Apple just signaled that the future of AI belongs to custom silicon built on ARM architecture, not massive data centers. The company's promotion of Johny Srouji to chief hardware officer, announced alongside John Ternus becoming CEO, represents a fundamental shift in how the tech industry will compete in the age of artificial intelligence. While headlines focused on the leadership transition, the real story is that Apple is doubling down on a silicon-first strategy that could reshape computing for the next decade .

Srouji, who joined Apple in 2008 from Intel and IBM, has spent nearly two decades building the company's in-house chip capabilities. He led the transition from Intel processors to Apple Silicon on Macs and oversaw every generation of the company's mobile and data center chips. His new role consolidates hardware technologies and hardware engineering under a single executive for the first time in over a decade, signaling that Apple believes the next competitive advantage in AI lives at the silicon layer .

Why Is ARM Architecture Becoming the Foundation for AI on Devices?

ARM architecture has quietly become the dominant platform for embedded artificial intelligence, powering everything from iPhones to industrial sensors. Unlike enterprise AI, which relies on massive GPU clusters and enormous datasets, embedded AI operates in a completely different universe. It runs on devices measured in milliwatts instead of megawatts and executes on microcontrollers rather than accelerator clusters .

Apple's A-series and M-series chips, both built on ARM architecture, now include neural engines, specialized hardware accelerators designed specifically for running machine learning models directly on the device. When Apple unveiled its latest A19 and M5 generations in 2025, they included built-in neural accelerators for powering AI on the device. The company's approach gives users and their data top-notch security and privacy because inference happens locally, not in the cloud .

The distinction between embedded AI and enterprise AI is crucial to understanding Apple's strategy. Embedded AI focuses on efficiency, running tightly defined models measured in kilobytes or a few megabytes rather than gigabytes. These systems solve specific problems locally and in real time, whether that's predictive maintenance analyzing vibration data, vision systems inspecting products on a production line, or voice interfaces recognizing commands .

How Is Apple Consolidating Control Over Its Entire Silicon Supply Chain?

Apple's strategy extends far beyond processors. The company is systematically bringing chip design in-house across multiple categories:

• Mobile Processors: Apple's A-series chips power iPhones and iPads, replacing reliance on external suppliers and allowing the company to optimize specifically for its products.
• Mac Processors: The M-series chips replaced Intel processors starting in 2020, giving Apple complete control over Mac performance and energy efficiency.
• Modems: Apple began moving away from Qualcomm in 2019 with the purchase of Intel's modem business for $1 billion. The company quietly released its first iPhone modem, the C1, in early 2025, and unveiled the C1X in the iPhone 19 in September.
• Wireless Chips: In September, Apple launched its own wireless chip for the iPhone, the N1, replacing Broadcom.
• Neural Accelerators: The Neural Engine, first announced in 2017, is now embedded in every recent iPhone, iPad, and Mac, handling all on-device machine learning tasks.

This consolidation is not accidental. By owning the silicon, hardware, software, and machine learning all in one team, Apple gains an advantage that competitors cannot easily replicate. As Srouji explained in a 2023 interview, "Because we're not really selling chips outside, we focus on the product and that gives us freedom to optimize. And the scalable architecture lets us reuse pieces between different products" .

As Srouji

Apple is also making a major commitment to manufacturing. As part of a $600 billion U.S. investment commitment through 2029, Apple announced in August that it is "leading the creation of an end-to-end silicon supply chain in the United States." The company is manufacturing at TSMC's Arizona campus and at Texas Instruments' two new U.S. factories .

Apple

What Does This Mean for the Future of AI on Devices?

The implications are profound. Gartner projects that 33% of enterprise software applications will include agentic AI by 2028, up from less than 1% in 2024. The companies that control the silicon those agents run on will capture outsized value in that transition . Apple is positioning itself to own that layer.

Unlike competitors who rent compute from cloud providers, Apple builds its own silicon from the ground up. Every agent running on a consumer device depends on inference, and Apple owns some of the best on-device inference silicon in consumer electronics. This is not a coincidence. Srouji's promotion signals that Apple believes the next wave of competitive advantage in AI lives at the silicon layer, and his architecture decisions have already built that capability .

The pairing of Ternus and Srouji at the top of Apple's organization is intentional. Ternus, who has shipped products alongside Srouji for 15 years, understands how to integrate hardware and software seamlessly. By elevating Srouji to command all hardware technologies while installing Ternus as CEO, Apple has formalized a partnership designed to move as a unit through the agentic AI era .

Apple is also ending support for Intel-based Macs with macOS 27, set to roll out in September. Intel-based models, including the 2019 16-inch MacBook Pro, the 2020 27-inch iMac, and the Mac Pro, will not be eligible for the new operating system, with macOS 26 Tahoe set to be the final version supporting Macs equipped with Intel processors . This move reflects Apple's strategic decision to use OS version requirements to drive user migration and advance an ecosystem transition across its installed base.

The lesson for the broader tech industry is clear: the next chapter of AI will be built on silicon, and the companies that control that silicon will control the future. Apple just told the market exactly where it is betting, and it is betting big on ARM-based custom chips designed for on-device intelligence.