Apple likely to start producing AI chips in 2026, eyes in-house data centres

Apple is reportedly gearing up to begin mass production of its first-ever AI server chips later this year, a move that could quietly reshape how it builds, powers and delivers its artificial intelligence services.

In the ongoing scramble for AI dominance, where Nvidia’s chips are worth their weight in gold and every major player is chasing “intelligence” in silicon form, the iPhone maker appears ready to throw a heavyweight punch of its own.

Apple to manufacture in-house AI chips

According to renowned Apple analyst Ming-Chi Kuo, the company’s self-designed AI server processors are moving swiftly toward large-scale manufacturing, with production expected to kick off in the second half of 2026.

Kuo, posting on X, suggested the chips are now mature enough for Apple to take them from lab prototypes to real-world deployment.

While Apple has long been known for designing its own silicon, think of the A-series chips inside iPhones or the M-series in Macs, this marks a bold new step into server-grade territory.

These chips aren’t designed for your pocket or laptop, but they’re built for the racks of data centres, where massive AI workloads are trained and executed.

AI server chips are the unseen engines driving modern artificial intelligence systems. Unlike traditional CPUs, they’re optimised for performing billions of calculations in parallel, allowing them to handle intensive tasks such as large language model training, image recognition and natural language processing.

In other words, they’re the powerhouses behind the chatbots, smart assistants and recommendation systems that are rapidly weaving themselves into daily life.

From cloud to chip, Apple wants it all

The reported development fits neatly into Apple’s broader strategy to control every layer of its ecosystem, from the chip inside your iPhone to the cloud running its AI models. The company has already built a reputation for tightly integrating hardware and software, but stepping into AI infrastructure marks new territory even by Apple’s standards.

Alongside chip development, the company is said to be laying plans to construct and operate its own data centres , possibly as soon as next year.

These would be tailored to run Apple’s proprietary AI workloads, enabling the company to keep sensitive data processing entirely under its own roof. That aligns with Apple’s long-standing focus on privacy, a value it has used to differentiate itself from competitors like Google and Meta.

Owning the full AI stack also means Apple could reduce its reliance on third-party providers such as Nvidia or AWS, gain more control over cost and performance, and fine-tune its systems for efficiency and security.

It’s a classic Apple playbook move: take something the rest of the industry depends on, build it in-house, and wrap it in the company’s signature focus on design and integration.

Interestingly, Apple’s push for independence doesn’t mean it’s cutting ties with the wider AI ecosystem. The company recently confirmed a collaboration with Google to bring Gemini AI capabilities to Siri , a reminder that even as Apple invests heavily in its own infrastructure, it’s keeping the door open to partnerships when it makes sense.