Last Friday, Intel CEO Pat Gelsinger again demonstrated his political prowess by hosting President Joe Biden for the groundbreaking of Intel’s $20 billion chip plant in Ohio. After the president promised thousands of jobs from the project, Gelsinger heralded the advent of the “Silicon Heartland.” That may be some distance ahead for the region, but in the meantime, Intel itself has both great opportunities and great challenges.
As noted previously, Intel has fallen behind its two major rivals, Samsung and Taiwan Semiconductor Manufacturing Company (TSMC). Gelsinger has taken bold action not only to catch up technologically but also to compete on manufacturing as well as the design of chips for customers around the world. He and his team argue that the staggering capital costs of constructing and operating a semiconductor plant (fab) make it imperative for Intel to go beyond its own needs and profit from chip sales to other tech companies. Thus, despite skepticism from Wall Street (Intel’s stock share prices have fallen 45 percent since Gelsinger became CEO), the company has embarked on a highly ambitious strategy of building new fabs or expanding existing facilities in the US and Europe. These include, as examples, $20 billion for construction of two new fabs at an existing site in Arizona, the aforementioned $20 billion project in Ohio (slated to expand to $100 billion over the next decade), and a $3.5 billion expansion of its advanced packaging facility in New Mexico.
There are also parallel plans for Europe, including an initial $17 billion fab in Germany, a $13 billion upgrade of an existing facility in Ireland, a $5 billion packaging plant in Italy, and a major research and development center in France. In total, Gelsinger has pledged to invest almost $90 billion in Europe over the next decade.
Intel, along with Samsung and TSMC, has made clear that its plans for future fab investment are contingent upon substantial public subsidy packages. Both the US and EU have now obliged: The Chips and Science Act in the US provides some $38 billion in grants and loans ($28 billion for advanced chips and $10 billion for less advanced chips), plus a separate investment tax credit, and the EU has allocated almost $50 billion to build a European chip industry. While these seem like outsized sums, these will not come close to covering construction and operating costs given the cost of new fabs ($10 billion and up) and the fact that there will be a number of qualified applicants for funds in both the US and EU well beyond the requests of the three giants (Micron and Infineon, for example). The public funds per fab will simply not cut it, particularly as new projects emerge.
Intel had hoped that in addition to public funds, it could finance its big future projects through cash flow and a backup in cash savings. This proved illusory, and recently Intel has stunned the tech world with a financing deal to sell a 49 percent stake in two new fabs under development in Arizona in return for up to $30 billion of equity from Brookfield to support the new plants. The unprecedented deal reflects the reality that Intel’s current capital spending, which has averaged 20 percent of its revenue over the past decade, could rise to 35 percent absent public aid or novel arrangements such as the Brookfield deal. The financing deal will free up cash for Intel’s future fab deals, but it is not clear if it will set a pattern for other companies. All of this comes as the notoriously boom-and-bust semiconductor industry seems poised for a sharp downturn. So Intel’s high-wire corporate agility will continue to be tested.
Finally, while Intel’s challenges stem in part from the company’s particular decisions and history, the US government—and other governments hoping to foster viable semiconductor industries—will face daunting technological and investment issues in coming years. As the former chip design firm ARM’s CEO Simon Segars has warned: “You’re going to have to spend a lot of money and you’re going to have to spend it over a long period of time. Whether governments have the stomach for that over the long term remains to be seen.”
The post Intel’s High Wire Act, Part II: Soaring Capital Costs appeared first on American Enterprise Institute - AEI.