Apple Takes a Big Leap Forward with Groundbreaking 2nm M6 Chip

Apple has been quiet about its move into the AI space, but it's made a significant step forward with the introduction of its new M6 chip. Built on TSMC's (TSM) most advanced 2nm process, this chip is the first of its kind in a consumer computer and marks Apple's shift towards putting AI directly onto devices rather than relying on remote data centers.

The M6 delivers nearly 30% more peak GPU compute for AI than the M5 and over eight times that of the M1. This allows it to run AI models on the device faster, with up to 32GB of memory enabling large language models (LLMs) to be run locally without sending data to the cloud.

Apple's bet against data centers is significant, as Amazon (AMZN), Microsoft (MSFT), Meta (META), and Alphabet's (GOOG) (GOOGL) Google have spent over $650 billion this year on AI data centers. By putting AI directly onto devices, Apple can potentially save billions in infrastructure costs.

However, there are concerns that Apple may struggle to deliver the software models needed to utilize its new chip fully, with its weak spot being the software meant to run on it. The company's revamped Siri slipped by more than a year and now leans on Google's Gemini.