Why Apple has remained with TSMC for more than a decade: Three structural advantages

1.     First access to leading-edge process nodes

Apple’s A-series and M-series processors are designed to maximize transistor density, power efficiency, and performance from day one. TSMC has consistently been the first foundry to bring advanced process nodes, including 7nm, 5nm and 3nm, into high-volume production. At the 2nm node, although Samsung announced production earlier, TSMC will be the first to achieve high-volume mass production at the scale required by customers such as Apple. This has enabled Apple to be the first major chip vendor to commercialize new manufacturing technologies at scale. As a result, each new iPhone and Mac series benefits from the latest advanced process technology. As Apple expands the use of its custom silicon across a broader range of products, its demand for advanced node wafers has also increased significantly. While iPhone continues to account for the majority of 3nm and future 2nm wafer demand, Apple is increasingly deploying advanced node across Macs, iPads, Vision Pro/MR devices, AirPods and its in-house cellular modem.

2.     Industry-leading manufacturing yields and quality

Apple ships 240-250 million iPhones annually, in addition to tens of millions of Macs, iPads, Apple Watches and other devices powered by its self-designed chips. At this scale, even a small reduction in manufacturing yield can translate into significant increases in production costs and constrain product availability. Therefore, unlike most chip vendors, Apple requires high manufacturing yields from day one. TSMC has consistently demonstrated industry-leading yields and manufacturing quality at advanced process nodes, providing Apple with predictable production, lower manufacturing costs and reliable launch schedules. These capabilities have been a key factor behind Apple's long-term reliance on TSMC for manufacturing its most advanced chips.

3.     Unmatched high-volume manufacturing capacity

Apple’s global product launches require hundreds of thousands of wafers within a very short time frame. TSMC is uniquely positioned to meet these requirements, with the ability to rapidly scale production of advanced nodes while maintaining high yields and consistent quality. No other foundry has demonstrated the same combination of advanced-node maturity, manufacturing scale, and execution at the 3nm and emerging 2nm generations. This manufacturing capability enables Apple to launch new products simultaneously across global markets without significant supply constraints.

Strategic advantages of a multi-foundry approach

Despite the advantages of partnering with TSMC, Apple's heavy reliance on a single foundry also creates several strategic risks:

• Dependence on a single foundry limits flexibility: Apple currently relies only on TSMC for manufacturing its most advanced SoCs. While this strategy has supported technology leadership, depending on a single foundry also limits Apple’s flexibility. Qualifying a second manufacturing partner would improve Apple's negotiating position, reduce reliance on a single supplier's technology roadmap, and provide greater flexibility in managing future production requirements.
• Growing demand for advanced manufacturing capacity: The rapid adoption of generative AI has significantly increased demand for TSMC's leading-edge manufacturing capacity. Companies including NVIDIA, AMD, Broadcom, Qualcomm, MediaTek, Amazon, Microsoft, Google and Apple are all competing for capacity on the same advanced process nodes to manufacture AI accelerators, GPUs, smartphone SoCs, PC processors, data center CPUs, and networking chips. This shift is also reflected in TSMC's segment revenues. According to Counterpoint’s tracker for foundry wafer demand by global top two foundries, HPC driven by AI chips now contributes more than half of TSMC's revenue, with NVIDIA emerging as one of its largest leading-edge customers. As AI demand continues to grow, Apple may need to commit to long-term capacity agreements or accept higher wafer prices to secure sufficient production for its product launches. Diversifying a portion of its chip production to an additional foundry could provide greater flexibility and reduce its dependence on a single manufacturing partner over the long term.

• Government support for domestic manufacturing: Governments across the US, Europe, Japan and South Korea are encouraging semiconductor companies to go in for local manufacturing through various incentive programs. The US CHIPS Act and similar initiatives are aimed at reducing dependence on a few key semiconductor manufacturing locations. As Apple continues to expand its investments in the US, increasing chip production in regions outside Taiwan could help it align with government priorities while making its supply chain more resilient.
• Leading-edge manufacturing capacity: Despite TSMC’s expansion into the US, Japan and Germany, most of its leading-edge manufacturing capacity remains concentrated in Taiwan. This leaves Apple exposed to potential disruptions from geopolitical tensions, export controls, natural disasters, and infrastructure-related challenges such as power or water shortages. Given Apple’s global product launches, even a temporary disruption could delay the production of iPhones, Macs, iPads, Apple Watches, and Vision products.

