The Great Economic War (GEW) (Part 10: The semiconductor technology battlefront)
Phuah Eng Chye (3 December 2022)
The US reacts strongly when it feels its economic and technology leadership is threatened; as it did against Japan in the 1980s. China’s rapid economic growth and technology advance is now triggering a similar response. According to the South China Morning Post (SCMP), “the US-China trade war started on July 6, 2018, when US President Donald Trump imposed a 25 per cent tariff on US$34 billion of Chinese imports…Further tariffs were imposed during 2018 and 2019”. “Initially, decoupling was dismissed as unrealistic because China was too big for US companies to ignore, and its economy was too intertwined with the global economy to be walled off”.
While there was little progress on the trade front where both participants seemed unwilling to make concessions, the spotlight soon shifted to supply chains and technology[1]. “The trade war was soon overshadowed by a tech war driven by US concerns that China was using unfair means[2], including state power and IP theft, to achieve its goal of becoming a global leader in core technologies like AI, semiconductors and 5G”. “A limited form of decoupling gained bipartisan support…a certain degree of bifurcation would be in US interests. That already exists to a large degree in the internet space, with Facebook and Google not allowed to operate their social network and search engine in China, and attempts under the Trump Administration to ban or restrict the US presence of Chinese social media like TikTok and WeChat. The global pandemic, which highlighted the world’s reliance on China for key medical supplies like masks, accelerated calls to bring supply chains back to the US. In February, President Joe Biden signed an executive order calling for a review of critical product supply chains for chips, batteries, pharmaceutical ingredients, and rare-earth minerals”.
SCMP notes “analysts generally point to Made in China 2025, Beijing’s 10-year blueprint for transforming the country from a manufacturing giant into a world manufacturing power, as the trigger for the tech war…when…announced in 2015. A series of unrelated events, including fines and sanctions imposed on Chinese telecoms giant ZTE for covering up its role in selling US technology to Iran, a report from the US Trade Representative (USTR) on China’s alleged theft of US intellectual property and force technology transfers, and concerns over Huawei Technologies Co’s dominance in 5G technology, morphed into a broader tech war that gained bipartisan support”.
SCMP notes “Washington’s concerns over Huawei date back to the early 2000s, and were largely driven by the belief that the Chinese company had close ties to the Beijing government. Years later, Huawei’s emergence as a global leader in 5G technology set off alarm bells in Washington…in January 2018…US lawmakers pressured American telecoms giant AT&T to pull out of a deal to distribute Huawei smartphones to US consumers. Seven months after the AT&T decision, Washington barred government agencies from purchasing equipment and services from the Chinese company. When Huawei was added to Washington’s Entity List in May 2019, banning it from buying products and services from US companies without US government approval, the US Department of Commerce said Huawei and its affiliates were deemed to be involved in activities…contrary to the national security or foreign policy interests of the United States. Washington also alleges that Huawei contravened US sanctions against Iran, leading to the arrest in Vancouver of Huawei CFO Meng Wanzhou in December 2018 on bank fraud charges”. She has since been released.
Kevin Wolf[3] notes “in 2015, BIS was worried about the behavior of a giant telecommunications company called ZTE…investigate its use of front companies to buy US origin items, to send to Iran and North Korea, in violation controls and sanctions. of US export. In the spring of 2016, we added ZTE to something called the Entity List, which prohibits the export from the US of anything and from outside the US of sensitive foreign made items…However, in an unusual move, in May of 2018, President Trump tweeted that President Xi of China and I are working to give massive Chinese phone companies ZTE a way back into business fast. Too many jobs lost in China. Commerce Department has been instructed to get it done.” Chad Bown comments “President Trump’s tweet and policy overruling his own national security enforcement officials at BIS put US export controls into the spotlight. Export controls had gone from technical work on national security and enforcement of US laws to a potential pawn or bargaining chip in the president’s suddenly escalating trade war with China. Something was changing”.
The other major enforcement was on Huawei. Kevin Wolf notes “there was a broader, more foundational issue involving Huawei with respect to its worldwide dominance in 5G telecommunications applications, which creates national security concerns for the US and its allies. Their dominance in the 5G ecosystem, which is of significant threat for cyber surveillance and cyber intrusion and theft and hacking. And if there was a company that could follow instructions from the Chinese government that could pull information at 5G speeds from anywhere on the planet”. Chad Bown adds “the US government and national security officials had been suspicious of Huawei dating back to at least the 2000s. But the problem was only three companies globally were supplying most of this 5G equipment. Nokia and Ericsson, two European companies were Huawei’s only real competitors. With no American companies to protect or promote, the US government was concerned that the critical telecommunications infrastructure of the future was going to end up entirely in the hands of this Chinese company”.
