Apple’s self-developed chip undergoes a generational upgrade: jumping directly from M2 to M5.

Fast Technology, Apple has completed a major hardware upgrade to its Apple Intelligence cloud infrastructure.

Apple’s private cloud computing servers have been upgraded directly from the M2 Ultra chip to the self-developed M5 chip, skipping the M3 and M4 chip nodes, resulting in a significant increase in Apple’s cloud AI computing power.

The core hardware model in this upgrade is J226C, with Apple’s M5 chip as its core processor. The accompanying private cloud computing architecture adds a new Private Cloud Compute Agent Worker component.

This component runs on a custom iOS-based system version and uses a new intelligent agent architecture to handle AI requests.

Furthermore, Apple’s latest iOS 26.4 system includes code that interacts with this new private cloud architecture, completing the hardware and software adaptation.

It is understood that Apple’s private cloud computing servers have long used the M2 Ultra chip, which was unveiled in June 2023. Even after the release of the M3 Ultra, Apple did not use it for cloud architecture updates, and the previously planned M4 chip server upgrade plan has not been implemented on a large scale.

The direct deployment of the M5 chip marks a significant iteration in Apple’s cloud computing architecture.

In addition to deploying the M5 chip to enhance general-purpose computing power in the cloud, Apple is also developing a server chip specifically for AI tasks.

It is understood that this dedicated AI chip is planned for mass production in the second half of 2026 and will be officially deployed in 2027.

In addition, the new hardware based on the M5 chip will further strengthen Apple’s localized production of private cloud computing servers in Houston, Texas, providing computing power for cloud AI functions.