Through the CHIPS Act, as well as tax and trade policies, both the Biden and the Trump administrations have implemented federal policies to encourage domestic semiconductor manufacturing. From 2021 through 2024, semiconductor and electronics companies invested an estimated $450 billion.
But capital, while necessary, is not sufficient. Producing semiconductors requires a multitude of chemicals and materials; if a single input is missing or late, the whole system falters. For example, fabs use hundreds of different chemicals, materials and gases, including ultra-high-purity chemicals (parts-per-trillion). These chemicals are used to achieve the core processing steps for semiconductor manufacturing: adding layers of metal and metal oxide; etching these layers or cleaning the wafer surface.
The US semiconductor industry market is expected to almost double in value, to $140 billion, by 2030; demand for chemicals will rise as much or more. The problem, according to a McKinsey analysis of US semiconductor supply chains, is that for more than half of them, there is not enough domestic supply to meet that level of demand. For chemicals, supply gaps could be serious by 2030.
All this would have ripple effects across the economy because critical sectors, such as cars, electronics, data storage, and telecoms, depend on semiconductors. The ongoing Fourth Industrial Revolution (4IR) is transforming manufacturing in particular and business in general. Smart computers and connected devices all need chips. That is a big part of the case for increasing domestic production. But doing so will be complicated and expensive. There is no one-size-fits-all approach.
Chemical mechanical planarization (CMP) pads, for example, are widely available and can be made cost competitively in the United States. But almost all ultra-high-purity hydrogen fluoride (HF), which is used to etch away silicon dioxides, is imported from Asia. That makes US manufacturers vulnerable—and building domestic capacity is difficult because of high capital and labor costs. In other cases, such as the separation and purification of global cerium ores, the United States faces technological or transportation challenges.
Stable supply chains are critical to semiconductor production. As these examples demonstrate, though, closing or at least narrowing the chemicals will require a range of solutions, including trade agreements, joint ventures, re-shoring, and finding new upstream options. Companies need to think strategically about resilience and ensure supply risks are factored into investment decisions.
Building a stronger supply base can also mean addressing related issues that are not directly about chemicals or the supply chain but that still make a difference. One has to do with talent. Last year, McKinsey estimated that the US semiconductor industry could see an additional $250 billion in investment. The good news is that this would translate into something like 160,000 new jobs in in engineering, technician support , construction, and the trades. The bad news is that at the moment the talent pipeline is not close to being able to fill these jobs. About 1,500 engineers a year are joining the semiconductor industry; by 2029, though, it could need 88,000.
There are a range of public, private, union, and company programs in the works or working. But even if these all function as intended, the talent gap would still be significant. Clearly, there needs to be more of them. In addition, semiconductor companies themselves can step up, doing more to reduce attrition, to re-train current staff, and to make themselves more attractive to new graduates. They could also think more broadly and seek out talent outside the traditional talent pools. One approach is to look for new technicians with transferable skills from adjacent industries such as pharmaceuticals and advanced manufacturing. Recent immigrants and military veterans could also be promising sources of untapped talent.
Another priority is to improve operations in order to make fabs as efficient as possible. More reliable equipment can improve tool availability (the share of time that a piece of equipment is ready to work) by 15 percent. That allows it to increase capacity without expanding its footprint. To make that happen requires prioritizing equipment recovery, planning maintenance, and manage parts efficiently. None of this is glamorous. All if it helps to expand production without adding people or plant.
No question: it is a tricky time for the industry. As technology historian Chris Miller told McKinsey late last year, “The industry is balancing the need to retain highly efficient international supply chains with risks of internationalization, including cross-border tariffs and regulations.”
By 2030, semiconductors could be a $1 trillion industry, with fabs producing a wider and more complex variety of chips and designs. The United States could account for 30 percent of advanced-node semiconductor fabrication capacity. But that potential will not become a reality if fabs cannot get the chemicals and materials they need when they need them. That is why establishing a reliable, resilient supply matters—and why the effort to do so must start now.
Guttorm Aase is a Partner in McKinsey & Company’s New York office. Chris Musso is a Senior Partner in Denver. Mark Patel is a Senior Partner in the Bay Area.