Apple (NASDAQ:AAPL) is quietly doing something big in the desert. It is using its wallet to pull more of its chip supply chain onto U.S. soil. The Feb. 23, 2026 Wall Street Journal report described rare access to the supplier sites. The headline project is TSMC’s Arizona mega-campus. Apple plans to buy more than 100 million chips from that site this year. Apple executives also pointed to U.S. work on wafers in Texas, packaging in Arizona, and server assembly in Houston. This push sits next to a noisy backdrop. Tariffs still matter. Taiwan risk still matters. And politics now cares about “where chips come from.” Apple’s recent earnings call added extra color. Management highlighted a $600 billion U.S. investment pledge over four years. It also flagged tight capacity at advanced chip nodes. So yes, the U.S. buildout is real. But it is also partial, slow, and still years behind Taiwan.
Supply-Chain Security & Geopolitics
The world’s chip map has a glaring bottleneck. Most leading-edge chips come from Taiwan. That is not a secret. It is also not a comfortable fact for Washington. China has threatened to take control of Taiwan. The U.S. sees that as a direct supply risk. Apple sits right in the blast radius. It designs the chips that run iPhones and Macs. It relies on foundries to make them. If Taiwan supply is disrupted, Apple cannot “code its way out” of that.
Tariffs add another kind of friction. They do not stop production. They do change costs and planning. Apple has every reason to reduce “single-region exposure.” A U.S. fab does not eliminate risk. It does diversify it. That matters for boards, insurers, and customers. It also matters for how fast Apple can ship during shocks. The pandemic car shortage showed what chip scarcity can do. That memory sticks.
There is also a politics layer that cannot be ignored. Two presidents have pushed large buyers to reshore. Apple’s $600 billion U.S. commitment is part optics and part procurement reality. Some of that spending is normal U.S. payroll and operations. Still, chips are where symbolism meets real industrial capacity. A fab in Arizona is a visible answer to “why are we dependent.” Yet, it is also a slow answer. The most advanced chips are still made in Taiwan. The U.S. effort is a hedge, not a replacement.
Apple’s Purchasing Power Reshaping Suppliers
Apple is not building a foundry. Apple is doing something more Apple-like. It is placing giant orders and forcing alignment. It is the biggest customer at the new TSMC Arizona site. Apple said it plans to buy over 100 million chips from that fab this year. It also said it will buy as much of the fab’s output as it can. That kind of demand is the difference between “pilot project” and “real ramp.”
This is how Apple changes industries. It commits to leading-edge process nodes. That gives TSMC confidence to spend huge sums on the next generation. Apple executives described a playbook. Apple helps ramp a node to high volume and high yield. Then other customers follow. In plain English, Apple goes first. Everyone else benefits from the learning curve. It is not charity. It is leverage.
Upstream, Apple’s influence also shows up in materials. GlobalWafers opened a new silicon-wafer plant in Sherman, Texas. The company said Apple helps it sell wafers by pushing chip makers to use them. That is subtle power. Apple is not just buying parts. It is shaping what its suppliers buy from their suppliers. Downstream, Apple is also funding or supporting packaging and assembly work. That fills gaps the U.S. still has. It also reduces cross-ocean shipping loops.
There is a practical reason Apple cares right now. Supply is tight at advanced nodes. On the latest earnings call, Apple said it faced constraints tied to advanced-node capacity. It pointed to 3-nanometer production as a key gating factor. When capacity is scarce, priority goes to the customer with scale and commitment. Apple has both. A domestic footprint can also help Apple negotiate, plan, and secure allocation. It will not remove all constraints. It can reduce the odds of being last in line.
The “How It’s Made” U.S. Buildout From Wafers To Packaging
Think of chips like a relay race. First you need a wafer. Then you carve circuits. Then you package chips so devices can use them. The Journal’s reporting walked through this chain in the U.S. The upstream piece starts with raw silicon. At GlobalWafers in Texas, giant machines grow silicon ingots. Those ingots get sliced into thin wafers. They are polished and shipped in special containers. The goal is consistent, clean wafers at scale. Without that, fabs cannot run smoothly.
