Apple Inc. is reworking its Mac chip strategy around artificial intelligence, moving the next generation of its M-series processors from general-purpose computing toward designs built for AI workloads, according to a report from Digitimes cited by 36Kr. The change marks a shift in architecture priorities that will show up in the M7 line, which is expected to carry a substantially more powerful neural engine.
The report describes a road map in which AI capability, not raw CPU speed, is the organizing principle. Apple has been adding neural engines to its chips for years, but the next generation is being designed the other way around: the AI components set the shape of the chip, and the general-purpose cores are arranged around them. Engineers familiar with the planning say the difference is visible in how the silicon is laid out, with the neural engine no longer treated as an add-on.
The shift has been building for a while. Apple’s software has leaned increasingly on on-device AI, from photo processing to text completion, and the company has made clear that it sees local intelligence as a feature, not a compromise. Moving the hardware to match means the Mac can run larger models and more complex tasks without sending data to a server, a design goal that also serves Apple’s privacy positioning.
The M7 series is the first concrete expression of the new approach. The neural engine in the M7 is expected to be several times more powerful than the current generation, according to the report, and the chip’s memory architecture is being tuned for the large models that AI workloads require. The result, if the plans hold, is a Mac that treats AI as its primary job rather than one of many.
The strategic reasoning is clear enough. Apple’s competitors in the PC market, led by Microsoft and Qualcomm, have been pushing AI features into laptops, and the company needs its own answer. But the deeper motive is architectural. Apple controls its entire stack, from chip to operating system to applications, and an AI-first chip lets it design features that competitors cannot copy without the same integration. The road map is a bet that this integration advantage will matter more as AI becomes the center of personal computing.
The neural engine has been part of Apple’s silicon for the better part of a decade. The company introduced its first dedicated neural processor in 2017 with the A11 Bionic chip, and every iPhone and Mac since has carried a version of it, growing larger with each generation. What has changed is the priority. In earlier designs, the neural engine was one component among many, sized to support the features Apple planned to ship that year. The M7 approach inverts that logic, making the AI capacity the primary constraint around which the rest of the chip is built. The difference matters for software developers, who will be able to rely on a consistent, large AI budget across the entire Mac line, rather than a capability that varies by model.
There are costs. An AI-first design risks optimizing for workloads that are still forming, and Apple has been burned before by bets on specific technologies. The company’s history is full of examples, from 3D displays to specific networking standards, where the industry moved in a different direction than Apple’s early choice. The AI road map carries the same risk, though the size of the industry’s investment in AI makes a broad miss less likely.
The transition also puts pressure on the software. An AI-first chip only matters if applications use it, and Apple has been steadily adding AI frameworks and developer tools to its platforms. The company’s approach has been to build the infrastructure first and let developers discover the uses, a pattern that worked with the original Mac transition to Apple silicon and that the company is repeating with AI.
For the broader industry, the Apple road map is a signal about where personal computing is heading. If the largest PC maker in the premium segment designs its chips around AI, the rest of the market will follow, and the definition of a good laptop will change accordingly. The neural engine is becoming a spec that buyers check, the way screen resolution and battery life became specs in earlier cycles.
The M7 will not arrive overnight, and the report leaves the exact timing open. What is clear is the direction: Apple has concluded that the Mac’s future is AI, and that the chip must be rebuilt around it. The company that spent a decade arguing that its chips were better at general computing is now arguing that general computing will be defined by AI. The architecture will have to prove it, one generation at a time.
