Apple and Intel Reach Landmark Chip Deal, Ending Apple's TSMC Exclusivity for M-Series Silicon

When Apple severed its relationship with Intel as a chip supplier in 2020 — replacing Intel processors in Macs with its own silicon — it looked like the end of a long and frustrating partnership. Six years later, Apple is bringing Intel back into the fold, not as a chip designer but as a chip manufacturer. And this time, the U.S. government helped broker the deal.

Apple and Intel have reached a preliminary agreement for Intel to produce Apple’s base M-class system-on-chip at Intel’s Arizona Fab 52 facility, targeting production ramp in the second half of 2027. The Wall Street Journal, which first reported the deal, described it as a significant milestone for Intel’s foundry ambitions and a notable strategic shift for Apple, which has relied exclusively on Taiwan Semiconductor Manufacturing Co. for Apple Silicon since the M1 era.

What Intel Will Make — and What It Won’t

The deal is narrower than it might first appear. Intel will manufacture Apple’s entry-level M-class SoC — the processor that ships inside the MacBook Air and base iPad Pro, the volume workhorses of Apple’s lineup. Higher-end variants, including the M-class Pro, Max, and Ultra chips, will continue to be manufactured by TSMC.

That scope is significant. The base M-class chip shipped roughly 20 million units in 2025, with 2026 and 2027 projections holding steady at 15 to 20 million units annually. For Intel Foundry, winning a contract of that scale from the world’s most valuable company is a proof point that no marketing campaign could replicate.

Intel will use its 18A-P process node for the Apple chips — a variant of the company’s 18A process, which Intel describes as its most advanced fabrication technology and which represents the culmination of Intel’s multi-year effort to reclaim process leadership from TSMC and Samsung.

Why Apple Is Diversifying

Apple’s decision to break its TSMC exclusivity is driven by supply and competition dynamics that have quietly intensified over the past two years.

TSMC is under extraordinary demand pressure. The AI computing boom has turned every advanced node — 3nm, 2nm, and the forthcoming 1.6nm — into a contested resource. NVIDIA, AMD, Qualcomm, and a constellation of AI chip startups all compete for allocation at TSMC’s most advanced fabs. Apple, despite being one of TSMC’s most important customers, is not immune to the squeeze.

During Apple’s Q1 2026 earnings call, CEO Tim Cook confirmed the constraint directly: iPhone 17 models had been supply-limited during the quarter because Apple could not procure sufficient A19 and A19 Pro chips from TSMC on the schedule it needed. At the same time, Apple’s plans to shift its entire modem lineup to TSMC’s 2nm process — representing hundreds of millions of units per year — are set to consume enormous capacity, leaving less room for M-class allocations.

Dual-sourcing the base M-chip to Intel removes one pressure valve while preserving the TSMC relationship for the chips where leading-edge performance matters most.

The Government’s Role

The Apple-Intel deal did not emerge purely from market logic. The Trump administration played an unusually direct role in making it happen.

After the CHIPS and Science Act allocated roughly $8.5 billion in grants to Intel in 2023, the Trump administration converted nearly $9 billion of that federal support into an approximately 10% equity stake in the company — an unusual move that gave the government a direct financial interest in Intel’s commercial success. Commerce Secretary Howard Lutnick has been actively lobbying tech executives to place orders with Intel Foundry, and President Trump personally raised the topic during a White House meeting with Tim Cook in early 2026.

For the administration, the deal represents a validation of its “America First” semiconductor strategy: two iconic American companies partnering to manufacture chips domestically, reducing dependence on Taiwanese fabs that sit within range of Chinese military capability.

Intel’s Foundry Vindication — With Caveats

For Intel CEO Lip-Bu Tan, the Apple deal is precisely the kind of headline win he has been working toward since taking over the company. Intel Foundry has spent years rebuilding its process technology, recovering from a string of manufacturing delays under previous leadership that allowed TSMC to pull far ahead on transistor density and yields.

The 18A process — which uses Intel’s proprietary RibbonFET gate-all-around transistor architecture and PowerVia backside power delivery — has shown promising early results in independent evaluations. Industry analysts have described 18A-P, the variant Apple will use, as competitive with TSMC’s N2 node that currently dominates leading-edge production.

But winning Apple’s confidence is not the same as delivering at scale. TSMC’s Apple Silicon production has been running for six years and achieves yields that are the envy of the semiconductor industry. Intel must demonstrate not just that it can make chips that pass Apple’s validation tests, but that it can make 15 to 20 million of them per year with the consistency Apple’s supply chain demands.

Intel shares surged 14% on the day the deal was reported, and have continued to outperform the broader semiconductor index in May, up more than 33% from the start of the month.

Geopolitical Calculus

The strategic subtext of this deal extends beyond the two companies. Every chip Apple buys from Intel Foundry’s Arizona facility is one fewer chip manufactured in a geopolitically exposed location. TSMC’s fabs in Taiwan remain central to global semiconductor supply chains and, consequently, to the geopolitical calculations of every government that depends on advanced electronics.

The Biden-era push for domestic semiconductor capacity gave the U.S. a framework; the Trump administration is now using direct commercial leverage — both carrots (equity stakes, contracts) and sticks (tariff threats, regulatory pressure) — to turn that framework into actual production volume.

Apple’s deal with Intel is one data point in that broader shift. But it is a significant one: when the company that designed and popularized the personal computer as a consumer product decides it is worth betting on domestic foundry capacity, other technology buyers tend to pay attention.

What Comes Next

The 2027 production target gives both companies two years to execute. Apple’s silicon team will need to design a version of the M-class chip for Intel’s 18A-P process — a substantial engineering undertaking, since process technology affects not just transistor density but power delivery, thermal characteristics, and interconnect architecture. Intel Foundry will need to demonstrate manufacturing readiness that satisfies Apple’s infamously exacting supplier certification standards.

If the deal delivers on its promise, it could expand. Apple could extend Intel’s mandate to cover additional chip types, higher-end variants, or future generations of Apple Silicon. If it stumbles — on yields, timelines, or performance parity with TSMC — the relationship may remain a one-generation experiment.

Either way, the semiconductor industry’s center of gravity is shifting. The question is no longer whether domestic alternatives to TSMC can exist. It is whether they can execute at the standard Apple requires.