We just learned something surprising about how Apple Intelligence was trained

A new research paper from Apple reveals that the company relied on Google’s Tensor Processing Units (TPUs), rather than Nvidia’s more widely deployed GPUs, in training two crucial systems within its upcoming Apple Intelligence service. The paper notes that Apple used 2,048 Google TPUv5p chips to train its AI models and 8,192 TPUv4 processors for its server AI models. Nvidia’s chips are highly sought for good reason, having earned their reputation for performance and compute efficiency. Their products and systems are typically sold as standalone offerings, enabling customers to construct and operate them as the best see fit. Google’s TPUs, on the other hand, are only available to users as part of the company’s larger cloud services package. That is, you don’t own Google TPUs so much as lease access to them as customers are required to develop their AI models within the Google Cloud ecosystem. This cloud requirement worked in Apple’s favor, per the research team. They noted that the ability to cluster Google’s TPUs enabled them to harness the necessary processing power to train Apple’s AI models and do so more efficiently than with a standalone system. Apple’s decision to use Google’s products is unexpected, and not just because of the two companies’ longstanding rivalry. Nvidia boasts the dominant market share in terms of AI chips, its accelerators constituting between 70% and 95% of all sales. However, Apple’s decision could be seen as a sign that tech companies are looking to move away from Nvidia’s pricey high-end chips. Amazon, for example, recently revealed that it is developing a new line of AI chips that are purportedly 50% more powerful than Nvidia’s offerings and operate using 50% less power. Microsoft in May announced that it will offer its cloud customers services built atop AMD’s AI chips, rather than Nvidia’s, while Google made similar moves in April. Google recently announced that Gemini will soon replace Google Assistant everywhere, from your phone and smartwatches to smart home speakers. ChromeOS has now joined the transition bandwagon, starting today.
The company has kicked off the stable rollout of Chrome OS M134, and it marks the silent exit of Google Assistant. “When triggering Assistant, you will automatically be directed to the Gemini app on your Chromebook,” Google says in a community update note.
Google says the feature update will be rolling out in a phased manner, so you might not be able to access the Gemini interface immediately after installing the latest software. Just to clear any confusion here, Gemini has been accessible on Chrome OS, but with the new build, it replaces the Google Assistant. Once the transition takes effect, users will see the sparkly Gemini icon in the top-right corner of the launcher window. For now, support for the “Hey Google” hotword for summoning Gemini is absent, even though it works fine on mobile platforms where Google Assistant is in the phase-out process.
Another noteworthy aspect is that Chrome OS will offer Gemini as a Progressive Web App (PWA), instead of a native application experience. Apple CEO Tim Cook should go public to explain the delay in integrating advanced Siri capabilities across its ecosystem, rather than Apple releasing the news quietly via a tech site last week, according to prominent Apple analyst Ming-Chi Kuo. The tech giant showcased an AI-powered Siri at its WWDC event in 2024, as part of its Apple Intelligence initiative. While the virtual assistant does now have some AI smarts, the more advanced features — including personalized responses, task completion across multiple apps, and on-screen awareness –have been delayed until next year at the earliest.