Apple is reportedly developing custom artificial intelligence (AI) chips for its data centers, expanding the company’s silicon capabilities beyond iPhones, iPads and Macs. As reported by the Wall Street Journal (WSJ), the project, known internally as “ACDC” (Apple Chips in Data Center), has been in development for several years.
According to WSJ, Apple is working with Taiwan Semiconductor Manufacturing Co. (TSMC) in the design and production of AI chips. Server chips are expected to focus on AI inference, which involves trained machine learning models pulling out fresh data.
Apple’s plans to set up AI servers by 2025 Haitong analyst Jeff Pu, based on supply chain control, suggests that Foxconn is currently assembling Apple AI servers containing the M2 Ultra chip. Pu also mentions plans to install AI servers powered by the M4 chip in delayed 2025.
Creating its own AI chips is consistent with Apple’s strategy of vertical integration and hardware control to provide greater security and design flexibility. By creating its own chips, Apple aims to enhance the efficiency of its data centers and future AI tools that rely on cloud computing.
CEO Tim Cook confirmed the company’s significant investment in AI, pointing to a major AI-related announcement later this year, potentially at the upcoming Worldwide Developers Conference in June. “We believe in the transformative power and promise of artificial intelligence, and we believe we have the advantages that will set us apart in a fresh era,” Cook said on Apple’s latest quarterly earnings call.
As the AI race heats up, everyone races to develop their tokens, and earning tokens becomes more and more tough and exorbitant. Apple, Amazon and Meta, which have announced their own AI chips in recent months, are following in the footsteps of Google, Amazon and Meta, which have announced their own AI chips, in moving towards developing their own AI chips.
However, Bloomberg’s Mark Gurman denies the WSJ report, noting that Apple has previously started a project dubbed “~’18” to develop its own chips to relieve Google. However, the project was canceled, according to Gurman. He now suggests that Apple is highly unlikely to produce its own server chips due to costs and the “device nature of its LLM.”
In addition to data center chips, Apple is also working on upgrading the neural engines in its next-generation M4 and A18 chips. These improvements support advanced AI capabilities on iPhones, iPads and Macs running the next-generation operating system, which is expected to be announced during the WWDC keynote.
According to WSJ, Apple is working with Taiwan Semiconductor Manufacturing Co. (TSMC) in the design and production of AI chips. Server chips are expected to focus on AI inference, which involves trained machine learning models pulling out fresh data.
Apple’s plans to set up AI servers by 2025 Haitong analyst Jeff Pu, based on supply chain control, suggests that Foxconn is currently assembling Apple AI servers containing the M2 Ultra chip. Pu also mentions plans to install AI servers powered by the M4 chip in delayed 2025.
Creating its own AI chips is consistent with Apple’s strategy of vertical integration and hardware control to provide greater security and design flexibility. By creating its own chips, Apple aims to enhance the efficiency of its data centers and future AI tools that rely on cloud computing.
Augment
As the AI race heats up, everyone races to develop their tokens, and earning tokens becomes more and more tough and exorbitant. Apple, Amazon and Meta, which have announced their own AI chips in recent months, are following in the footsteps of Google, Amazon and Meta, which have announced their own AI chips, in moving towards developing their own AI chips.
However, Bloomberg’s Mark Gurman denies the WSJ report, noting that Apple has previously started a project dubbed “~’18” to develop its own chips to relieve Google. However, the project was canceled, according to Gurman. He now suggests that Apple is highly unlikely to produce its own server chips due to costs and the “device nature of its LLM.”
In addition to data center chips, Apple is also working on upgrading the neural engines in its next-generation M4 and A18 chips. These improvements support advanced AI capabilities on iPhones, iPads and Macs running the next-generation operating system, which is expected to be announced during the WWDC keynote.
