As we noted in a recent article, “Semiconductor shortage: How the automotive industry can succeed,” demand for automotive chips is skyrocketing as vehicle technology becomes more sophisticated.1 Some OEMs have been forced to shut down or slow production because they could not obtain sufficient quantities of semiconductors, and our article noted that shortages will probably continue over the next few years based on our analysis. With no end in sight to the supply–demand mismatch, we recently took a closer look at the automotive-chip market to get a better understanding of the forces in play.
Our analysis showed that overall revenues for automotive chips could rise from $41 billion in 2019 to $147 billion by 2030 (Exhibit 1). Three areas—autonomous driving, connectivity, and electrification—will drive most of the demand, accounting for $129 billion in revenues, or about 88 percent of the total.
In a related development, our analysis showed that annual demand for 12-inch–equivalent automotive wafers could increase from about 11 million in 2019 to 33 million by 2030—a CAGR of 11 percent. Continuing the pattern seen today, most future automotive-wafer demand will involve nodes of 90 nanometers (nm) and above because many vehicle controllers and electric powertrains, including electric drive inverters and actuators, rely on these mature chips. Such nodes will account for about 67 percent of automotive demand in 2030 (Exhibit 2).
Semiconductor companies are increasing production of 90 nm chips, but our analysis suggests that the CAGR will remain at only 5 percent or so from 2021 through 2026—not enough to eliminate the supply–demand mismatch. What’s more, OEMs that rely on 90 nm chips for many applications have little incentive to migrate to smaller nodes, because the shift would require additional development and qualification costs, as well as more R&D staff. Those disadvantages often outweigh the technological benefits. (Since drive inverters and actuators require high voltages and currents, chips used in these applications don’t benefit from the high transistor density characteristic of smaller node sizes.)
Of course, OEMs do sometimes need leading-edge chips—for instance, to enhance autonomous-driving systems significantly. These chips have higher CAGRs (about 9 percent from 2021 through 2026) than mature nodes do. But OEMs may still have trouble obtaining sufficient quantities because cross-industry competition is intense: high-tech, consumer electronics, and other companies all want leading-edge chips to improve the performance of their products and are willing to pay a premium to ensure supply. The supply–demand mismatch will therefore persist across all node sizes.
Our previous article discussed many potential solutions to the chip shortage, such as creating better technology road maps and improving short- and long-term demand planning. One overriding theme was the need for greater cross-industry collaboration along the automotive supply chain. For instance, tier-one suppliers might hold discussions with OEMs when creating technology road maps to identify opportunities for drop-in components that can be used in multiple vehicles. They could also jointly invest in projects with OEMs to share the financial burden of creating mature or leading-edge node designs and manufacturing capacity—a strategy that both reduces costs and increases supply. In addition, OEMs and tier-one suppliers may be able to ensure a more reliable supply of automotive chips by working more closely with semiconductor companies on demand forecasting and other activities. To ensure the best possible solutions are found, semiconductor companies will likely benefit from thinking about such collaborations now.