Kevin Wolf notes in May of 2019 , the Trump administration’s BIS added Huawei to the Entity List, just like with ZTE, prohibited all exports from the United States of any kind any US origin items and sensitive foreign made items made with US technology to Huawei and most of its affiliates”. “Over the course of 2019 and into early 2020 the administration realized that their objective of shutting down the ability of Huawei to make 5G items for distribution worldwide was ineffective. US companies and foreign companies, so long as they made their commercial products outside the United States, they weren’t subject to any US export controls. They could be sent from outside the US to Huawei and it was creating an unlevel playing field…The problem was that only a tiny share of semiconductor production globally took place on US soil”. The new US policy wasn’t stopping the massive shipments to Huawei of semiconductors made outside of the US…and the only companies suffering from the US export controls were the ones manufacturing chips in the US.
To address this, in August 2020 the Trump administration expanded its interpretation of the scope of foreign direct product rule to apply “to any foreign made item made with any kind of semiconductor, electronic, computer, or telecommunications technology of any sensitivity, that was US origin or – here’s the kicker – if the foreign-made item such as the semiconductor was made in a facility that used US equipment”. Kevin Wolf explains “since every foundry on the planet uses US tools and equipment then, by definition, every foreign-made semiconductor on the planet – if they were destined to Huawei – became subject to US export controls and licensing requirements in August of 2020”. This subjected all plants worldwide to US export controls on sanctioned Chinese companies.
Foreign semiconductor manufacturers like TSMC and Samsung were given a choice:
“They could stop selling chips to Huawei and continue to use their US-made equipment, or they could keep selling to Huawei, but they would be cut off from access to these top-of-the-line American tools…Companies all over the planet immediately started complying and applying for licenses”. The Trump administration converted the Entity List into a licensing policy that denied Huawei access to inputs for 5G or in support of 5G applications but permitted access to older 4G inputs. Kevin Wolf notes “lots of other Chinese companies suddenly became worried about the Trump administration’s new use of export controls…To protect themselves, these other Chinese buyers started to hoard semiconductors”. “The December 2020 action targeted SMIC – the Semiconductor Manufacturing International Corporation…Like the second action taken against Huawei, these new controls would cut off SMIC from buying US tools to make chips”.
The US steps up offensive
The US government has been broadening its actions. The US pressured Netherlands to block ASML from selling its DUV (deep ultraviolet) and EUV (extreme ultraviolet) lithography tools and on Japan to ban Nikon’s sales of similar chip-making technology to China. It pressed Samsung and SK Hynix not to expand their China chip-making operations. This was followed by new license requirements for Nvidia and AMD exports of its highest-performance products for servers to “address the risk that the covered products may be used in, or diverted to, a military end use or military end user in China and Russia.”.
In August 2022, export restrictions were issued on electronic computer-aided design (ECAD) software affecting certain Electronic Design Automation (EDA) tools for gate-all-around (GAA). Scott Foster explains the restrictions cover the 2-nm technology of TSMC and Intel as well as the 3-nm and 2-nm technologies of Samsung Electronics” and “products not only from EDA industry leaders Synopsys, Cadence Design and Siemens (Mentor Graphics), but also from Ansys and other suppliers as well”. “The same tools are used for 5-nm to 28-nm design rules, meaning that if all EDA tools used for GAA are excluded, then all EDA tools will be effectively blocked from being licensed in China. Not only that but the shipment to China of integrated circuits designed using these EDA tools might also be banned. That would include products designed by Qualcomm, NVIDIA, MediaTek, Intel, Samsung and other companies and made by Intel, Samsung, TSMC and other foundries. This broad definition not only would deprive Chinese makers of cell phones and other electronics of the advanced components they need, but would also lead to a serious loss of sales for foreign semiconductor design and manufacturing companies”.
Matthew Reynolds points out “on September 16, U.S. national security advisor Jake Sullivan…announced a major shift in the strategy underpinning U.S. export controls against China. Sullivan stated that the United States would no longer focus primarily on using export controls to maintain relative advantages over competitors in certain key technologies.” Instead, he declared that the United States must maintain as large of a lead as possible. The previous paradigm of export controls had allowed adversaries to move up the critical technology frontier at a deliberate lag of a few generations behind the United States. The U.S. government now appears intent on blocking China’s ability to advance in certain semiconductors that have specific critical applications. This strategic shift is taking place against the backdrop of China devoting immense resources to advance its semiconductor sector, with some success, and implementing its military-civil fusion doctrine, making it increasingly difficult to distinguish civilian end-users from military end-users…the Biden administration’s controls explicitly establish a point beyond which they seek to stop China’s advancement, and importantly, in a departure from previous measures, actively seek to degrade China’s current chipmaking capabilities, not just forestall their advance”.
