Nvidia Reports Million-Unit Shipment of Grace Standalone Servers and Focuses on CPU-Based AI Data Centers

As Tom’s Hardware reports (https://www.tomshardware.com/pc-components/cpus/nvidia-has-shipped-hundreds-of-thousands-of-grace-standalone-servers-gpu-firm-pivots-messaging-as-cpus-take-center-stage-in-agentic-data-centers), Nvidia has reached a significant milestone with its Grace CPU architecture: the company has already shipped hundreds of thousands of standalone servers equipped with the Grace CPU. This development underscores the shift in Nvidia’s strategy, where CPUs play a central role in modern, especially agentic AI data centers.

Grace Standalone Servers: A Success Beyond GPUs

The Grace CPU, originally designed for demanding AI workloads and as a complement to Nvidia’s GPU-based solutions, is now also being used as a standalone server platform. These standalone servers not only serve agentic AI applications but are also used in traditional data center environments that do not require dedicated GPU acceleration. The high volume of shipped Grace servers demonstrates that Nvidia has achieved broad acceptance of its CPU architecture among enterprise customers. This is remarkable since Nvidia is traditionally known as a GPU specialist and CPUs have so far been more of a complementary part of its portfolio.

Vera: Nvidia’s New Focus on Agentic AI

Alongside the successful Grace generation, Nvidia is intensively working on Vera, the first CPU specifically developed for agentic AI systems. Agentic AI describes systems that independently make decisions and autonomously perform complex tasks. Vera is intended to help make the data centers of the future more efficient and powerful by optimizing CPU performance for these new requirements. With Vera, Nvidia is betting on a CPU architecture tailored specifically to the needs of AI agents, marking a departure from the company’s previous GPU-centric messaging. This strategic realignment shows that Nvidia is significantly elevating the role of the CPU in AI infrastructure.

Why It Matters

The shipment of hundreds of thousands of Grace standalone servers and the development of Vera mark a turning point in the hardware industry. While GPUs continue to play an important role in AI computations, the CPU is gaining importance as an independent computing unit for AI applications. This could sustainably change the architecture of data centers, as CPUs can be more flexible and, in certain scenarios, more efficient than GPUs. For companies, this means more choice and specialized hardware options to optimally run AI workloads. Agentic AI, which controls complex autonomous processes in particular, benefits from the new CPU generation. Nvidia is positioning itself as a pioneer in a market that is increasingly moving away from pure GPU acceleration toward hybrid and CPU-centered solutions.

Outlook

Nvidia will release further details about Vera in the coming months and expand the availability of the new CPUs. The successful introduction of the Grace standalone servers lays a solid foundation for the acceptance of the Vera architecture. Observers of the hardware industry should closely follow this development, as it could have an impact on the entire server and AI infrastructure. The combination of massive shipments and strategic realignment shows that Nvidia is further solidifying its role as a leading provider of AI hardware while redefining the importance of CPUs in the AI era.