What Intel fabrication plants do and why they matter

Intel fabrication plants, called fabs, are the factories where Intel designs and manufactures the processors that power computers, servers, and data centers. A fab is not a straightforward assembly line — it is a highly controlled environment where raw silicon wafers are transformed into finished chips through hundreds of precise steps. Each fab costs billions of dollars to build and requires constant upgrades to stay competitive as chip designs become more complex.

Intel operates fabs in multiple countries, and the location of a fab affects which chips get made there, how long production takes, and what the final cost is. Understanding where Intel manufactures chips and what each facility does helps explain why chip shortages happen, why some products cost more than others, and how geopolitics influences the semiconductor industry.

Key Takeaways

  • Intel operates fabrication plants in the United States, Ireland, and Israel, with each fab specializing in different chip sizes and production volumes.
  • The manufacturing process inside a fab involves depositing, etching, and testing silicon wafers through dozens of steps that take weeks to complete.
  • Fab capacity directly affects how many chips Intel can produce, so when demand spikes or a fab shuts down for maintenance, chip availability drops.
  • Intel has announced plans to build new fabs in Arizona, Ohio, and Germany as part of a strategy to reduce dependence on overseas manufacturing.

Intel's current fabrication plants and their locations

Intel operates fabs in three countries. In the United States, Intel runs fabs in Arizona, New Mexico, and Oregon. The Arizona facility in Chandler is one of Intel's largest and most advanced plants. The New Mexico fab in Rio Rancho produces older chip designs. The Oregon fab in Hillsboro is Intel's oldest active facility and historically served as a testing ground for new manufacturing techniques.

Outside the United States, Intel operates a major fab in Kildare, Ireland, which produces chips for European and global markets. Intel also runs a fab in Kiryat Gat, Israel, which manufactures processors and memory chips. Each location has different capabilities — some fabs are equipped to make the newest, smallest chips, while others focus on producing chips using older, proven processes that still have strong market demand.

The choice of which fab produces which chip depends on the chip's design size (measured in nanometers), the volume needed, and the fab's current workload. A fab running at full capacity cannot take on new orders, which is why Intel sometimes announces it will shift production of certain chips to a different location.

How chip manufacturing happens inside a fab

The process of turning raw silicon into a finished chip involves dozens of steps and takes several weeks from start to finish. The process begins with a silicon wafer — a thin, flat disk of pure silicon about the size of a dinner plate. Workers place the wafer into a machine that deposits a thin layer of material on top, then uses light and chemicals to etch away parts of that layer, creating patterns. This cycle repeats many times, building up layers of different materials.

Each layer serves a purpose: some conduct electricity, some block it, and some store data. The patterns on each layer must align perfectly with the patterns on the layers below and above, or the chip will not work. This is why fabs use extreme ultraviolet light and other advanced tools — the features being etched are so small that visible light cannot even see them.

After all layers are complete, the wafer moves to a testing area where machines check whether the circuits work correctly. Defective chips are marked and discarded. The remaining chips are cut from the wafer, packaged in plastic or ceramic cases, and shipped to computer makers. A single wafer can contain hundreds of chips, but not all of them pass testing — the percentage that pass is called the yield, and improving yield is a constant focus for fab engineers.

Why fab location and capacity matter for chip availability

When a fab is running at full capacity, Intel cannot produce more chips even if customers want them. This is why chip shortages happen — not because fabs break down, but because demand exceeds the total capacity of all fabs combined. During the pandemic and the cryptocurrency boom, demand for chips spiked, and fabs could not keep up.

Fab location also matters because moving production from one fab to another takes time. If Intel decides to shift production of a certain chip from Arizona to Ireland, the fab in Ireland must first be set up with the correct equipment and processes, which can take weeks or months. During that transition, production may slow or stop entirely.

Maintenance and upgrades also reduce capacity temporarily. Fabs must periodically shut down sections for cleaning, equipment replacement, or software updates. A major upgrade can take a fab offline for days or weeks. Intel schedules these shutdowns during periods of lower demand when possible, but they still reduce the total number of chips the company can produce.

