Apple’s technical paper on Monday revealed that its two artificial intelligence models, Apple Intelligence, were pre-trained on Google-designed cloud chips. One of these chips is called TPU (Tensor Processing Units), developed by Google.
Google’s TPU tensor processing unit was introduced in 2015 and publicly released in 2017. Initially designed for internal workloads, it now has broader applications. Apple’s move validates that top technology companies in the field of AI training models are seeking alternatives to Nvidia’s GPUs.
Nvidia’s expensive graphics processing units (GPUs) have dominated the high-end AI training chip market and have experienced high demand in recent years, leading to supply shortages. OpenAI, Microsoft, and Anthropic have used Nvidia’s GPUs in their models, and other tech companies, including Google, Meta, Oracle, and Tesla, have also purchased Nvidia chips to build their AI systems and products.
Although Apple did not explicitly state that it did not use any Nvidia chips, the company’s description in its 47-page technical paper did not mention the company. Apple stated in the paper that its Apple Foundation Model (AFM) and AFM server were trained on Google Cloud TPU clusters. This implies that Apple rented servers from cloud providers to perform computations. Apple stated in the paper that this system allows them to efficiently scale the training of AFM models, including AFM on devices, AFM servers, and larger models.
Reuters pointed out that while Apple did not explicitly state that it did not use Nvidia chips, its description in the technical paper did not mention Nvidia. Apple’s engineers stated in the paper that using Google’s chips enables them to create larger and more complex models than the two models studied in the paper. Apple introduced a series of new AI features at its developer conference in June, including integrating OpenAI’s ChatGPT technology into its software.
Investors are concerned about the future application development of Apple Intelligence on the iPhone. Apple’s report indicates that it will rely on Google chips instead of Nvidia’s graphics chips, which dominate about 80% of the market share in the AI graphics processor market. Chips produced by Google, Amazon, and other cloud computing companies also require cooperation with Nvidia.
In its research report, Apple did not explicitly state that it did not use Nvidia chips, but its description of AI hardware and software did not mention Nvidia’s hardware. To train its AI models, the iPhone manufacturer used two forms of Google Tensor Processing Units (TPUs), which were organized into large chip clusters. Apple used 2,048 TPUv5p chips to build AI models that can run on iPhones and other devices. For its server AI models, Apple deployed 8,192 TPUv4 processors.
Nvidia focuses on graphics processing units (GPUs) widely used for AI tasks. Google and Nvidia differ in their sales approach, with Google selling access to TPUs through the Google Cloud Platform. Customers interested in purchasing access to TPUs must create software through Google’s cloud platform system to use these chips.
Apple has introduced partial features of Apple Intelligence to its beta testers. Reuters reported on the use of TPU chips in June, but Apple’s recently published research disclosed the extent of its dependence on Google’s hardware.
Apple’s engineers stated in the report that using Google’s chips enables them to create larger and more complex models than the two models studied in the report. Apple’s new iPhone and Siri features in recent years have been underwhelming. The newly introduced Apple Intelligence is only available for iOS 18 and requires at least an iPhone 15 Pro or higher, MacOS 15, and iPadOS 17 to use.
Experts point out that based on Bloomberg’s report, Apple’s delay in launching Apple Intelligence until the end of the year means it will miss the timing of the upcoming iPhone firmware update in September. Developers can only start testing the new AI features this week at the earliest. Investors and Apple fans are looking forward to the Apple Intelligence AI-enhanced features in the new Apple iPhone 16, expected at the end of the year, but may be disappointed.
Apple has been slow in hardware and software development and has a strong relationship with Google. Will Apple Intelligence be a remarkable product that replaces Siri or just a rehashed Siri 2.0? The iPhone once was a groundbreaking product launched by Steve Jobs, but with the rapid development of AI, the iPhone is no longer the only choice for smartphone users. Apple’s innovative AI features will undoubtedly affect its position and stock price in the technology industry.