Kristoffer Tripplaar/Sipa USA via AP Images
The offices of semiconductor chip maker GlobalFoundries in Santa Clara, California
What do Uber and GlobalFoundries, the world’s largest U.S.-based semiconductor fabricator, have in common? They’re both unprofitable companies, and they’re both propped up by Gulf petrostates seeking to gain political access.
Semiconductor chips are found in almost every electronic product, and while their manufacturing infrastructure was once close to invisible, COVID supply chain breakdowns have made Americans far too aware of it, amid the increasing shortages of affordable products. But while Saudi Arabia’s role in propping up more prominent unprofitable companies like Uber has been widely reported, Gulf States’ investments in semiconductor firms is less well known, though just as hazardous.
In the 2000s, fearing the long-term decline of the oil economy, UAE and Saudi leadership both released 2030 plans that involved “diversification.” As an IEEE publication notes, “Abu Dhabi’s major global investment company, Mubadala, played a leadership role in implementing Vision 2030 and began to invest strategically in the diversification of the economy.”
It’s clear that the investment in the tech industry has already afforded the Saudis, and particularly their leader Mohammed bin Salman, access to the tech elite, though at great cost to our economic balance. It also notably led members of this elite—who work closely with the U.S. security state—to provide cover for Saudi crimes, such as the murder of Jamal Khashoggi. The seemingly endless subsidization of Uber not only made it seem profitable, but encouraged other companies to follow its model. This intervention distorted our transportation market, putting many in the taxi industry into massive debt, and diminished demand for less carbon-intensive transportation options.
The U.S. ought to think very carefully about the extent of Gulf petrostate money and its role in distorting our tech industry from the bottom up.
Could we be headed down a similar path with semiconductors? Especially in light of Russian aggression in Ukraine and recent reporting on Saudi and Emirati support for that effort, the U.S. ought to think very carefully about the extent of Gulf petrostate money and its role in distorting our tech industry from the bottom up.
According to their website, Mubadala controls $894 billion in assets. They boast investments in aerospace, agribusiness, health care, information and communications technology, metals and mining, real estate and infrastructure, petroleum and petrochemicals, pharmaceuticals and medical technology, and renewables and utilities. Outside of the U.S., these investments are almost entirely in oil and energy–related businesses (with the exception of two commodities companies). In the U.S., however, they branch into many crucial sectors. Many of these companies, like GlobalFoundries, are not good investments or leaders in their fields, but the Emiratis are apparently more interested in political dividends than monetary ones.
GlobalFoundries, which claims it is the “world’s largest U.S.-based, independent semiconductor foundry with a global manufacturing footprint,” is owned by the Emiratis through Mubadala. The government investment fund evens out GF’s balance sheets, giving it the appearance of profitable stability. In reality, GlobalFoundries only produces two million wafers a year—not nearly enough to make them a competitive player. To make a profit, a foundry would need to produce at least ten million.
The company also has an extremely low utilization rate, which refers to the percentage of chips produced that are usable. GlobalFoundries was at 84 percent in 2020, their best number in recent years, up from 80 percent in 2018 and 70 percent in 2019.
By contrast, TSMC, the world’s leading semiconductor manufacturer, has a utilization rate of over 95 percent. Since the industry is extremely capital-intensive, and margins are small even for the best companies, GlobalFoundries’ rate is abysmal. Other analyses have come to the same conclusions.
GlobalFoundries is a so-called “Trusted Foundry” site for production of chips intended for use by the Department of Defense. Though they claimed to be in the race for leading-edge chips when they bought IBM’s foundries in upstate New York (now the basis of a breach-of-contract suit), these are not cutting-edge chips like those produced by TSMC. Rather, GlobalFoundries produces the general-use chips ubiquitous in almost every consumer (and defense) product.
The company claims it is indispensable, because most semiconductor production takes place in East Asia, using this rationale to pitch building domestic foundries that can soak up European and American subsidies. However, the majority of GlobalFoundries production by wafer volume is in Singapore. Subsidizing this company would solve neither supply chain issues nor national-security ones.
GlobalFoundries has so far received an incredible $698 million in government grants, plus $22.4 billion from the Emiratis. Yet even according to their own distorted financials, they have lost money every year except 2021, which seems largely due to the largest demand for chips in our lifetimes, paired with U.S. and EU subsidies. If demand subsides, the company seems unlikely to stay in the black, given the reality that they are not serious players in the race for cutting-edge chips and have continued to reduce R&D spending.
Despite this gloomy outlook, GlobalFoundries chose this moment to go public, hoping to boost the company’s balance sheet. The IPO took place last fall, and when it turned out less successful than anticipated (despite occurring during a global chip shortage), Mubadala made up the difference, selling only a small portion of the company.
