According to a Monday report, Apple avoided industry leader Nvidia in preference of Google’s developed processors.
Apple will rely mainly on Google processors as the pillar of AI-related features and capabilities slated to release instead of using Nvidia’s GPUs for its artificial intelligence software infrastructure.
Since Nvidia has already taken over this industry and brought Google, Amazon.com, and other cloud computing businesses to 80% of the total share, Apple’s decision to employ Google’s cloud infrastructure is noteworthy. Apple’s strategic choice to rely on some other source to reduce its hardware reliance appears wise.
Apple has never said that it is avoiding Nvidia chips in any way. However, a comment in a study paper outlining the hardware and software underneath the corporation’s AI tools and capabilities is drawing criticism. This description quite glaringly does not mention Nvidia processors. Apple said no when questioned about this, allowing industry experts to make informed judgments.
The study claims Apple’s AI model training procedure arranges two types of Google’s tensor processing units (TPUs) in large chip clusters. Designed to run on iPhones and other items sold by the firm, the digital behemoth Google started distributing at least 2,648 of its powerful TPUv5p processors as it got ready to develop its new AI model. Developing the server-side for this model included at least 8,192 TPUv4 processors, which is an even more critical component.
Google’s choice to use TPUs is counter to Nvidia’s emphasis on GPUs, which are now almost exclusively related to machine learning research. The decision becomes much more fascinating as Google’s TPUs vary in design and access. Unlike Nvidia’s products, which are accessible outside of other providers, Google exclusively leases TPUs via the Google Cloud Platform, thereby requiring consumers to create their program with the support of Google’s cloud to use them.
Apple disclosed this data just as it was about to provide beta users with its Apple Intelligence tools. Its developer beta program, which offers early access to upcoming software releases, combines many AI-powered capabilities. Above all, the use of its writing tools is improved today. Users could discover a better Siri, which now recognizes text and speech. Thanks to email summarising, AI substitutes may also be utilized for Mail and Smart Replies. Bonus: beta testers may now use AI to help them transcribe their dictations.
Real-world use and feedback will help Apple enhance its artificial intelligence capabilities, including ChatGPT integration, and increase Siri’s capacity to communicate with other applications. This approach will help the company create the best general release in autumn by letting people acquaint themselves with AI and provide feedback.
Had Monday’s thorough investigation not taken place, the extent of Apple’s use of Google hardware would have remained undisclosed even if Reuters revealed in June that the firm employs TPU processors. Moreover, Apple’s developers claimed to be able to create even more important and sophisticated models using Google’s processors, implying that the present deployment marks merely the start of Apple’s artificial intelligence initiatives.
This development follows Apple’s annual developer conference in June, when the company unveiled a suite of fresh artificial intelligence capabilities, including Apple software ecosystem integration of ChatGPT technology from OpenAI. Some perceived this as evidence of Apple’s dedication to the future, combining ambiance with the newest technology advances, which caused general excitement in the tech industry.
Though this news is important, Apple’s stock hardly reacted; it dropped 0.1% to $218.24 in regular trade on Monday. Given Apple’s past propensity to enter new technical terrain purposefully and with polished solutions, this subdued reaction might suggest that investors are waiting and seeing Apple’s AI approach.
