Apple admits to using Google Tensor hardware to train Apple Intelligence
New artificial intelligence research published by Apple reveals that Apple has been using Google hardware to build the early foundations of Apple Intelligence.
The research paper, called "Apple Intelligence Foundation Language Models" is pretty technical, and details the already-known sources of the language model at the core of the company's new technology. However, a quote buried inside the paper hints that Apple may have been using Google hardware in early development.
In the paper, it says that Apple's Foundation Model (AFM) and the server technology that drives it, were initially built on "v4 and v5p Cloud TPU clusters" using Apple software. There is a great deal of information in the research about how that's done, and what data sources they used to train.
A CNBC report on Monday suggests that Apple rented time on existing Google-hosted clusters, but the research doesn't directly support that, nor say anything about Google or Nvidia at all. What's more likely is that Apple bought the hardware outright from the company, and used it within its own data centers.
The model's initial training being performed on Google-designed hardware ultimately doesn't mean much in the long-run. Apple has been said to have its own hardware derived from Apple Silicon in its data centers to process Apple Intelligence queries.
Said to be called "Project ACDC," Apple is reportedly planning to optimize AI applications within its data centers.
Apple is significantly increasing its investment in the artificial intelligence sector, planning to allocate over $5 billion to AI server enhancements over the next two years. The company aims to match the technological capabilities of industry leaders like Microsoft and Meta by acquiring tens of thousands of AI server units, likely driven by Project ACDC.
Apple has also acquired firms in Canada and France that both work on compressing data used in AI queries to data centers.
Read on AppleInsider
Comments
Is there any source for this claim? Wall street has been asking Google to sell TPU hardware basically forever but as far as anyone knows it's never happened.
EDIT: Apple seems to have more than a passing interest in using Google Tensor hardware for their intelligence features.
"Apple's engineers said in the paper it would be possible to make even larger, more sophisticated models with Google's chips, than the two models it discussed in the paper"
“Reveals” would be more appropriate haha.
I think it’s no surprise given that Chris Lattner went to Google and worked on TensorFlow and created Swift for TensorFlow so….seems logical to me to tap into the leading APIs for machine learning.
https://www.reddit.com/r/swift/comments/1b85ucx/an_interested_read_on_why_google_engineers_chose/
quote: “However, a quote buried inside the paper hints that Apple may have been using Google hardware in early development.”
Apple… almost aways… uses other's technology while it develops it own!
In 2006 the first iPhones used a Samsung processor… And where are we now?
Until 2020 the Mac used Motorola, IBM, Intel chips… And where are we now?
Back in 1997 it included Michorsoft's browser… In 2006… Google Maps…
And we all know that Apple has already surpassed Google, Meta, OpenAI, and Microsoft in user privacy and security in the Artificial Intelligence realm!
But… showing Apple fails bring pageview… even if what you say is useless or partially fake.
Keep your friends close and your enemies closer. In other words, use their hardware to achieve your success. Smart move.
When did Apple do this? Did they rent or buy? How much did they pay? Are they still using this hardware? Are there plans to do this in the future?
It's the same with Waymo vehicles. Google has used a large number of vehicles from various manufacturers with a ton of different modifications over years of testing (they still aren't done).
I would not be surprised if there are Nvidia GPUs sitting in an Apple data center somewhere. Perhaps not for production purposes but for testing.
In the end we will never know all the details of Apple's development process. They certainly want the least amount of this disclosed because it gives their competitors insight into Apple's methodology. Apple's secrecy is a competitive advantage. Note that the open source AI model they released is not the one they will use for Apple Intelligence.
That mobile hardware, software power also includes the M4 iPad Pro and 15 Pro and 15 Pro Max which run programs like Morpholio Trace with ease. But Apple is definitely missing the boat in the cad/3D/construction area it's comical and sad watching current PC laptops getting 2 hours of battery time on site.
How much of that kind of work is done away from a socket?