The Saudis have also announced a number of new tech ventures. Mostly recently, in early February they announced a $6.4 billion investment in future technologies, including the metaverse and the blockchain. These two technologies, like Uber, seem to be overinflated and without significant use cases. Previously, the Saudis were also major (in 2017, $4.4 billion) investors in the memorably successful company WeWork. SoftBank, a major Uber and WeWork investor among others, has also received billions from the Saudis. The fact that it was major news in the aftermath of Khashoggi’s murder that the CEOs of U.S. tech firms refused to go to Riyadh’s global investment forum testifies to the oversized influence they enjoy.
Unlike China, which uses indirect market power, the Gulf States use direct investment to exert direct control.
While the Saudis opt for flashy, more obvious distortions of the industry and purchase of political influence, the Emiratis are in the same game though with more seemingly infrastructural investments. It seems that, despite extensive rhetoric about diversification, neither the Saudis nor the Emiratis have actually diversified their economies with these moves. Rather, they have bought access and leverage instead, forcing U.S. tech companies to become beholden to Saudi and Emirati political power via investment largesse. Unlike China, which uses indirect market power, the Gulf States use direct investment to exert direct control.
Since GlobalFoundries’ claim to be a competitive semiconductor manufacturer has become increasingly unbelievable, it has now adopted the line that it is a leading “specialty” chip manufacturer. This claim is easily challenged. It does, however, provide a glut of jobs to the upstate New York area and claims to be “one of the largest public/private partnerships in New York’s history.” This gives it significant leverage in state politics, and over Senate Majority Leader Chuck Schumer—a fact that the company is unafraid to signal in a less than subtle manner. The desire amid shortages and a lack of economic security to reshore chips and bring back American manufacturing makes this trend even more salient.
Before being purchased from IBM, these upstate New York chip factories had been the site of a kind of fortress model (i.e., wholly controlled by the U.S. security state). But because they made so few chips, they never emerged as cutting-edge production facilities. The sale of this foundry in 2015 was hailed by UAE boosters as “one of most important investments for the UAE in global markets in terms of value and quality high technology” that “underscores American confidence in the UAE as a trusted partner.” The sale also caused the Pentagon to panic about the availability of onshore chip manufacturers.
The CHIPS Act, pushed strongly by Schumer and now part of a bipartisan competitiveness bill being negotiated in Congress, invests $52 billion in reshoring U.S. semiconductor capacity. While some of the CHIPS subsidies are going to projects that seem sensible, like Samsung fabrication facilities in Texas and TSMC in Arizona, fundamentally unprofitable and foreign-owned companies like GlobalFoundries are also reaping the benefits.
The bill’s sponsors conflated two major issues: Intel no longer being at the leading edge of semiconductor manufacturing (instead, TSMC and Samsung dominate), and COVID-induced supply chain issues. Industry and national security state officials are very concerned about the former, while the latter has a greater impact on consumers. The national security state has worked extremely closely with the semiconductor industry for almost all of its existence, sometimes blurring the lines between state and industrial sovereignty—similarly to how Adam Tooze describes networks of central bankers in his recent book Crashed.
If Congress’s priority is planning for future chip shortages and shoring up the supply chain, it should encourage and subsidize more projects in the U.S. like the TSMC and Samsung factories. If, on the other hand, they are more concerned about secure access to leading-edge chips (a DOD priority), they should consider more serious long-term investments in their own production facilities. Rather than a one-shot payment, this would require billions of dollars a year. In addition to being designed so that it encourages buybacks rather than investment, experts generally think that the CHIPS Act is too small for the U.S. to regain the edge in semiconductor manufacturing.
Because GlobalFoundries does most of its manufacturing in Singapore, its receipt of U.S. subsidies may not address the national-security issue of geographic concentration of chip plants in East Asia. Similarly, even the more sensible companies like TSMC and Samsung are unlikely to put their leading-edge facilities in the U.S. for a variety of reasons, including the possibility of being kicked out of the country for political or nationalist reasons by a future president like Donald Trump.
The reason for the flaws in the CHIPS Act, and Congress’s inability to legislate well around tech more broadly, goes back to the Gingrich-controlled House. In 1995, the Gingrich Congress gutted the Office of Technology Assessment, which left Congress relying on national security state actors and lobbyists to make tech policy. Nearly three decades later, America is faced with a critical domestic chip shortage, and anemic high-tech factories propped up by foreign money, which are primarily paying dividends of petrostate influence, not economic security.
Ilhan Omar is right: “It’s past time we dispense with the fiction that the United Arab Emirates are a reliable security partner for the United States. Between the allegations of their passing U.S.-origin weapons on to terrorist groups in Yemen to their refusal to condemn Putin’s illegal war in Ukraine, it is clear time and again that they do not share our values or our sense of security.” The U.S. should not waste this opportunity to improve both our tech sector and national security, and to turn away from petro-authoritarian states with which we do not share interests.
