The “Sputnik Moment” for the Chinese Semiconductor Industry
By Henry Hing Lee Chan – IPP Review
Starting with the naming of China as a strategic competitor in the US National Security Strategy at the end of 2017, to the full implementation of the presidential order to slap a special duty on Chinese imports worth around USD 50 billion under Section 301 by August 2018, and the subsequent threat to increase the tariff by another USD 200 billion in July, there is an emerging trend in the US of pinning China as the key strategic challenge and the imposition of measures to contain the ascendancy of China, particularly in the high technology arena.
In the new Defense Authorization Act signed by President Trump, the US government will largely prohibit US government agencies and any company wishing to contract with the government on telecommunication systems from using Chinese components or services. The ban will be implemented over a two-year period, and Chinese parts not involved in routing or viewing data are the only exemptions from the ban.
The trade dispute and the US denial of export privileges of semiconductors to ZTE in April and the subsequent paralysis of ZTE’s operations until the ban was lifted in July have given rise to concerns that China has technological shortcomings in some critical areas. The possibility of restricting access to the technology and products in these vital areas from the US is increasing.
Most of the observers agreed that the first product on the list of the weakest links in the Chinese industrial sector is the semiconductor. The country used more than 45 percent of global semiconductor production, and it only makes 15 percent of worldwide semiconductors locally. The country ran a trade deficit of USD 193 billion in 2017 on semiconductor trade.
China invested heavily in the semiconductor industry in the 2010s, but the results were less than impressive. In 2009, local semiconductor production accounted for less than 8 percent of the market, and by last year this had grown to 15 percent. The chance of China making enough semiconductors to account for 20 percent of market share by 2020 looks good with some foreign-invested factories and local memory factories coming online in the next two years. Despite the slow increase in market share, Chinese semiconductor manufacturers are still considered a third-tier player in the global semiconductor industry. The US is the undisputed leader, with Japan, South Korea, Taiwan, and Europe in the second-tier ladder. The Chinese remain two-three generations behind the leaders.
Out of the three stages of semiconductor manufacturing: Integrated Circuit (IC) design, IC fabrication, and IC assembly and testing, China has had some notable success in the IC design space, and it is next to the US in the global industry, though its size at less than 20 percent of the global market share is around 1/3 of that of the US.
The fabrication segment in the semiconductor production chain remains the weakest link of China. The Chinese manufacturers are still working on the 14nm design, while their overseas peers had passed that technological stage four years ago and have successfully migrated to 10 nm and now mass producing 7nm products. International fab leader Taiwan Semiconductor Manufacturing Corporation (TSMC) is working on a 5 nm design at the moment.
China is catching up in the less technology demanding assembly and test space, and the Chinese company Jiangsu Changjiang Electronics Technology (JCET) is one of the top five players in the industry.
The increasing restriction on learning from overseas peers means that domestic technological breakthroughs have become indispensable if China wants to break the technical barrier.
There are peculiar industrial factors as well as organizational mistakes that account for the unimpressive semiconductor industrial policy in China. The semiconductor industry is one of the most capital-intensive sectors, and capital expenditure spending often accounts for 25-40 percent of sales. Existing players who have a high capacity utilization rate enjoy the cash flow associated depreciation necessary for the next generation capital expenditure. This financial advantage of the incumbent has become more apparent in the last ten years as the cost of the latest generation fabrication rose from less than USD 1 billion in the 1990s to more than USD 10 billion recently. The market concentration among the top players has increased in the last decade, and one should note that aside from China, there is virtually no new player in the fabrication business around the world since the 2000s.
The semiconductor industry is also technology-intensive and fast moving. More than 40 academic disciplines are involved in the industry, and production steps can be as much as several thousand, while the product life is only around 2-3 years. Experienced incumbents who can ramp up production fast enjoy a considerable incumbency advantage. China has suffered from a technological embargo imposed by the Western countries on latest semiconductor fabrication equipment, and this access denial is probably the most important reason for its relative backwardness. The failure of the earlier plan to integrate the more than 40 academic disciplines to build a native technology ecosystem has not helped the matter too.
While there is no disagreement on the importance of the semiconductor industry, there is always the argument that importation is more cost-efficient and reinventing the wheel does not fit the modern economy. The social and economic incentives given to research personnel in semiconductor fabrication research is very far behind that given to product development engineers in other fields. There is a severe shortage of semiconductor fabrication related personnel at the moment. Some experts have estimated that if China is going to hit a semiconductor production output value of RMB 1 trillion by 2020 as government plans have indicated, the country will need around seven hundred thousand related personnel, but the available workforce today is only three hundred thousand.
The semiconductor industry is a general purpose industry that anchors many other sectors, such as computing, digital communications, high-performance machinery, and others. Aside from meeting the technical requirements of the application sector, the semiconductor industry must produce its products at a competitive price. How to design an industrial development plan to fit such a dynamic industrial environment while satisfying both cost and availability criteria is a challenge. Traditional state-centric high-tech industrial development policies often emphasize availability and neglect the cost efficiency aspect. In an industry that talks about obsolescence in a matter of months, the old way of doing things must be changed.
The semiconductor success story of the US is often associated with Thomas Moore through the famous Moore’s Law and the company Intel that he set up. And the success of Taiwan semiconductor industry is also associated with Morris Chang and the Taiwan Semiconductor Manufacturing Corporation (TSMC) that he established. In the fast-moving high-technology area that calls for both technical excellence and enlightened leadership to guide the technology development path, the presence of inspirational leadership is often as significant as the availability of risk-taking capital and a vast army of technologically proficient researchers and engineers.
The increasing restriction on learning from overseas peers means that domestic technological breakthroughs have become indispensable if China wants to break the technical barrier. The country needs new ideas, needs to set up a new research and development structure, and needs to seek out inspirational leaders to accelerate the development of the semiconductor industry. China is facing a “Sputnik moment” in semiconductor technology and the need to control core technology. Industry observers are watching how the country will respond to the challenge, and the outcome will likely affect the country’s drive on industrial transformation.