Apple plans to begin its foray into generative AI by offloading complex queries to its M2 Ultra chips running in data centers before moving to the more advanced M4 chips.
bloomberg Apple reports that it plans to put the M2 Ultra on cloud servers to run more complex AI queries while handling simple tasks on the device. wall street journal It was previously reported that Apple wanted to create custom chips to deploy in data centers to ensure security and privacy in a project called Project ACDC (or Apple Chips in Data Center). But the company now believes existing processors already have sufficient security and privacy components.
The chip will be deployed in Apple's data centers and eventually in servers run by third parties. Apple operates its own servers across the U.S. and is working on a new center in Waukee, Iowa, which it first announced in 2017.
Apple hasn't been as early into generative AI as competitors like Google, Meta, and Microsoft, but it has published research on the technology. In December, Apple's machine learning research team released MLX, a machine learning framework that can efficiently run AI models on Apple silicon. The company has also published other research into AI models that suggests what AI might look like on its devices and how existing products like Siri might be upgraded.
Apple focused on AI performance in announcing its new M4 chip, calling its new Neural Engine a “very powerful chip for AI.”