Limited number of potential partners for Apple

TSMC is expected to remain Apple’s primary manufacturing partner due to its leadership in advanced process technology, manufacturing yields and production capacity. However, adding a second foundry does not mean replacing TSMC. Instead, it would help Apple strengthen its supply chain and reduce its dependence on a single manufacturing partner. If Apple decides to diversify its foundry strategy, the number of potential partners is extremely limited. Among the available options, Samsung Foundry and Intel Foundry are the leading candidates, although each differs significantly in terms of technology maturity, manufacturing capability and strategic fit.

1.     Samsung Foundry: Samsung remains the most mature alternative to TSMC, with advanced process technologies including GAA transistors and a 2nm roadmap. However, the company has faced challenges in consistently achieving manufacturing yields comparable to TSMC, particularly during the ramp-up of its 3nm process. Given Apple’s historical shift away from Samsung to avoid relying on a direct competitor, combined with Samsung's current yield challenges at advanced nodes, a return to Samsung appears unlikely. Samsung's yield challenges with its in-house Exynos SoCs have also led it to rely on Qualcomm chipsets for some Galaxy S-series smartphones.

2.     Intel Foundry: Intel could emerge as a strong long-term partner if Apple decides to diversify its foundry strategy. The company is investing heavily in advanced process technologies, including Intel 18A, while also expanding its manufacturing capacity in the US. Intel's growing manufacturing presence in the US aligns with the industry's increasing focus on supply chain resilience and government support for domestic semiconductor production. Apple could initially consider Intel for selected M-series chips or lower-volume products before expanding to higher-volume SoCs as Intel's foundry capabilities mature.

However, Intel has yet to prove that it can manufacture advanced SoCs for external customers at the scale and quality Apple requires. To become a reliable second foundry, Intel will need competitive process technology, consistently high manufacturing yields, reliable execution and the ability to support Apple’s large-volume product launches. If Intel achieves these milestones, it could become a viable secondary manufacturing partner for Apple over the long term.

Why Intel can be a favored partner for Apple

Intel has invested significantly to improve its foundry business, and its 18A process technology is yielding results, positioning Intel Foundry among the leading foundries for leading-edge nodes. Intel is also working on RibbonFET transistors along with PowerVia backside power delivery (Intel’s version of GAA technology), apart from Samsung Foundry and TSMC, and Intel’s CPU based on this technology is showing promising results. Apart from improving its foundry business, Intel is also investing in advanced packaging through its EMIB technology.

Partnering with Intel will give Apple several advantages, such as reducing the geopolitical risks associated with Taiwan and reducing its dependency on a single foundry. Apart from this, partnering with Intel will also help Apple increase domestic manufacturing and procurement, as the US government is promoting more domestic manufacturing.

Most of the leading-edge node capacity at TSMC has been reserved by AI players, and Apple might have to pay a premium for capacity allocation. With a partnership with Intel, Apple might gain greater bargaining power and negotiate pricing with TSMC for leading-edge nodes.

Challenges for Intel foundry migration

Apple needs large volumes of SoCs across its product segments. As Intel is an IDM and does not yet have comparable high-volume foundry capacity, relying on Intel for most of its requirements will not be possible for Apple. Apart from this, Apple also needs extremely high yields for leading-edge nodes and stable production volumes.

Given the domestic manufacturing push by the US government, Apple can start with small volumes, likely for products sold in the US initially, and then gradually increase its allocation to Intel.

Conclusion

We expect TSMC will remain Apple’s primary foundry partner over the next few years, supported by its leadership in advanced process nodes, manufacturing scale and high yield rates. Samsung remains an alternative, but its yield challenges and Apple's earlier move to TSMC make it difficult for Apple to shift back. At the same time, growing demand for advanced-node capacity from AI and HPC customers is tightening capacity at TSMC, which could increase Apple's cost of securing advanced node manufacturing capacity. Also, we expect Apple's premium smartphone shipments to continue growing over the coming years, increasing its requirement for advanced-node wafer capacity and consistently high manufacturing yields from day one.

In this scenario, Intel Foundry could become Apple's second foundry partner over the long term. Intel's progress in advanced process technologies, growing manufacturing capacity in the US, and investments in advanced packaging can support Apple's aim of building a more diversified and resilient supply chain. We believe Apple could start with Intel for advanced packaging and gradually expand the partnership to M-series and then A-series chips.

Intel is more likely to complement Apple's existing foundry strategy, given Intel's manufacturing capacity and the US government's focus on domestic semiconductor production. Apple could initially use Intel to manufacture M-series chips for products sold in the US, while continuing to rely on TSMC for the majority of its global production.

Overall, adding Intel as a second foundry partner would help Apple diversify its supply chain, strengthen its negotiating position with TSMC, and support its long-term manufacturing strategy. For Intel, securing Apple as a foundry customer would validate its manufacturing roadmap and strengthen its position in attracting other leading fabless semiconductor companies.