Nvidia has announced plans for a new AI chip called Vera Rubin, designed to accelerate inference processing. The announcement, reported exclusively by the Wall Street Journal, signals the company's continued expansion in addressing computational demands across the AI infrastructure market.

The Vera Rubin chip targets inference workloads, a critical component of AI deployment where trained models process real-world inputs at scale. By focusing on inference acceleration, Nvidia is positioning itself to serve the growing number of enterprises and cloud providers deploying AI applications in production environments.

The move reflects intensifying competition in AI chip design, with the market seeing increasing demand for specialized processors optimized for different stages of AI workflows. Nvidia's development of dedicated inference hardware underscores the company's strategy to maintain dominance across the full spectrum of AI computing, from training to deployment.