Apple Will Improve AI Cloud Computers With M4 Chips Starting Next Year

M4 Ultra chip, the latest variant of the M4 family (photo: x @bakri3d)

JAKARTA - Apple is reportedly planning a major update to the cloud computers used to process Apple Intelligence demand. Currently, these computers rely on M2 Ultra chips to handle various AI-based requests privately in the cloud. According to a report from Nikkei Asia, Apple plans to upgrade this device with the M4 chip starting next year.

Apple is working with Foxconn to build a new server specifically for artificial intelligence (AI) in Taiwan. This step was taken to accelerate the development of AI-based features, where Taiwan was selected for offering engineering talent and R&D resources also used by Nvidia, another client of Foxconn.

Apple Intelligence uses two types of processing: on-device and online. For tasks that require heavy computing, Apple utilizes Private Cloud Compute (PCC) modules that use end-to-end encryption to maintain user data privacy.

According to Apple, PCC is the most advanced security infrastructure ever used for large-scale AI cloud computing, with dedicated silicon and operating systems designed for high security.

References on PCC first appeared in early 2023, and most of the current PCC modules are using the M2 Ultra chip, while some modules with the M1 chip handle lighter tasks. With significant progress on the M4 chip in AI processing, it's not surprising that Apple wants to upgrade this device.

Nikkei Asia reports that the PCC module to be produced next year is likely to be equipped with an M4 Ultra chip, the latest variant of the recently announced M4 family. This chip is expected to be able to increase efficiency and speed in data processing from Apple Intelligence.

Apple still hasn't confirmed the specific variant of the M4 chip to be used in PCC modules, but tech observers predict that the M4 Ultra variant will be the primary option to optimize AI performance in the cloud.