The following Bureau of Industry and Security (BIS) measures[4] probably stunned the semiconductor industry and China with their far-reaching implications are still being digested. Gregory C. Allen notes “on October 7…The Biden administration announced a massive policy shift on semiconductor exports to China as well as revised rules for how the lists of restricted parties are managed…With the new policy, which comes on the heels of the CHIPS Act’s passage, the United States is firmly focused on retaining control over chokepoint technologies in the global semiconductor technology supply chain. The most important chokepoints…are AI chip designs, electronic design automation software, semiconductor manufacturing equipment, and equipment components…these actions demonstrate an unprecedented degree of U.S. government intervention to not only preserve chokepoint control but also begin a new U.S. policy of actively strangling large segments of the Chinese technology industry – strangling with an intent to kill…the Biden administration is trying to (1) strangle the Chinese AI industry by choking off access to high-end AI chips; (2) block China from designing AI chips domestically by choking off China’s access to U.S.-made chip design software; (3) block China from manufacturing advanced chips by choking off access to U.S.-built semiconductor manufacturing equipment; and (4) block China from domestically producing semiconductor manufacturing equipment by choking off access to U.S.-built components”.
Gregory C. Allen notes earlier efforts “to allow commercial Chinese access to high-end chips while blocking military access” was “ineffective at stopping indirect sales to the shell companies that helped the Chinese military evade export controls”. “For high-end AI and supercomputing chips…can no longer be sold to any entity operating in China, whether that is the Chinese military, a Chinese tech company, or even a U.S. company operating a data center in China. The Biden administration is essentially saying to China: If your policy is military-civil fusion, then the only realistic way of implementing our policy of no military end use is to end all sales to China, and we are now willing to take that step”.
In this context, Gregory C. Allen explains “in reality, the rules set a performance threshold for what sorts of chips can be sold to China, and everything above the performance threshold requires the seller to seek an export license from the Department of Commerce. However, policymakers have told all sellers that license applications for sales of such chips to China will face a presumption of denial. Thus, this new policy is de jure a new license requirement but de facto a ban. Because the United States is broadly invoking the foreign-direct product rule, this export license requirement applies not only to Nvidia and AMD but to any would-be competitors around the world”. In the case of Nvidia chips, it is not only the loss of access to the chips but also the loss of the benefit of Nvidia’s CUDA software ecosystem which makes it easy for programmers to write parallelized software and to ensure compatibility with older chips. “Any customer who seeks to stop using Nvidia chips has to leave the CUDA ecosystem, which requires solving a lot of incredibly hard software problems for which CUDA already provides free answers…explaining why Nvidia accounts for 95 percent of AI chip sales in China”. While there are risks that Chinese or non-US firms could “band together and build an alternative ecosystem to CUDA, one that potentially could harm Nvidia sales not only in China but around the world”.
“However, Biden administration officials foresaw this risk and blocked “China from designing AI chips domestically by choking off its access to U.S.-made chip design software and U.S.-built semiconductor manufacturing equipment”. “By invoking the foreign-direct product rule, the United States is prohibiting any semiconductor manufacturing company worldwide from providing services to any Chinese chip design company that is seeking to make high-end chips for AI or supercomputing. The Foreign-Direct Product Rule offers two hooks to block Chinese chip design companies from having their chips manufactured: on the EDA software and again on the U.S.-built equipment that is an irreplaceable (at least for now) part of every semiconductor manufacturing operation anywhere in the world. Any chip manufacturing operation – whether Chinese or otherwise – that seeks to build Chinese chip designs will risk losing its own access to U.S. semiconductor manufacturing equipment. With the new rules, no Chinese chip design company will be allowed to outsource manufacturing abroad for advanced AI and supercomputing chips. For those Chinese organizations on the BIS Entity List, they will be blocked from outsourcing the manufacturing of any types of chips at all…these recent actions are designed to put the roughly 28 Chinese chip design and supercomputing firms on the parties of concern lists out of business”.
Gregory C. Allen points out the hypothetical path to escape the US semiconductor chokepoints “would be to design the chips in China, pirate older EDA software, and manufacture the chips in China using older manufacturing technology. This all-Chinese semiconductor supply chain strategy is in fact a logical extension of the one Huawei announced that it intends to pursue this year with a resurrected HilSilicon business”. “However, the Biden administration has taken steps to stomp out that approach as well…With these latest rules, the United States has dramatically restricted the supply of advanced semiconductor manufacturing equipment to all of China. Additionally, the rules also restrict the supply of a much more broadly applicable set of semiconductor manufacturing equipment (e.g., inspection and metrology equipment) to those companies in China that are currently making advanced chips (such as SMIC and YMTC). In doing this, the Biden administration is going beyond the traditional U.S. policy of seeking to keep the United States’ two semiconductor technology nodes ahead of China and is now trying to actively degrade China’s technological maturity below its current level…The United States is blocking the export of advanced semiconductor manufacturing equipment to all of China, but in the case of SMIC’s 14 nm production facility, the United States is even blocking the export of older, less advanced equipment. In other words, the United States is attempting to put SMIC’s 14 nm production line out of business”. The equipment sale restrictions apply to Gate-All-Around Field Effect Transistors (GAAFET), DRAM at the 18 nm process node and below and for NAND at 128 layers or higher. “It will be especially difficult for Chinese companies to evade these equipment export controls. Without access to the support teams and spare parts of U.S. equipment suppliers, these facilities may be forced to slow or halt production. Even if the existing advanced production facilities at SMIC and YMTC somehow manage to continue producing, they are unlikely to be joined by other companies and facilities in China anytime soon. Overall, these new restrictions represent a devastating blow for the makers of advanced chips in China”.
