Google and Meta have entered into a multi-billion-dollar AI chip deal, which represents a big part of Meta’s AI infrastructure strategy. This announcement highlights the increasing competitive landscape for major technology companies seeking advanced computing capacity as they continue to see a large increase in global demand for training and deploying large AI models.
Meta has signed a multi-year, billion-dollar deal with Google to use Google's AI chips to build its future AI models. According to reports, the deal is primarily about purchasing Google’s TPUs (Tensor Processing Units); that can be used to run machine learning workloads on the chips.
Meta has also signed an agreement with AMD for AI processing chips, only a short time after announcing a plan to purchase millions of GPUs from NVIDIA, thereby demonstrating a growing and aggressive strategy to expand the company's infrastructure for AI.
Google, for its part, has been working to turn its AI hardware capabilities into a broader cloud growth engine. The growth of the external TPU market may allow Google to provide proof of returns from its extensive AI investment.
Meta's ongoing trend to reduce its dependence on a single vendor by expanding its supplier base will help develop more resilient supply chains, increase pricing power, and optimize performance across a variety of AI workloads.
This partnership highlights the competitive nature of major tech companies looking to gain computing power to meet growing demand for advanced artificial intelligence. Meta is proactively pursuing multiple suppliers across the semiconductor industry instead of depending solely on its internal facilities.
The need to fulfill AI compute demand worldwide means that by having a number of different chip suppliers, Meta can maintain a scalable, flexible, and stable long-term infrastructure that can accommodate its growing demand for generative AI capabilities.
Also Read: Meta Stock Climbs Near $656 as Revenue Hits $59.89 Billion
Google is promoting its own Tensor Processing Units (TPUs) against NVIDIA's established Graphics Processing Units (GPUs) amid the global rise in AI-based training and inference.
While Meta and Google are also in competition in those spaces of: Advertising, Social Networking, and AI-based Services, companies are developing relationships with each other to secure their access to these limited computing resources.