NVIDIA and Meta establish long-term partnership to drive AI infrastructure innovation

On February 17th, American tech giants Meta and NVIDIA announced the establishment of a new long-term strategic partnership. This collaboration not only involves large-scale deployment of chips but also comprehensive optimization from hardware to software. In the context of increasingly fierce competition in the artificial intelligence field, this news quickly drew significant industry attention.

Following the announcement, the stock prices of Meta and NVIDIA rose in after-hours trading, while AMD’s stock price temporarily dropped more than 4%. According to disclosures from both parties, one of the core aspects of this cooperation is that Meta will deploy millions of NVIDIA chips in its data centers, including Blackwell architecture GPUs, next-generation Rubin architecture GPUs, and Arm-based Grace processors. This marks NVIDIA’s Grace CPU achieving its first large-scale independent deployment, posing a strong challenge to the traditional x86 architecture market.

Additionally, Meta plans to introduce the more powerful Vera series processors by 2027 to further solidify its position in high-efficiency AI computing power. This partnership is not just about simple hardware procurement; the engineering teams of both companies will collaborate on the software and hardware design for Meta’s next-generation large-scale language models (such as successors to Llama4) to optimize underlying performance.

NVIDIA, through integrating CPU, GPU, network technologies (such as Spectrum-X Ethernet platform), and software ecosystem, provides Meta with a unified solution covering training, inference, and data processing. Jensen Huang stated that no other company currently deploys AI at Meta’s scale, combining cutting-edge research with industrial-grade infrastructure to build the world’s largest system for billions of users.

Zuckerberg emphasized that expanding cooperation with NVIDIA and leveraging its Vera Rubin platform to build leading clusters is a key foundation for realizing the vision of “providing personal superintelligence for everyone worldwide.” Although Meta has been actively developing its own AI chips in recent years, this large-scale procurement of NVIDIA chips is seen as a long-term lock-in of external computing power needs, ensuring it remains competitive in AI commercialization against rivals like Google and Microsoft.

Meta also mentioned that it will integrate NVIDIA’s security technology into the AI features of its communication app WhatsApp. Zuckerberg stated that by introducing NVIDIA’s confidential computing technology, Meta can enhance performance while meeting strict data security and privacy requirements.

Analysts believe that the scale of this cooperation could reach hundreds of billions of dollars. Last month, Meta announced that its capital expenditure for 2026 could reach up to $135 billion, much of which will be allocated to AI infrastructure development. Chip analyst Ben Bajarin from Creative Strategies pointed out that Meta’s large-scale adoption validates NVIDIA’s “full-stack” infrastructure strategy, which includes both CPU and GPU deployment.

It is worth noting that Meta has been pushing its own AI chip strategy in recent years, aiming to optimize performance and reduce costs for its unique workloads. However, according to informed sources, the project has faced technical challenges and deployment delays, leading the company to continue relying on NVIDIA’s mature solutions.

This cooperation has directly impacted NVIDIA’s competitor AMD. After the announcement, AMD’s stock price temporarily fell over 4% in after-hours trading, while Meta and NVIDIA’s stocks rose by 1.5% and 1.8%, respectively, indicating strong market recognition of this partnership.

From an industry perspective, the collaboration between NVIDIA and Meta reflects a shift in AI computing focus from model training to inference. Inference tasks require efficient, low-latency computing power, and NVIDIA’s Grace CPU is specifically designed for this need, with performance-per-watt significantly better than traditional CPUs.

The signing of this cooperation agreement also signals a major reshuffle in industry competition. For traditional chip giants like Intel and AMD, Meta’s large-scale shift toward Arm architecture CPUs is a stark warning, suggesting that the power structure in the hyperscale data center market is undergoing a structural shift.

(China Economic Observer, Li Qiang / Text)