Apple used Tensor Processing Units (TPUs) developed by Google instead of Nvidia's widely-used graphics processing units (GPUs) to construct two critical components of Apple Intelligence.
The decision is detailed in a new research paper published by Apple that highlights its reliance on Google's cloud hardware (via CNBC). The paper reveals that Apple utilized 2,048 of Google's TPUv5p chips to build AI models and 8,192 TPUv4 processors for server AI models. The research paper does not mention Nvidia explicitly, but the absence of any reference to Nvidia's hardware in the description of Apple's AI infrastructure is telling and this omission suggests a deliberate choice to favor Google's technology.
The decision is noteworthy given Nvidia's dominance in the AI processor market and since Apple very rarely discloses its hardware choices for development purposes. Nvidia's GPUs are highly sought after for AI applications due to their performance and efficiency. Unlike Nvidia, which sells its chips and systems as standalone products, Google provides access to its TPUs through cloud services. Customers using Google's TPUs have to develop their software within Google's ecosystem, which offers integrated tools and services to streamline the development and deployment of AI models.
In the paper, Apple's engineers explain that the TPUs allowed them to train large, sophisticated AI models efficiently. They describe how Google's TPUs are organized into large clusters, enabling the processing power necessary for training Apple's AI models. Apple has announced plans to invest over $5 billion in AI server enhancements over the next two years, which should bolster its AI capabilities and reduce its dependence on external hardware providers.
In addition to detailing its use of Google's TPUs, the paper addresses ethical considerations in AI development. Apple emphasized its adherence to responsible data practices, claiming that no private user data was used in training its AI models. The company relied on a mix of publicly available, licensed, and open-sourced datasets for training purposes. Apple added that its training data set, which includes publicly available web data and licensed content, was curated to protect user privacy.
Article Link: Apple Used Google Tensor Chips to Develop Apple Intelligence
Next you'll be telling me that Apple will say good things about the iPhone.Apparently, many many companies...
Real-world gen AI use cases from the world's leading organizations | Google Cloud Blog
Gen AI is everywhere, as top companies, governments, researchers, and startups showcase how they're already using Google's AI solutions to enhance their work.cloud.google.com
Apple aren’t using phone chips found in the Pixels.
I’m sure those in the know know more than MR posters’ speculations.
Building off of and using German WW2 rocket technology doesn’t mean you have to subscribe to German WW2 ideology. 🤦🏼♂️
Apple doesn't make enterprise or data-center-grade AI accelerators. The premise of this comment is incorrect.
They are using Tensor Processing Units (below), not the mobile SoC named Tensor.What an ABSOLUTE JOKE!
Tensor ... chips that are 2yrs behind the industry standard and using old Samsung Foundry manufacturing processes until next year!
This is a colossal failure of endeavor on Apple's part. The ONLY reason to use Tensor chips in 2024 is simply for being efficient in your code as the SoC is garbage! I'm not just talking out my arse here cause even Samsung's much hated SoC the Exynos 2400 outperforms on ALL measurements and in real world tasks and applications as well.
Samsung Galaxy S24 vs. Google Pixel 8
Wondering which compact flagship phone suits you best? Consider the Galaxy S24 and the Pixel 8. If you are in the market for a compact flagship in the...www.gsmarena.com
Sorry ONLY in video playback minutes the Google Pixel 8 which is the tensor chip Apple supposedly used is 'SLIGHTLY' better by just a few minutes more.
While your statement is true it has no bearing on the current topic.
In addition Nvidia demanded core system access for their slotted GPU cards. Apple refused and never used Nvidia again.
I don't know that this last statement is entirely accurate. Apple has been shipping NPUs with its consumer SoCs since 2017 for phones, and since 2020 for the Mac. You can quibble on the pros and cons between NPU and TPU cores, but both give the A series and M series a boost for AI and machine learning tasks (as long as they're paired with enough RAM).
Smart choice after the recent Crowdstrike Windows outage, as an out of bounds memory access crashed Windows.
