Overview
- Jensen Huang’s keynote is set for 11 a.m. PT at the SAP Center to outline Nvidia’s hardware and software plans for the four‑day conference.
- Nvidia is expected to formally kick off its Vera Rubin platform and detail rack‑scale systems, with reporting pointing to NVL72 demos and larger designs under discussion.
- The company plans to showcase how licensed Groq inference technology plugs into its stack, with reports suggesting Groq‑based servers or a new inference chip could be introduced.
- Analysts and reporters expect a preview of the next‑generation Feynman chips along with updates to CUDA, AI agents, and physical robotics platforms.
- Nvidia’s recent investments in optics suppliers and potential deeper work with Intel on CPUs and packaging are in focus as investors assess efficiency, supply, and competitiveness.