Opinion What the chip industry and the petroleum sector have in common

The supply of petroleum is dominated by the “cartel” OPEC and mega-sized public and private multinational companies, often referred to as the “super majors”.

The semiconductor value chain is comparably close-knit. The US is arguably the most powerful player. Every chip produced in the world has a direct or indirect connection with the country. The software for chips, for instance, is provided by three US based companies — Cadence, Synopsys and Mentor. ASML, the sole producer of the equipment EUV (Extreme Ultraviolet Lithography), is a Dutch company but dependent on its wholly-owned San Diego-based subsidiary Cymer for the manufacturing tools. Samsung and Hynix, which together produce 44 per cent of the world’s memory chips, and TSMC, which fabricates 37 per cent of the world’s logic chips and 92 per cent of the most advanced chips, are Korean and Taiwanese respectively. Both are protected by the US military security blanket.

The US is, however, not the only cog in the semiconductor life cycle. There are other critical cogs. Were TSMC’s fabrication facilities to fall into an earthquake fault or destroyed by military action, one third of the worlds computing power would grind to a halt, the 5G network would collapse and, according to Miller, the economic loss would be trillions of dollars.

Geopolitics is at the core of both industries. Every oil import dependent country has beaten a path to the Middle East to secure access to petroleum and at times “weaponised” their efforts to safeguard this objective. The Saudi embargo of exports to the pro-Israeli Western world in 1973; the US intervention in Iraq in 2003 and the current cutback of Russian gas to Europe are three examples of this phenomena.

Comparably, semiconductors have also been part cause and consequence of the “technology Cold War” between the US and China. The US has imposed sanctions on the physical and intellectual export of chip technology to China and President Xi has, in turn, called for a “full scale assault” to “rejuvenate” China. The full consequences of these actions have yet to play out but the world is clearly fragmenting along the fault lines of chip geopolitics.

Both petroleum and semiconductor chips have a global footprint. Crude oil is tradable because it is easy to store and ship. Gas on the other hand has a narrower market. It was earlier limited to the overland routes defined by the pipeline infrastructure but in recent years with the commercialisation of gas liquefaction, cryogenic shipping and regassification, this limitation has eased. One reason the price of gas in Europe has declined over the past weeks is because many LNG carriers destined for East Asia have been redirected to the regas terminals in Europe. Oil and gas prices are cyclical, reflecting the capital intensity and long lead times of the investment cycle.

The chip industry mirrors these economic dynamics. The value chain (design, equipment, fabrication and testing and assembly) straddles the globe. Investment to create part or all of this chain runs into billions and the returns depend on engineering precision and technical talent. Currently, efforts are underway to drive a wedge into this supply chain and to “reshore” fabrication facilities. This is because of the technology Cold War. The US has passed the Chips and Science Act and earmarked $52 billion for the creation of domestic chip fabrication. INTEL has laid the foundation stone for a $-20-billion fabrication plant. There are, however, economic constraints to how far this process can be sustained. Goldman Sachs has estimated that the cost of building a high-end semiconductor fabrication facility in the US could be up to 44 per cent more expensive than locating it in Taiwan, Korea or Singapore.

Finally, both industries are marked by technological change. The mistake made by the theorists of “peak oil” was the assumption that oil exploration and production technology would be linear and incremental in progress. They did not anticipate the cutting-edge innovations that would open up the hydrocarbon resources in deep waters and within the pores of shale rock. A similar mistake could be made in electronics. Those countries that decide to copy, steal and subsidise technology are, according to Miller, condemned to technological backwardness because this strategy will not allow them to keep pace with the speed of cutting-edge innovation.

India has struggled for more than five decades to reduce its dependence on external sources of petroleum supply. It has not been successful. India’s import dependence was around 20 per cent in the early Eighties. It is now more than 80 per cent. India has recently embarked on a journey to develop domestic chip fabrication facilities. Given the above commonalities the decision makers involved with electronics might draw useful lessons from our “failed” experience in petroleum. There are two lessons they should internalise. One, chip nationalism will be economically costly and could be technologically regressive. They should be cautious about decoupling from the international supply chain. And two, government support should be limited to financial support, nimble cooperation, and the creation of an innovative ecosystem. Bureaucratic intervention should be minimal.

The writer is chairman and distinguished fellow, Centre for Social and Economic Progress