Nvidia announced on February 17 that it has entered into a multiyear agreement to supply Meta Platforms with millions of its current and future artificial intelligence chips, positioning itself as a strong competitor in the market against Intel and Advanced Micro Devices. While the financial details of the deal were not disclosed, the agreement encompasses Nvidia’s existing Blackwell chips and its upcoming Rubin AI chips, along with standalone installations of its Grace and Vera central processors.
This strategic move indicates Nvidia’s intention to expand the application of its central processors, which were introduced in 2023 and built on technology from Arm Holdings. The chips are not only aimed at enhancing AI functionalities but also at streamlining everyday technical tasks such as database management. Ian Buck, Nvidia’s general manager for hyperscale and high-performance computing, emphasized that the Grace processors are effective in energy consumption, using half the power for certain tasks compared to previous models, and they already show promise in collaboration with Meta on the Vera processor.
The announcement comes at a time when Meta is also developing its own AI chips and exploring collaboration with Google to utilize its Tensor Processing Units (TPUs) for AI applications. Despite Nvidia not publicly disclosing its sales figures to Meta, analysts suggest that the company has solidified its business relationship, which accounts for a significant portion of its revenue, particularly in its latest fiscal quarter.
This partnership represents a significant step for both companies as they aim to redefine capabilities in artificial intelligence and data processing, capturing emerging markets and enhancing data center efficiencies in the long term.