The “carving” happens at TSMC. In Arizona, TSMC is producing 4- and 5-nanometer chips. The campus plan is massive. TSMC said it plans to spend $165 billion on six chip plants and more. One fab is built and producing chips. A second is set to come online next year. A third is targeted for 2030, with more planned after. The site could become a company town. The Journal described apartments and even a Costco nearby. That is how big this is.
Packaging is the downstream gap the U.S. is trying to close. Chips are often packaged in Asia today. Near TSMC’s site, Amkor is building advanced packaging facilities in Arizona. The first is expected to complete in 2027. It will take wafers from TSMC, dice them into chips, and add connectors. Amkor has said it plans to spend about $7 billion on the site. Apple is investing, though the size was not specified. This matters because packaging is not glamorous, but it is essential.
Finally, assembly shows up in pockets. Apple and Foxconn run a Houston operation for AI servers. That line produces about 10 servers an hour. Apple also plans to assemble the Mac mini in Houston, using more than 200,000 square feet of space. Still, iPhone assembly remains abroad. Apple sells far more iPhones than Mac minis. That scale is the hard part. For now, the U.S. push is focused on silicon components and selected products.
Reality Check On Tech Gap, Scale & Timelines
Here is the part that gets lost in the patriotic headlines. The U.S. is still behind Taiwan on both volume and the most advanced technology. TSMC’s Taiwan operations produce more than 100,000 wafers per month at several facilities. Arizona will not match that until all six fabs are built. Even then, the ecosystem depth is different. Taiwan has decades of clustering. It has dense supplier networks. It has specialized labor. The U.S. is rebuilding that from scratch, one crane at a time.
The technology gap is also real. Taiwan makes the most advanced chips today, with transistors as small as two nanometers. Arizona currently makes 4- and 5-nanometer chips. TSMC has said two-nanometer production in Arizona is not expected until 2030. That is a long wait in chip years. Apple’s newest “brain” chips for iPhone and Mac tend to need the leading edge. Arizona can still make plenty of other chips. Devices have dozens of chips inside. Many use older nodes. Some advanced AI chips also use older process tech.
Constraints are not just about machines. They are about supply flexibility. On Apple’s earnings call, management said it was constrained by advanced-node capacity. It also said it was hard to predict when supply and demand would balance. That is a reminder. Domestic fabs do not create infinite capacity. They allocate scarce capacity. They also come with higher costs and slower ramp times. The U.S. also needs more packaging depth, more tooling, and more trained technicians. Those are solvable problems, but not fast ones.
So what is being reshored, really? Some wafer supply is moving to the U.S. Some leading-edge fabrication is starting in Arizona, though not at the newest nodes. Packaging is being built in Arizona, with 2027 as a key date. Some server and Mac mini assembly is moving to Houston. What is not being reshored is the iPhone assembly engine, and the cutting-edge node crown, at least for now. The U.S. effort is meaningful. It is not yet a full substitute for Asia.
Final Thoughts
Apple’s U.S.-based chip push has clear upsides. It adds redundancy. It reduces single-region risk. It can help with tariff uncertainty. It also nudges suppliers to invest locally, from wafers to packaging. The projects in Arizona and Texas look substantial. The Houston buildout adds a “made here” story for some products. And Apple’s order book makes these projects bankable.
The downsides are also clear. The U.S. will remain behind Taiwan for years on the most advanced nodes. The timeline matters. Two-nanometer in Arizona is not expected until 2030. Volume at Taiwan scale requires all six planned fabs. Packaging is improving, but it is still catching up. Costs may also be higher, which can pressure margins if not offset by mix and efficiency.
Investors should also keep valuation in mind. On trailing numbers, Apple recently traded around 8.85x LTM EV/Revenue, 25.20x LTM EV/EBITDA, and 33.76x LTM P/E (per the table provided, dated Feb. 23, 2026). Those are not bargain multiples. They suggest the market already prices in durability. A steadier supply chain can support that story. It does not guarantee it. The reshoring effort looks like a risk-management move first, with strategic benefits over time.
Disclaimer: We do not hold any positions in the above stock(s). Read our full disclaimer here.
