Apple’s decision to use Google-designed chips over industry leader Nvidia’s for its artificial intelligence software infrastructure has been revealed in a research paper published on Monday. Despite Nvidia’s dominance in the AI processor market, Apple opted for Google’s cloud infrastructure for its AI tools and features. The research paper did not explicitly mention the use of Nvidia chips, indicating Apple’s reliance on Google hardware for its AI models.
In developing its AI models, Apple utilized two variations of Google’s tensor processing unit (TPU) organized in large clusters of chips. While Nvidia focuses on graphics processing units (GPUs) widely used for AI efforts, Google sells access to TPUs through its cloud platform. Apple’s decision to use Google’s TPUs showcases a different approach to AI infrastructure compared to Nvidia, with its chips being integrated into the forthcoming suite of AI tools and features set to be rolled out to beta users this week.
Apple’s engineers mentioned in the research paper that Google’s chips could potentially enable the creation of larger and more sophisticated models than the two discussed in the paper. The company’s unveiling of new AI features at its June developer conference, including integrating OpenAI’s ChatGPT technology into its software, demonstrates its commitment to advancing AI capabilities. The stock price of the Cupertino-based company saw a slight decline in regular trading on Monday following the news of its reliance on Google hardware for its AI infrastructure.