Alexandra Alper and Karen Freifeld notes export restrictions on tools to make NAND chips with more than 128 layers is notable because these memory chips do not have specialized military applications. “Micron and Western Digital are under pressure from Yangtze Memory Technologies Co Ltd’s (YMTC) low prices, as the White House wrote in a June 2021 report. YMTC’s expansion and low-price offerings present a direct threat to Micron and Western Digital…report described YMTC as China’s national champion and the recipient of some US$24 billion in Chinese subsidies”. “YMTC accounts for about 5% of worldwide NAND flash memory chip production, almost double from a year ago. Western Digital stands at about 13% and Micron 11%…Production of NAND chips in China has grown to more than 23% of the worldwide total this year from under 14% in 2019, while production in the US has decreased from 2.3% to 1.6% over the same period”. While protecting “the only US memory chip producers, Western Digital Corp and Micron Technology Inc, which together represent about a quarter of the NAND chips market”, this policy would “hurt South Korean memory chip juggernauts Samsung Electronics Co Ltd and SK Hynix Inc with production facilities in China” and “LAM Research Corp and Applied Materials, both based in Silicon Valley, are the primary suppliers of such tools”.
Gregory C. Allen notes the US is also blocking “China from developing its own semiconductor manufacturing equipment by choking off access to U.S.-built components…Of all the semiconductor supply chain chokepoints where the United States has significant leverage over China, this is the one with by far the most durable competitive advantage”. His view is that “in weaponizing its dominant chokepoint positions in the global semiconductor value chain, the United States is exercising technological and geopolitical power on an incredible scale. The targeted nature of the Biden administration’s actions suggests three important implications about its worldview. First, the United States believes that China is willing to take extraordinary measures – including but not limited to spending hundreds of billions of dollars, hacking U.S companies, and creating networks of shell companies – in order to evade export controls and free itself from dependence on U.S. semiconductor supply chains…the past decade of Chinese technological progress shows that anything less simply will not work…When it comes to China’s AI and semiconductor industries, the United States is done with half measures”. Second, the US “is wielding its power forcefully, but at a limited set of Chinese targets, in order to preserve U.S. chip power and leverage over the long term. The Biden administration is signaling that it believes that – despite all the United States’ considerable technological advantages in semiconductors – its lead is not so overwhelming that China would never have caught up”. “Third, this policy signals that the Biden administration believes the hype about the transformative potential of AI and its national security implications is real…the Biden administration intends to hold those benchmarks constant, meaning that the gap in performance will grow over time as the world advances and China remains stuck behind. It is as though the United States is saying to China, AI technology is the future. We and our allies are going there. You can’t come”.
Reva Goujon, Lauren Dudley, Jan-Peter Kleinhans and Agatha Kratz estimates the “immediate costs to US semiconductor manufacturing equipment (SME) firms arising from a narrow implementation of the controls would be fairly limited (between $1.4-$3 billion in annualized sales, according to our estimates). However, if these controls were to be applied more broadly, for example to target foreign fabs in China, costs could rise quickly (to $4.6-$5.2 billion in annualized sales, based on our calculations)”.
Another major shock was the extension of export controls to “block U.S. persons — both citizens and permanent residents — from supporting the development or production of certain high-end chips that could advance Chinese tech applications from military use and artificial intelligence to supercomputers”. Cheng Ting-Fang points out “previously, Washington only regulated U.S. persons’ activities if they were related to nuclear or chemical and biological weapons programs in China”. “The latest export controls are Washington’s first explicit attempt to stop the sea turtles from helping China develop advanced semiconductors for supercomputers, AI and other key tech applications…the export control document said. We don’t have actual numbers of how many people could fall under the U.S. new regulations this time. But from public information, many key Chinese chip companies are led by Chinese Americans or have several Chinese American senior executives”. Hence, “U.S. persons related to the development of certain grades of high-end chips must cease until they receive a license”. Christopher Timura clarified the “new restriction applies not only to engineers who might be supporting the development and production of the types of targeted ICs, but also to other business personnel because it also extends to U.S. persons that are involved in other business functions such as shipping or transmitting commodities or technologies…the restriction applies equally to U.S. expatriates supporting projects and chipmaking facilities located in China and to American companies located in the U.S…Violators of the licensing requirement can be subject to both civil and criminal penalties”. Cheng Ting-Fang adds “the new regulation caught many of these Chinese American executives off guard…Many of them have family overseas in the U.S. and have assets and properties in the U.S…It will be very tough for them if the U.S. can really enforce and execute this new regulation”. In addition, “it would not be easy for these Chinese Americans to abandon their citizenship overnight” either.
