of compute.
- Nvidia and Cadence agreed to use Cadence physics engines so Nvidia can train robots inside high-fidelity simulations, a shift that can shrink training time and improve model quality.