MENLO PARK, California, November 28, 2025: Meta Platforms is in advanced discussions with Google to spend billions of dollars on the search giant’s custom artificial intelligence chips, according to The Information. The deal, if finalized, would mark one of the largest technology supply agreements in the AI hardware sector and could reshape relationships among the world’s top data center operators. The report said Meta is considering deploying Google’s tensor processing units, or TPUs, across its global data centers beginning in 2027.
The companies are also discussing a short-term arrangement that would allow Meta to rent access to Google Cloud’s TPUs as early as next year. The proposed commitment underscores the intensifying competition among major technology firms to secure advanced AI computing infrastructure amid sustained demand for generative AI services. Google’s TPUs are high-performance chips designed to accelerate machine learning and large-scale data processing.
Until now, the technology has primarily been reserved for Google’s internal operations and select cloud customers. An agreement with Meta would represent the first major external deployment of TPUs on a global commercial scale, positioning Google as a direct hardware supplier to another of the world’s largest social and technology companies. The talks come at a time when Google, through its parent company Alphabet, is seeking to strengthen its position in the AI chip market, which has been dominated by Nvidia.
Alphabet shares rise on report of Meta chip discussions
By expanding TPU availability, Google aims to offer a credible alternative to Nvidia’s graphics processing units that currently power most large AI models and services. The Information report said that executives within Google Cloud have been promoting the TPUs as a cost-effective and scalable option for enterprise-level AI workloads. The financial markets responded sharply to news of the potential deal. Alphabet’s shares rose more than 3 percent in early trading this week, pushing the company closer to a valuation of nearly four trillion dollars.
Nvidia’s stock declined about four percent following the report, reflecting investor sensitivity to any sign of increased competition in the AI computing market. Shares of Meta also traded higher on optimism about potential cost efficiencies in its infrastructure spending. Meta has been investing heavily in data centers and custom AI chips as it expands its artificial intelligence capabilities. The company has previously relied extensively on Nvidia’s GPUs to train large language models and to support AI features across Facebook, Instagram, and WhatsApp.
Global chip demand drives new corporate partnerships
Incorporating Google’s TPUs could diversify Meta’s hardware base and increase the resilience of its AI infrastructure. Neither Meta nor Google has issued a public statement regarding the ongoing negotiations. Nvidia also declined to comment. The Information said the discussions are continuing, and no final contract has been signed. The report reflects a broader trend among major technology companies seeking to secure long-term access to AI compute power, a critical resource driving the next phase of digital services, content moderation, and machine learning products.
As demand continues to outpace chip supply globally, the involvement of multiple large-scale suppliers such as Google could further accelerate the expansion of AI infrastructure worldwide. The potential deal underscores the rapidly evolving dynamics of the semiconductor and cloud computing industries, where the ability to deliver reliable, high-capacity AI processing is increasingly shaping corporate partnerships, market performance, and technological leadership. – By Content Syndication Services.