Apple's failed self-driving car program led to the development of the Neural Engine and powerful on-device AI chips, which now underpin Apple's AI hardware strategy, including upcoming M7 chips with significant Neural Engine upgrades.
<figure>
<img alt="" data-caption="" data-portal-copyright="" data-has-syndication-rights="1" src="https://platform.theverge.com/wp-content/uploads/sites/2/2026/03/apple_chips.jpg?quality=90&strip=all&crop=0,0,100,100" />
<figcaption>
</figcaption>
</figure>
<p class="wp-block-paragraph">Apple's self-driving car program <a href="https://www.theverge.com/2024/3/3/24085995/apple-car-project-titan-timeline-driverless-ev-doomed">never really got off the ground</a>, but it may have been what made the company's chips the powerful AI performers they are. Early in the development of the self-driving platform, Apple realized that it would need <a href="https://www.theverge.com/2017/12/10/16758160/apple-machine-learning-autonomous-car-project-titan-ruslan-salakhutdinov">powerful on-device AI processing</a>. While the car processor was <a href="https://www.theverge.com/24085629/apple-car-dead-project-titan-ev-driverless-robotaxi">never finished</a>, as Mark Gurman details in his latest <a href="https://www.bloomberg.com/news/newsletters/2026-07-12/apple-s-chip-plans-m6-m7-pro-m7-max-m7-ultra-m8-details-touch-macbook-pro"><em>Power On</em></a> newsletter, it did lead to the development of the Neural Engine, the backbone of Apple's on-device AI processing. </p>
<p class="wp-block-paragraph">The Neural Engine made its debut with the <a href="https://www.theverge.com/2017/9/13/16300464/apple-iphone-x-ai-neural-engine">iPhone X and the A11 Bionic</a>. In those early days, it was primarily used for computer vision, powering FaceID, Animoji, and a …</p>
<p><a href="https://www.theverge.com/tech/964519/apple-silicon-self-driving-car-ai-m7-ultra">Read the full story at The Verge.</a></p>
# Apple’s failed self-driving car program left a legacy of powerful AI chips
Source: [https://www.theverge.com/tech/964519/apple-silicon-self-driving-car-ai-m7-ultra](https://www.theverge.com/tech/964519/apple-silicon-self-driving-car-ai-m7-ultra)
[](https://www.theverge.com/authors/terrence-obrien)
Terrence O'Brien
is the Verge’s weekend editor\. He’s covered the tech industry for over 18 years and knows a thing or two about synths\.
Apple’s self\-driving car program[never really got off the ground](https://www.theverge.com/2024/3/3/24085995/apple-car-project-titan-timeline-driverless-ev-doomed), but it may have been what made the company’s chips the powerful AI performers they are\. Early in the development of the self\-driving platform, Apple realized that it would need[powerful on\-device AI processing](https://www.theverge.com/2017/12/10/16758160/apple-machine-learning-autonomous-car-project-titan-ruslan-salakhutdinov)\. While the car processor was[never finished](https://www.theverge.com/24085629/apple-car-dead-project-titan-ev-driverless-robotaxi), as Mark Gurman details in his latest[*Power On*](https://www.bloomberg.com/news/newsletters/2026-07-12/apple-s-chip-plans-m6-m7-pro-m7-max-m7-ultra-m8-details-touch-macbook-pro)newsletter, it did lead to the development of the Neural Engine, the backbone of Apple’s on\-device AI processing\.
The Neural Engine made its debut with the[iPhone X and the A11 Bionic](https://www.theverge.com/2017/9/13/16300464/apple-iphone-x-ai-neural-engine)\. In those early days, it was primarily used for computer vision, powering FaceID, Animoji, and augmented reality features\. But by laying the groundwork for on\-device AI processing, Apple established itself as an early leader by bringing the Neural Engine to desktops with the[M\-series chips](https://www.theverge.com/2024/5/7/24148451/apple-m4-chip-ai-ipad-macbook)\. While Apple’s AI software efforts have lagged behind the rest of the industry, its hardware has been impressive\. It’s also what has allowed Apple to tout its privacy features, since less data is sent to the cloud\.
Apple is making its AI hardware a cornerstone of its strategy going forward\. According to Gurman, the company is skipping the Pro, Max, and Ultra versions of its upcoming M6 chip\. Instead, it’s accelerating development of the M7, which should arrive in the first half of 2027 with significant Neural Engine upgrades\. The M7 Ultra is expected to be the basis for a new server product from Apple as well, with support for up to 1\.5TB of RAM\.
**Follow topics and authors**from this story to see more like this in your personalized homepage feed and to receive email updates\.
- Terrence O'Brien
A developer reverse-engineered Apple's private APIs to enable training neural networks directly on the Apple Neural Engine (ANE) in M4 Macs and iPhones, bypassing CoreML and GPU. The project demonstrates that ANE hardware is capable of training, though with limitations like low utilization and CPU fallbacks for some operations.
Apple plans to skip M6 Pro/Max chip variants to accelerate development of M7 chips for enhanced on-device AI, aiming for a 2027 release. The move responds to competition from Nvidia, AMD, Intel, and Qualcomm in AI processing.
This paper presents a reverse-engineered account of Apple Neural Engine (ANE) architecture, programming, and performance across A11-A18 and M1-M5 chips, based on direct measurements and static analysis.