Apple has opted to use chips designed by Alphabet Inc.’s Google, rather than industry leader Nvidia to develop its new artificial-intelligence software infrastructure. This infrastructure will support Apple’s forthcoming suite of AI tools and features, Apple detailed in a research paper.
Apple’s decision to utilize Google’s cloud infrastructure is noteworthy given Nvidia’s dominance in the AI processor market, Reuters news report said. Nvidia’s GPUs are highly sought after. Nvidia controls 80 percent of the market for AI processors, including those made by Google, Amazon.com, and other cloud computing firms.
Apple did not explicitly state the exclusion of Nvidia chips. The research paper did not mention Nvidia hardware. Apple could not be reached for immediate comment.
AI investment by Apple is expected to bolster its share in the smartphone market. Apple’s smartphone market share dropped to 15.8 percent in the second quarter of 2024 from 16.6 percent in Q2 2023. Apple’s iPhone shipments grew only 1.5 percent year over year as compared to the overall smartphone-market’s growth of 6.5 percent.
Apple is planning to make Apple Intelligence available to software developers for the first time for testing as soon as this week via iOS 18.1 and iPadOS 18.1 beta, Bloomberg news report said earlier.
According to the research paper, Apple used two types of Google’s tensor processing units (TPUs) organized in large clusters to train its AI models. For the AI model intended for iPhones and other devices, Apple used 2,048 TPUv5p chips.
For its server AI model, the company deployed 8,192 TPUv4 processors. Unlike Nvidia, which focuses on standalone GPU products, Google offers access to its TPUs through its Google Cloud Platform, requiring customers to build software within the platform to use the chips.
Apple is set to release portions of Apple Intelligence to beta users this week. This development follows Apple’s unveiling of several new AI features at its June developer conference, including the integration of OpenAI’s ChatGPT technology into its software.
Google did not respond to requests for comment, and Nvidia declined to comment. Apple’s engineers noted in the paper that it is possible to create even larger and more sophisticated models with Google’s chips than the ones currently discussed.