Jeff Pao reports “most mainland Chinese IT writers admitted that the US curbs are having a big negative impact on China’s chip sector but some of them said the situation remains manageable”. Some Chinese American engineers resigned but continued to work informally. Given the restrictions applied only to US citizens, semiconductor firms offered huge salary packages to lure talent from Taiwan, South Korea, Japan and Singapore”. While some Chinese American engineers and executives likely leave China as they want to retain their US citizenship but some expect “many of the departed Chinese American engineers will eventually give up their American citizenship and return to China” as it will be difficult for them to enjoy equivalent job opportunities in the US. In addition, “mainland Chinese commentators said the US curbs will have only a short-term impact on China, as the country has a strong talent pipeline from its technology schools”.
On top of the export restrictions, Jeff Pao notes “the BIS added Yangtze Memory Technologies Co (YMTC), China’s top memory chipmaker, and 30 other Chinese entities to an unverified list. The bureau said it was unable to verify the bona fides of the Chinese firms because an end-use check could not be completed satisfactorily for reasons outside the control of US authorities”. These companies are now “required to pass an end-use check within 60 days or they will be added to the entity list of the US Commerce Department’s Bureau of Industry and Security (BIS)”. “Some Chinese semiconductor makers are reported to have taken the initiative to contact the United States embassy in Beijing” to remove themselves from the unverified list.
The sweeping approach is intended to address weaknesses of past controls. Jonathan Goldberg notes “the United States has gone from blocking specific companies in China, to blocking all companies and focusing on specific products”. “Until just recently, the US government had largely relied on its Entity List…The entity list process is unlikely to be remembered for its success. As soon as a large number of large Chinese companies started appearing on this list around 2017-2018, US companies dispatched teams of lawyers and lobbyists to poke as many loopholes as they could. From the Chinese perspective, the entity list appeared terrifyingly random. True, Huawei was a primary target, but many other companies ended up on the list for reasons not entirely clear to them. We have spoken with many people at Chinese companies who have grown incredibly wary of doing business with anyone in the US for fear that they will somehow end up on the list…We are now six years into this process, how are these units just now being added to the list? Judging by what came next, we have to think that the people putting the list together realized they faced a hopeless task. Chinese companies have become incredibly adept at creating byzantine corporate structures, making tracking their affiliations close to impossible. Our guess is that someone in the government realized the futility of this approach and made the decision to instead switch to banning shipments of specific products”.
Gregory C. Allen notes “the Department of Commerce needs the intent and resources to manage the Entity List much more rapidly…Chinese organizations can change shell companies to evade export controls in a matter of days or weeks…Fortunately, another set of rule changes…changed the handling of the Unverified List (a sort of precursor to the Entity List for cases where the government has not yet been able to verify an organization’s bona fides) and suspended the use of export license exemptions for companies on the Unverified List. With this rule change, the Department of Commerce can now effectively block exports to any Chinese organization in 60 days. This provides a much faster and more flexible response to Chinese shell companies and other export evasion tactics. These export restrictions apply to U.S.-owned entities operating in China, but they do not apply to Chinese-owned entities operating in, for example, India. In fact, there are no U.S. export controls that apply specifically to a company’s country of ownership. This greatly increases the challenges associated with preventing the international subsidiaries of Chinese corporations from smuggling chips into China in violation of U.S. export controls. Imagine if several Chinese cloud computing giants purchased 10 percent more AI chips than they need at dozens of their hyperscale datacenters outside of China and then worked to smuggle those excess chips back into China. While this policy dramatically raises the obstacles to China’s production of an indigenous supply chain, China is not going to give up. The local government of Shenzen, a key hub of China’s tech sector, has already established a series of subsidies for organizations that purchase open source or Chinese alternatives to U.S. EDA software and chip intellectual property”. However, US funding of “research and development into open-source technologies that may compete with the dominant proprietary solutions offered by U.S. semiconductor companies, thereby offering China a non-export-controlled path to desirable capabilities”.
