Apple's self-driving car effort did not become a finished vehicle platform. Its more lasting result may be inside the company's chips.
According to details from Mark Gurman's latest Power On newsletter, early work on the car project made Apple confront a central technical need: powerful on-device AI processing. The car processor itself was never completed, but that push helped lead to the Neural Engine, the hardware foundation behind Apple's on-device AI work.
A car project that pointed Apple toward AI hardware
The self-driving car program never really got off the ground, but the technical direction behind it mattered. A self-driving platform would have required serious AI processing close to the device, rather than depending only on remote computing.
That requirement helped shape Apple's thinking. The company realized early in the self-driving platform's development that it needed stronger on-device AI performance. Even without a completed car processor, the work contributed to a chip strategy that became important far beyond the car project.
The key result was the Neural Engine. It became the backbone of Apple's on-device AI processing and later spread across the company's hardware lineup.
The Neural Engine started with iPhone X
The Neural Engine debuted with the iPhone X and the A11 Bionic. At first, its most visible uses were tied to computer vision.
Those early uses included Face ID, Animoji, and augmented reality features. Each depended on processing information directly on the device, giving Apple a hardware base for features that needed fast local analysis.
That mattered because the Neural Engine was not only a phone feature. Apple later brought the same on-device AI foundation to desktops through the M-series chips. In doing so, the company gave its computers dedicated hardware for AI processing at a time when its AI software efforts were lagging behind the rest of the industry.
The hardware story has been stronger. Apple's chips have remained impressive, and the Neural Engine has supported one of the company's recurring privacy arguments: when more processing happens on the device, less data needs to be sent to the cloud.
Why on-device AI fits Apple's broader pitch
On-device AI is not only a performance decision. For Apple, it also connects directly to privacy positioning.
The source describes Apple's ability to tout privacy features because less data is sent to the cloud. That is the practical value of doing more processing locally: the device can handle more work without relying as heavily on external servers.
For users, that distinction is easy to understand. A chip that can process AI tasks on a phone, desktop, or other Apple hardware gives the company more control over the experience. It can also reduce the need to move data away from the device for certain kinds of processing.
That does not erase Apple's software gap in AI. The source is clear that Apple's AI software efforts have trailed the broader industry. But it also shows why the company's hardware remains strategically important: strong local AI processing gives Apple a base to build on, even when software is still catching up.
The M7 Ultra points to a bigger AI hardware plan
Apple is now making AI hardware a cornerstone of its strategy going forward. Gurman reports that the company is skipping the Pro, Max, and Ultra versions of its upcoming M6 chip.
Instead, Apple is accelerating development of the M7. That chip should arrive in the first half of 2027 and is expected to include significant Neural Engine upgrades.
The most ambitious part of the plan is the M7 Ultra. It is expected to serve as the basis for a new server product from Apple and support up to 1.5TB of RAM.
That detail is important because it suggests Apple's AI hardware strategy is not limited to consumer devices. The same lineage that started with the need for on-device AI processing in a self-driving project is now tied to a possible server product.
A failed program with a lasting chip legacy
The self-driving car program may be remembered as an Apple project that did not produce a finished car. But the article's bigger point is that unfinished projects can still leave behind core technology.
In this case, the car program helped push Apple toward the Neural Engine. The Neural Engine then became central to Face ID, Animoji, augmented reality features, M-series chips, and the company's on-device AI story.
Now, with M7 development accelerating and the M7 Ultra expected to support up to 1.5TB of RAM, Apple appears to be extending that hardware-first AI approach into its next chip generation. The original car platform never arrived, but the AI processing challenge it created helped shape hardware that Apple is still building around.