Apple is using generative AI cautiously, first using its current M2 Ultra CPUs in data centers and then switching to its next M4 chips.
Apple thinks the security protections on the current M series CPUs are adequate
Apple intends to transfer sophisticated AI inquiries to M2 Ultra processors housed in their data centers, according to Bloomberg. This approach replaces an earlier proposal known as Project ACDC (Apple Chips in Data Center), which called for the creation of unique chips made just for data centers. The trillion-dollar corporation currently feels, meanwhile, that the security protections on its current M series chips are enough for their AI requirements.
These M2 Ultra processors will be initially installed in Apple’s own data centers, with the possibility of later being expanded to servers owned by other parties. Apple has a nationwide network of data centers, and a new one is now being built in Waukee, Iowa.
Apple has prioritized research and development, while rivals like Google, Meta, and Microsoft have made significant strides in the field of generative AI. A framework called MLX, created by the company’s machine learning team to optimize AI models for Apple silicon, was launched in December. Additionally, they have released studies looking into how AI might be used on gadgets and how it might improve already-existing services like Siri.
With its recent focus on AI performance at the M4 processor introduction and its “outrageously powerful” neural engine, Apple appears to be preparing for a more significant role in the field of generative AI.