The huge endeavor to produce a tiny microchip

Some feature more than 50 billion tiny transistors that are 10,000 times smaller than the width of a human hair. They are made on gigantic, ultraclean factory room floors that can be seven stories tall and run the length of four football fields.

Microchips are in many ways the lifeblood of the modern economy. They power computers, smartphones, cars, appliances and scores of other electronics. But the world’s demand for them has surged since the pandemic, which also caused supply-chain disruptions, resulting in a global shortage.

That, in turn, is fueling inflation and raising alarms in the United States that the nation is becoming too dependent on chips made abroad. The U.S. accounts for only about 12% of global semiconductor manufacturing capacity; more than 90% of the most advanced chips come from Taiwan.

Intel, a Silicon Valley titan that is seeking to restore its longtime lead in chip manufacturing technology, is making a $20 billion bet that it can help ease the chip shortfall. It is building two factories at its chipmaking complex in Chandler, Arizona, that will take three years to complete, and recently announced plans for a potentially bigger expansion, with new sites in New Albany, Ohio, and Magdeburg, Germany.

Why does making millions of these tiny components mean building — and spending — so big? A look inside Intel production plants in Chandler and Hillsboro, Oregon, provides some answers.

What chips do

Chips, or integrated circuits, began to replace bulky individual transistors in the late 1950s. Many of those tiny components are produced on a piece of silicon and connected to work together. The resulting chips store data, amplify radio signals and perform other operations; Intel is famous for a variety called microprocessors, which perform most of the calculating functions of a computer.

Intel has managed to shrink transistors on its microprocessors to mind-bending sizes. But the rival Taiwan Semiconductor Manufacturing Co. can make even tinier components, a key reason Apple chose it to make the chips for its latest iPhones.

Such wins by a company based in Taiwan, an island that China claims as its own, add to signs of a growing technology gap that could put advances in computing, consumer devices and military hardware at risk from both China’s ambitions and natural threats in Taiwan such as earthquakes and drought. And it has put a spotlight on Intel’s efforts to recapture the technology lead.

How chips are made

Chipmakers are packing more and more transistors onto each piece of silicon, which is why technology does more each year. It’s also the reason that new chip factories cost billions and fewer companies can afford to build them.

In addition to paying for buildings and machinery, companies must spend heavily to develop the complex processing steps used to fabricate chips from plate-size silicon wafers — which is why the factories are called “fabs.”

Enormous machines project designs for chips across each wafer, and then deposit and etch away layers of materials to create their transistors and connect them. Up to 25 wafers at a time move among those systems in special pods on automated overhead tracks.

Processing a wafer takes thousands of steps and up to two months. TSMC has set the pace for output in recent years, operating “gigafabs,” sites with four or more production lines. Dan Hutcheson, vice chair of market-research firm TechInsights, estimated that each site can process more than 100,000 wafers a month. He estimated the capacity of Intel’s two planned $10 billion facilities in Arizona at roughly 40,000 wafers a month each.

How chips are packaged

After processing, the wafer is sliced into individual chips. These are tested and wrapped in plastic packages to connect them to circuit boards or parts of a system.

That step has become a new battleground, because it is more difficult to make transistors even smaller. Companies are now stacking multiple chips or laying them side by side in a package, connecting them to act as a single piece of silicon.

Where packaging a handful of chips together is now routine, Intel has developed one advanced product that uses new technology to bundle a remarkable 47 individual chips, including some made by TSMC and other companies as well those produced in Intel fabs.

What makes chip factories different

Intel chips typically sell for hundreds to thousands of dollars each. Intel in March released its fastest microprocessor for desktop computers, for example, at a starting price of $739. A piece of dust invisible to the human eye can ruin one. So fabs have to be cleaner than a hospital operating room and need complex systems to filter air and regulate temperature and humidity.

Fabs must also be impervious to just about any vibration, which can cause costly equipment to malfunction. So fab clean rooms are built on enormous concrete slabs on special shock absorbers.

Also critical is the ability to move vast amounts of liquids and gases. The top level of Intel’s factories, which are about 70 feet (21 meters) tall, have giant fans to help circulate air to the clean room directly below. Below the clean room are thousands of pumps, transformers, power cabinets, utility pipes and chillers that connect to production machines.

The need for water

Fabs are water-intensive operations. That’s because water is needed to clean wafers at many stages of the production process.

Intel’s two sites in Chandler collectively draw about 11 million gallons (about 42 million liters) of water a day from the local utility. Intel’s future expansion will require considerably more, a seeming challenge for a drought-plagued state like Arizona, which has cut water allocations to farmers. But farming actually consumes much more water than a chip plant.

Intel says its Chandler sites, which rely on supplies from three rivers and a system of wells, reclaim about 82% of the freshwater they use through filtration systems, settling ponds and other equipment. That water is sent back to the city, which operates treatment facilities that Intel funded, and which redistributes it for irrigation and other nonpotable uses.

Intel hopes to help boost the water supply in Arizona and other states by 2030, by working with environmental groups and others on projects that save and restore water for local communities.

How fabs are built

To build its future factories, Intel will need roughly 5,000 skilled construction workers for three years.

They have a lot to do. Excavating the foundations is expected to remove 890,000 cubic yards (about 680,000 cubic meters) of dirt, carted away at a rate of one dump truck per minute, said Dan Doron, Intel’s construction chief.

The company expects to pour more than 445,000 cubic yards of concrete and use 100,000 tons of reinforcement steel for the foundations — more than in constructing the world’s tallest building, the Burj Khalifa in Dubai.

Some cranes for the construction are so large that more than 100 trucks are needed to bring the pieces to assemble them, Doron said. The cranes will lift, among other things, 55-ton chillers for the new fabs.

Patrick Gelsinger, who became Intel’s CEO a year ago, is lobbying U.S. Congress to provide grants for fab construction and tax credits for equipment investment. To manage Intel’s spending risk, he plans to emphasize construction of fab “shells” that can be outfitted with equipment to respond to market changes.

To address the chip shortage, Gelsinger will have to make good on his plan to produce chips designed by other companies. But a single company can do only so much; products like phones and cars require components from many suppliers, as well as older chips. And no country can stand alone in semiconductors, either. Though boosting domestic manufacturing can reduce supply risks somewhat, the chip industry will continue to rely on a complex global web of companies for raw materials, production equipment, design software, talent and specialized manufacturing.

This article originally appeared in The New York Times. © 2022 The New York Times Company