Nvidia delivers over 2.5 million Grace CPUs for AI computing
Nvidia has delivered more than 2.5 million Grace CPUs, its ARM-based processors for AI and high-performance computing. The company is increasing production capacity for its next-generation products.
Chinese technology outlet IT Home reports that Nvidia has surpassed a significant milestone in the delivery of its Grace CPUs, processors designed for artificial intelligence and high-performance computing (HPC).
The company has shipped over 2.5 million Grace chips. Nvidia utilizes these ARM-based processors in servers developed for data centers and AI workloads. The Grace CPU marked Nvidia's first custom-designed processor architecture for data centers.
Nvidia is accelerating its CPU roadmap to solidify its market position as a full-stack systems provider. The company has set a target to produce 1,000 Vera server racks daily. Demand for the next-generation Vera CPU is expected to significantly exceed previous generations, driven by the growing needs of agentic AI.
To meet this demand, Nvidia is collaborating with over 300 global partners to ensure adequate supply of Vera chips. With approximately 12 manufacturing partners, Nvidia aims for a daily production capacity of 1,000 Vera racks. These Vera Rubin racks are described as foundational elements for "AI factories," tailored for large-scale agentic AI workloads.