Understanding CHIPS, Part One: The Semiconductor Manufacturing Challenge

In recent decades, the United States has come to rely heavily on foreign manufacturers for advanced semiconductors today produces only 10% of global chip capacity, down from 37% in 1990. A 2022 bipartisan law, the CHIPS and Science Act, creates opportunity to reverse the trend and bring chip production back to the U.S. This is a win for our economy and national security–the lack of domestic production capacity has been identified by American lawmakers as a significant threat to national competitiveness.

The CHIPS and Science Act is the largest investment in the American semiconductor industry in decades, providing $52.7 billion in funding for semiconductor manufacturing, R&D, and workforce training. After three decades of decline, the CHIPS and Science Act affords a once-in-a-generation opportunity to restore U.S. competitiveness in manufacturing, but it’s a long road ahead to accomplish the goals of the legislation by 2030.

The U.S. Department of Commerce, through the National Institute of Standards and Technology (NIST), launched the first semiconductor related funding opportunity created by the CHIPS and Science Act in February 2023. Applicants selected for the program will receive the awards in the form “of direct funding, federal loans and/or federal guarantees of third-party loans.”

This funding mechanism is not to be confused by another incentive created by the CHIPS and Science Act, the Advanced Manufacturing Investment Credit, administered by the U.S. Department of the Treasury and the Internal Revenue Service. This tax credit, already available to companies, is equal to 25% of qualifying investments in facilities manufacturing semiconductors or semiconductor manufacturing equipment.

Now that implementation of the CHIPS and Science Act is in full swing, it’s critical that policymakers understand the complex manufacturing process behind semiconductors as well as the current state of the domestic industry. Semiconductor chips are the basis of many products we use in our daily lives as well as an enormous range of electronics and machines—from mobile phones to Javelin missiles. They play a vital role across industry while enabling technological advancements and driving innovation. In a way, they are the foundation of modern competitiveness. But what does it take to build a cutting-edge chip, and what does that mean for policymakers?

How Are Semiconductor Chips Made?

Source: ASML, Intel, ASE Group

Semiconductor manufacturing is extremely complex and capital-intensive. First, silicon dioxide (silica sand) is heated with carbon to form a block of pure silicon, or an ingot. Ingots are then sliced into wafers to become the basis of the semiconductor chip. Wafers can vary in size with each containing between dozens and thousands of chips.

The next step is “front-end manufacturing” at a semiconductor fabrication plant (more commonly known as a “fab”). The wafer then proceeds along a complicated process which is, in summary:

1. Deposition: Wafers are polished and a blanket of conducting, semi-conducting, or isolating materials is deposited on its surface.
2. Coating: A light-sensitive coating called photoresist is applied.
3. Lithography: The wafer then undergoes photolithography, in which light hits the photoresist and creates a microscopic pattern. This requires extremely specialized lithography systems.
4. Etching & Doping: The “printed” pattern is physically cut into the wafer and ions are implanted to better conduct the flow of electricity. Now, transistors (switches that turn electric currents on and off) can be created.
5. Packaging: Wafers are cut down to individual chips and then placed onto a substrate to direct input/output signals. Packaging is complete when a heat spreader is placed over the chip to protect the fragile product while dissipating heat buildup.

Many of these steps contain hundreds of precise sub-processes, and the entire sequence is repeated to increase the number of transistors (and the complexity of the chip). By the time these steps are completed, there can be up to billions of transistors on chips the size of coins—Apple’s M1 Ultra, for example, has a mindboggling 114 billion transistors. At this point, the chip is tested, packaged, and assembled so that it can be mounted in a printed circuit board and integrated into a product.

These subprocesses often require highly specialized and expensive equipment. For example, photolithography machines can cost more than $300 million apiece. Not to mention the power bill; each machine can consume as much electricity as a thousand homes. Machinery, energy, and other process costs contribute to the enormous price tag for building and maintaining semiconductor fabs. Intel estimates that a single fab can take more than $10 billion and three years to complete. Other nations’ government subsidies typically focus on supporting the development, construction, and operating costs of these fabs.

Nanometers Matter

From a competitiveness standpoint, the quality, such as feature size, of semiconductor technology is as important as manufacturing capacity. Feature size refers to the size of the individual components or transistors on a microchip. As feature size decreases, the number of transistors that can be placed on a chip increases, leading to more processing power and improved efficiency. This has had a significant impact on competitiveness in the technology industry as smaller feature sizes have enabled faster and more advanced electronics to be developed. Additionally, the ability to pack more transistors into a smaller area also leads to smaller and lighter devices, making them more appealing to consumers. Overall, reduction in feature size has played a key role in driving innovation and competitiveness in the semiconductor and technology industries.

The most advanced semiconductors are critical components for consumer electronics like smartphones and laptops and underpin technologies like artificial intelligence and cloud computing. These technologies are not just important for the growth of America’s innovation economy, but also enable essential military equipment.

Increasingly advanced semiconductor generations are called process nodes. Each process node represents a set of feature sizes; the smaller a chip’s features are, the more transistors it can contain, allowing it to execute more complex tasks. While the names of these nodes (e.g., 250 nanometers, nm) were historically based on approximate distances between parts of the chip, they are more arbitrary “branding” specifications today. In 2022, the most advanced mass-produced process node was the 5 nm node, with larger-scale 3 nm manufacturing expected soon. On the other hand, chips designed to more mature process nodes (e.g., 90 nm) are called legacy chips.

Implications for Policymakers

Source: Semiconductor Industry Association

Due to semiconductors’ importance in the electronics supply chain and the cost of manufacturing, many countries have subsidized domestic industries with tax credits and direct funding for R&D and/or manufacturing. In many cases, these efforts are specifically aimed at the most advanced process nodes. The semiconductor manufacturing process is complex and unforgiving but a modern requirement for national competitiveness. The CHIPS and Science Act represents a step toward ensuring that the supply of this essential technology is secure.

The next blog post in this series will describe how the recently enacted law offers a framework to renew American competitiveness in semiconductor manufacturing.