Political pressure is also being applied to discourage US companies from sourcing from Chinese manufacturers. Roslyn Layton and Jeff Ferry warned Apple’s intention to ship the iPhone 14 with Yangtze Memory Technologies Company (YMTC) memory chips could “compromise iPhone users’ security and privacy with untrusted Chinese technology, concentrate more chip production in China, increasing supply chain risk, force the exit of at least one memory chip supplier from the U.S. or another democratic country, thereby potentially reducing U.S. jobs, give legitimacy to state-subsidized military-linked YMTC as a chipmaker, bolster the Chinese government’s quest to dominate global semiconductor production, reducing U.S. leadership and innovation. Ideally Apple will voluntarily end its partnership with YMTC”.
Anton Shilov notes while “Apple argues that it plans to use chips from Yangtze Memory only for iPhones to be sold in China”, U.S. legislators remain upset and argue “there is credible evidence that YMTC is breaking export control laws by selling goods to Huawei. Apple will effectively be transferring knowledge and know-how to YMTC that will supercharge its capabilities and help the CCP achieve its national goals.” They also “accuse the Chinese government of illegally subsidizing YMTC, giving it a competitive edge over American makers of memory, such as Micron” and want YMTC to be place “on the entity list for its ties with CCP, and therefore ban it from getting U.S. tools and software”. Anton Shilov points out though “even without YMTC, Apple’s products are full of technology developed and made in China. That might change eventually as Apple shifts some of its production to India and other countries, but that will of course be a slow process”. Apple has since changed its decision to source chips from YMTC.
Scott Foster notes on August 2022, the US Department of Commerce issued a ruling that identified gallium oxide semiconductor substrates as an emerging and foundational technology essential to the national security of the United States. “While not superior in all respects, gallium oxide has the virtue of being relatively easy and potentially cheaper to manufacture. It should make inroads into the power device market over the next several years and could eventually make a major contribution to electric vehicles once economies of scale are achieved”. “Tellingly, one major investor in the leading US company in the field is the US Department of Defense”. “The development of gallium oxide wafers and devices is underway in the US, Japan, Europe, South Korea, Taiwan and China. While the US government is ringing alarms about gallium oxide’s national security implications, the Japanese are leading way to its commercialization. Three companies stand out as developers and manufacturers of gallium oxide substrates, wafers and devices, namely Kyma Technologies of the US and FLOSFIA and Novel Crystal Technology of Japan”.
Reva Goujon, Lauren Dudley, Jan-Peter Kleinhans and Agatha Kratz note “more controls are in the pipeline. The White House is advancing plans to create an outbound investment screening regime to prevent US capital from contributing to the development of force-multiplying technologies in China, including advanced semiconductors, high-performance computing, and possibly bio-manufacturing and high-capacity batteries”. More Chinese entities could be added to the unverified list (UVL) to demonstrate non-military end-use within 60 days or the foreign direct product rule (FDPR) could be modified to enables the US to target a broad set of tech categories that could be linked to weapons of mass destruction (WMD) end-use. Foreign partners could be compelled to “follow the US lead by expanding the use of FDPR, invoking far-reaching restrictions on the activities of US citizens linked to China’s development of advanced semiconductors, and implicitly threatening to expand extraterritorial measures if US allies fail to implement their own controls”. Jake Sullivan[5] noted the importance of export controls to maintain as large of a lead as possible and highlighted computing-related technologies, including microelectronics, quantum information systems and artificial intelligence as critical areas.
Another provocative move would be to sanction tech companies for human rights violations. Jon Bateman highlights “Washington is weighing whether to sanction Chinese tech company Hikvision for human rights violations under the Global Magnitsky Act…Hikvision is a large, publicly traded Chinese video surveillance equipment maker…with international sales to government agencies and private companies in more than 180 countries…The firm is already on several lists that prevent Hikvision from importing U.S.-origin goods without a license; receiving American investment; or selling products to U.S. telecoms, federal agencies, or federal contractors. Moreover, Hikvision will be unable to sell any new products in the United States once the Federal Communications Commission finalizes pending regulations required by Congress. Yet the current restrictions on Hikvision pale in comparison to what Washington is now considering. The Global Magnitsky Act empowers the Treasury Department to impose a range of powerful sanctions on foreign companies and individuals involved in human rights violations, to include placement on the Specially Designated Nationals (SDN) List…the harshest financial penalty in Washington’s tool kit, often used against terrorists, drug lords, and the worst human rights abusers…It would grievously (perhaps fatally) wound the company, depending on how sanctions are implemented – specifically, what scope of transactions are blocked and whether any exemptions are offered…The most severe sanctions would freeze all of Hikvision’s assets and create civil and potential criminal penalties for anyone globally who sends Hikvision money or property, provided there is some nexus to U.S. legal jurisdiction…U.S. sanctions could stop Hikvision from selling anything to or buying anything from all countries friendly with (or at least afraid of) the United States. The sanctions could block Hikvision from using reputable international banks. Hikvision’s online presence could be eliminated in much of the world if internet hosts and social media platforms take down its accounts. Many of Hikvision’s employees could technically face U.S. civil and criminal liability for their involvement with a designated entity. And Hikvision would suffer incalculable reputational damage. It would also find that many parties choose to over-comply with sanctions – doing more than U.S. law requires to avoid any scintilla of risk”.
