Apple is developing a new AI chip that could power server-based AI functions: Report

Apple is reportedly working on developing its artificial intelligence (AI) chipsets to run software in data centers. If true, it would mark a major change of plans for the Cupertino-based tech giant, as previous reports have highlighted that the company is only focused on on-device AI features for its devices. However, the development of AI chips suggests that Apple is also considering server-based AI functions. Some of these features may be unveiled at the Worldwide Developers Conference (WWDC) 2024, which is scheduled to take place on June 10.

According to a report from The Wall Street Journal, Apple has been working on its own server project, known internally as Project ACDC, for several years. Citing unnamed people familiar with the matter, the report highlighted that the tech giant now plans to introduce specific chipsets for these data centers to make them capable of AI computing.

The tech giant is reportedly working with Taiwan Semiconductor Manufacturing Co. (TSMC) for the design and production of these AI chips, but it is not clear whether the manufactured processors have shown a final result. These chips are said to be used only for inference purposes (running the AI ​​model) and not for training large language models (LLM). It should be noted that TSMC is Apple’s primary chip manufacturing partner and produces most of the silicon used in the company’s devices.

If the claims are true, it could mean a major shift in Apple’s AI strategy. Many reports have previously claimed that the company is only interested in bringing AI features to the device that can be handled entirely locally. Evidence of this plan was seen when it unveiled the new iPad Pro 2024 with the M4 chip, which comes with a new Neural Processing Unit (NPU) to perform AI tasks.

So far, however, no smartphone manufacturer has been able to offer a set of AI functions that function entirely locally due to the high hardware requirements and instead split the processes between the server and the device. Notable examples are Samsung Galaxy AI and Oppo’s AI features. Apple seems to have realized this issue as well, and now plans to combine AI features on both the device and the server for its ecosystem.

This likely means that functions that do not violate privacy can be performed by servers, while those functions that access the user’s data can be handled locally. This combination could also allow Apple to extend these features to its older devices that don’t have the hardware to support on-device AI features. While this is just speculation, we’ll know for sure if Apple unveils its AI features at WWDC 2024.