Intel's plans to expand fabrication capacity

Intel has announced plans to build new fabs in Arizona, Ohio, and Germany over the next several years. The Arizona expansion will add capacity for advanced chip designs. The Ohio fab, located near Columbus, is intended to produce chips for the U.S. government and defense contractors. The Germany fab, in Magdeburg, is part of Intel's strategy to increase manufacturing in Europe.

These expansions are driven by several factors. First, the U.S. government has provided subsidies through the CHIPS and Science Act to encourage domestic chip manufacturing. Second, geopolitical tensions have made companies and governments nervous about relying too heavily on fabs in Asia. Third, Intel faces intense competition from other chip makers and needs more capacity to meet demand and win new customers.

Building a new fab is extremely expensive — each facility costs $10 billion to $20 billion — and takes several years from groundbreaking to first production. Intel has stated that some of these new fabs will not reach full production until the mid-2020s.

How fab technology changes over time

Chip designs shrink over time, meaning the features etched onto silicon become smaller and closer together. Intel measures this in nanometers — a smaller number means smaller features and more transistors on the same piece of silicon. Moving from one size to the next (for example, from 7 nanometers to 5 nanometers) requires new equipment, new chemical processes, and extensive testing.

Not all fabs can handle the newest sizes. A fab built to produce 28-nanometer chips cannot be easily converted to produce 5-nanometer chips — the equipment is different, the processes are different, and the building itself may need modifications. This is why Intel maintains fabs at different technology levels: some produce the newest designs, while others produce older designs that still have strong demand and do not require the most advanced equipment.

As technology advances, older fabs become less profitable because they cannot compete on performance or power efficiency. Intel sometimes sells older fabs to other companies or converts them to produce specialty chips. This allows Intel to focus its newest fabs on the most advanced and profitable products.

The relationship between fab location and geopolitics

Where chips are manufactured has become a matter of national security and trade policy. The United States, Europe, and China all want to reduce their dependence on fabs located in other regions. The U.S. government has restricted the sale of advanced chip-making equipment to China, which has pushed China to invest heavily in its own fabs. Europe has launched initiatives to increase chip manufacturing on the continent.

Intel's expansion into Ohio and Germany reflects these geopolitical pressures. By building fabs in the United States and Europe, Intel positions itself to serve government contracts and reduces the risk that trade restrictions or political conflict will disrupt its supply chain. At the same time, Intel must balance these political considerations against the cost and complexity of operating fabs in multiple countries with different labor costs, regulations, and tax policies.

Frequently Asked Questions

What is the difference between a fab and a foundry?

A fab is a manufacturing plant owned by a chip company like Intel that makes chips designed by that company. A foundry is a fab that makes chips designed by other companies — Taiwan Semiconductor Manufacturing Company (TSMC) is the world's largest foundry. Intel operates fabs, not foundries, though Intel has discussed offering foundry services to other chip designers in the future.

Why does it take so long to build a new fab?

Building a fab requires constructing a massive facility with extremely precise environmental controls, installing billions of dollars in specialized equipment, training workers, and testing every system. The building itself must be designed to prevent vibrations, dust, and temperature changes that could ruin chips. From planning to first production typically takes three to five years.

Can Intel move production of a chip from one fab to another quickly?

Moving production takes weeks to months because the receiving fab must be configured with the correct equipment, processes, and quality controls. Intel must also validate that chips made at the new location meet the same specifications as chips made at the old location. During the transition, production may be slower than normal.

What happens to chips that fail testing in a fab?

Chips that do not pass testing are discarded and cannot be sold. The percentage of chips that pass testing is called yield. Improving yield is a major focus for fab engineers because higher yield means more usable chips from each wafer, which reduces manufacturing cost and increases profit.

Does Intel own all of its fabs, or does it lease some?

Intel owns and operates all of its major fabs. This is different from some other chip companies that use foundries owned by other companies. Owning its fabs gives Intel control over production schedules and technology, but it also means Intel bears the full cost of building and maintaining these expensive facilities.