Jon Bateman points out “Hikvision’s provocative placement on the SDN List…Beijing might respond in any number of ways – perhaps by freezing the Chinese assets of a large U.S. company, halting bilateral talks on stock market rules, or further closing ranks with Russia…Washington should also be aware that a powerful Chinese counterpunch could trigger tit-for-tat dynamics and escalatory spirals. If Beijing reacts by placing U.S. firms on its Unreliable Entity List, for example, then American political figures would doubtless demand further punishment of China. At the same time, the U.S. precedent set by a Hikvision designation would create domestic political pressure for future SDN listings. Dahua, iFLYTEK, SenseTime, Megvii, and DJI – leading Chinese sellers of digital cameras, artificial intelligence software, and drones – have all been cited for supporting Xinjiang repression, and each already faces some of the same restrictive measures currently placed on Hikvision. If Hikvision goes on the SDN List, members of Congress and U.S. commentators would quickly call for designating these other firms, too. Uncorking the SDN List for Xinjiang abuses would also embolden those who have already called for even broader designations of Chinese firms involved in ordinary mass surveillance and cyber espionage. Any further SDN listings would effectively declare economic war on growing portions of the Chinese tech sector – with serious and unpredictable consequences for bilateral relations, global stability, and the U.S. economy”.
Jon Bateman points out “if U.S. leaders can’t convincingly justify whatever line they may wish to draw around Hikvision, then Beijing, other governments, and the global private sector could come to expect a much more serious and abrupt technological decoupling than U.S. leaders really intend. These actors could choose to self-decouple now on their own terms – preemptively avoiding and winding down investments, relationships, and other long-term commitments that depend on enduring U.S.-China tech ties. Yet much of these commercial dealings and scientific collaborations benefit the U.S. economy and national security. The United States risks setting in motion a broader, faster, and messier decoupling than it wants or can afford. To avoid spooking others and ceding control to them, Washington’s restrictive actions must be narrowly targeted and have definite limiting principles. Placing Hikvision on the SDN List is the opposite of this prudent approach”. “Yet Washington must recognize the limits of its power and the risks of overreach. It is hard to imagine that placing Hikvision on the SDN List could substantially alter China’s Xinjiang policies. U.S. sanctions often fail to change their target’s behavior, and Beijing has adhered to its treatment of Hong Kong, Tibet, and Xinjiang despite overwhelming international outrage. Any hope of changing China’s calculus would require many countries – representing the bulk of global trade and diplomatic heft – to coordinate in carefully planned ways. Such a full-throated international campaign would call for much greater leverage and delicate direct negotiations with China. Ad hoc, unilateral U.S. sanctions on Hikvision would accomplish little while risking great harm to American interests”.
CHIPS and Science Act
The offensive on technology chokeholds was preceded by the CHIPS and Science Act[6] which was signed into effect on 9 August 2022. Its aim is to revitalize “American semiconductor manufacturing and strengthen the global semiconductor supply chain”. Gary Clyde Hufbauer and Megan Hogan explains the “US semiconductor subsidy bill, dubbed the Creating Helpful Incentives to Produce Semiconductors Act (or CHIPS Act), with a headline cost of $76 billion, including $52 billion in grants to support advanced chip manufacturing and R&D in the United States and $24 billion in investment tax credits for chipmakers by letting them write off 25 percent of capital expenses incurred in manufacturing semiconductors and related equipment. The $76 billion is divided between increasing fabrication capacity ($65 billion) and supporting R&D ($11 billion). The CHIPS Act included guardrail statutes to discourage firms from expanding their plants in China by prohibiting recipients of US federal funding from expanding or upgrading their advanced chip capacity in China for 10 years.
Gary Clyde Hufbauer and Megan Hogan estimates the CHIPS Act will “provide US-headquartered firms less than 8 percent of their prospective intellectual and physical capital outlays…between 2022 and 2030…of $1 trillion and are marginally lower than the “subsidies reported for Taiwan and South Korea, and substantially smaller than subsidies reported for China”. They note “CHIPS Act subsidies, concentrated on US soil, are lopsided in favor of physical plants (85 percent of total subsidies) rather than R&D (15 percent). The political economy of this division is obvious (jobs, jobs, jobs!), but the history of US industrial policy shows that the greater national strength is advanced research, not physical plants. A larger share of the subsidy pie should have been delivered to R&D. But an increase in US chip production is almost guaranteed”.
