Over the past two decades, Apple’s stock price has experienced remarkable growth, primarily fueled by its iconic consumer devices. This journey began with the introduction of the iPod and iMac, followed by groundbreaking innovations such as the iPhone, iPad, Apple Watch, and AirPods.
However, Apple’s significance extends far beyond its renowned gadgets. Deep within its Silicon Valley headquarters, concealed within an unassuming room filled with hundreds of humming machines and a select group of engineers donned in lab coats, Apple is diligently crafting custom chips that empower its most beloved products.
The debut of Apple’s proprietary semiconductors dates back to the iPhone 4 in 2010. As of this year, all new Mac computers rely on Apple’s in-house silicon, marking the end of a 15-plus-year reliance on Intel.
John Ternus, who oversees hardware engineering at Apple, noted, “One of the most profound changes at Apple, particularly in our product lineup over the last two decades, is how we now manage many of these technologies internally. At the top of the list is our silicon.”
However, this transformation has also exposed Apple to new risks. Its cutting-edge silicon primarily originates from a single supplier, the Taiwan Semiconductor Manufacturing Company (TSMC). Furthermore, the smartphone market has faced challenges, including a sales downturn, and competitors like Microsoft are making significant strides in the field of artificial intelligence.
In November, CNBC had the unique opportunity to visit Apple’s Cupertino campus, becoming the first journalists permitted to film within one of the company’s chip laboratories. During this visit, they conducted an exclusive interview with Johny Srouji, the head of Apple’s silicon division, shedding light on Apple’s endeavors in the intricate realm of custom semiconductor development—a pursuit also embraced by industry giants like Amazon, Google, Microsoft, and Tesla.
Srouji emphasized, “We have a substantial workforce of engineers, but if you examine our chip portfolio, it is remarkably streamlined and efficient.”
Diverging from conventional chip manufacturers, Apple’s focus is not on selling chips to external entities but rather on optimizing products. Srouji elaborated, “Because we’re not primarily chip vendors, we concentrate on the end product. This affords us the freedom to fine-tune, and our scalable architecture allows us to reuse components across different products.”
The Journey of Apple’s Custom Chips: Srouji joined Apple in 2008 to lead a small team of engineers, initially around 40 to 50, dedicated to designing custom chips for the iPhone. Shortly after his arrival, Apple acquired P.A. Semiconductor, a 150-person startup, for $278 million, signaling its intent to produce its chips.
According to Ben Bajarin, CEO and principal analyst at Creative Strategies, Apple’s inherent design focus aims to “control as much of the stack” as possible, with the purchase of P.A. Semi being a pivotal step in this direction.
Two years post-acquisition, Apple introduced its maiden custom chip, the A4, featured in the iPhone 4 and original iPad. Srouji explained, “We created what we call the unified memory architecture that is adaptable across products. We built an architecture starting with the iPhone, then scaling it to the iPad, watch, and eventually the Mac.”
Apple’s silicon team has since burgeoned to comprise thousands of engineers stationed in labs worldwide, including Israel, Germany, Austria, the U.K., and Japan. In the United States, Apple boasts facilities in Silicon Valley, San Diego, and Austin, Texas.
The central focus of Apple’s chip development is the system on a chip (SoC), which amalgamates the central processing unit (CPU), graphics processing unit (GPU), and other components. Apple also integrates a neural processing unit (NPU) to power its neural engine, setting it apart from conventional SoCs.
Apple’s key SoCs include:
- The A series, from A4 in 2010 to the recently announced A17 Pro, serving as the central processor for iPhones, iPads, Apple TVs, and HomePods.
- The M series, introduced in 2020, powering new Macs and advanced iPads.
- The S series, a smaller system in package (SiP) for Apple Watch.
- H and W chips utilized in AirPods.
- U chips facilitating communication between Apple devices.
- The forthcoming R1 chip, designed for Apple’s Vision Pro headset, dedicated to processing input from cameras, sensors, and microphones.
Apple’s approach enables precise tailoring of chips for specific products. For instance, the H2 chip in the 2nd generation AirPods Pro enhances noise cancellation, while the S9 chip in the Series 9 Apple Watch unlocks unique features like double tap. In iPhones, the A11 Bionic introduced the first Apple Neural Engine for on-device AI tasks.
The most recent A17 Pro, featured in the iPhone 15 Pro and Pro Max, brings significant advancements in computational photography and advanced rendering for gaming, including hardware-accelerated ray tracing and mesh shading acceleration. This innovation has led to iPhone-native versions of popular games, such as Ubisoft’s Assassin’s Creed Mirage, The Division Resurgence, and Capcom’s Resident Evil 4.
Apple proudly claims that the A17 Pro is the industry’s first 3-nanometer chip to reach high-volume production. Srouji emphasized the choice of 3-nanometer technology, explaining that it allows for more transistors in a given space, enhancing both product performance and power efficiency.
In conclusion, despite not being a traditional chip manufacturer, Apple’s commitment to custom silicon development has positioned it as an industry leader, demonstrating its ability to innovate and adapt to ever-evolving technological landscapes.