Gary Clyde Hufbauer and Megan Hogan note “between January 2021 and 2022, nearly $80 billion in new semiconductor investments have been announced in the United States through 2025, including a $12 billion Taiwan Semiconductor Manufacturing Corporation (TSMC) plant in Phoenix, Arizona; a $20 billion Intel plant outside Columbus, Ohio; a $17 billion Samsung plant in Taylor, Texas (near Austin); and a $30 billion Texas Instruments plant in Sherman, Texas (near Dallas). Construction delays, unreliable access to vast amounts of water and electricity, and shortages of key equipment (especially chip machines) and skilled engineers will continue to constrain the expansion of chip manufacturing capacity in the United States and abroad…New fabs announced in anticipation of the CHIPS Act will not come online for at least two years – meaning they will not arrive in time to alleviate current shortages…Consequently, the main impact of subsidies will be to shift the location for constructing new fabs to US soil, rather than to enlarge chip output in the near term”.
The dilemma for allies
Reva Goujon, Lauren Dudley, Jan-Peter Kleinhans and Agatha Kratz argue “the US is establishing itself as the de facto regulator of the global semiconductor industry, and is willing to apply far-reaching provisions, when it is deemed necessary, to compel foreign partners to fall in line. By including several extraterritorial measures in this package of controls, the administration is signaling to partners that they could be facing more pressure after the 2024 presidential election if they don’t align controls independently or in plurilateral working groups like the US-EU Trade and Technology Council (TTC) and the fledgling Chip 4 Alliance (involving the United States, Japan, Taiwan, and South Korea). Looking forward, it will be important to watch where gaps are created, as partners – who take a traditional country-agnostic, product-specific approach to export controls – shy away from matching the new US restrictions. It is likely to be easier for the US and Japan to align their approaches to semiconductor controls than it will be for the US and EU…Under a more assertive GOP-controlled Congress, or should the White House conclude that US allies and partners are moving too slowly in aligning their own export controls, Washington could choose to ratchet up entity listings targeting Chinese semiconductor firms using FDPR designations”.
The TTC currently focuses on policies and processes related to investment screening and to sensitive technologies such as export control on dual-use and emerging technologies, technology acquisition, compliance and enforcement approaches. Arjun Gargeyas notes “the Chip 4 alliance covers all the major areas of the value chain. Although there are other dependencies and bottlenecks, the four states can run semiconductor production more efficiently together”. The US has strengths in design; Taiwan is the global hub for semiconductor manufacturing – including all Assembly, Testing, Marking, and Packaging (ATMP) processes; South Korea’s Samsung has leading design and manufacturing capability; and Japan is dominant in the production of critical manufacturing equipment and materials such as photoresists”.
Scott Foster asks “the critical question: Is US policy aimed at controlling China, South Korea, or both? The answer is that the US government would like all three of its proposed Chip 4 Alliance partners – South Korea, Japan and Taiwan – to turn their attention away from China and back to the US, no matter what losses they might incur. To some degree, through subsidies provided under the CHIPS Act, it is willing to bribe them to do so. R&D is a different issue. Protecting intellectual property is a common interest and defense against China’s light-fingered approach to trade secrets and practice of hiring away experienced engineers could benefit from consultation and coordination. But the Koreans, Japanese and Taiwanese might also have issues with technology-sharing and joint R&D projects directed from Washington, DC…Korean, Japanese and American companies compete in memory chips and Samsung competes with the Taiwanese and Intel in the foundry (contract manufacturing) market. So why would they want to share their R&D with market competitors? Fears have been expressed, notably in Korea, that Chip 4 might funnel East Asian technology to Intel and Micron. On the other hand, the four also collaborate and depend on each other. Intel is both a competitor and a customer of Taiwan Semiconductor Manufacturing Company (TSMC), which recently became a joint venture partner of Sony. Without Japanese equipment and materials (e.g., photoresists), none of them would be making much of anything. That leaves the inconvenient fact that China is the world’s largest market for semiconductors and, at least recently, the largest market for semiconductor production equipment. All players thus have a lot to lose from cutting off the Chinese, including American fabless IC design companies (e.g., Nvidia), integrated device manufacturing companies (e.g., Intel), and makers of semiconductor production equipment (e.g., Applied Materials).
Scott Foster adds “caught between a rock and hard place, the Koreans appear to be reassuring the Chinese before negotiating a non-threatening membership in America’s Chip 4 initiative. On August 2022, “China and South Korea have agreed to establish a new Collaborative Supply Chain Council to address any disruptions of their extensive and interdependent economic relationship in a timely fashion”. He points out “in 2021, almost 40% of South Korea’s semiconductor exports went to China, according to the Korea Chamber of Commerce and Industry, up from only 3% in 2000. Moreover, Samsung makes about 40% of its NAND flash memory in the Chinese city of Xian at the world’s largest NAND flash memory factory. SK Hynix makes DRAM in Wuxi and has bought Intel’s NAND flash operations in Dalian”. “Japan is likely to take a similar but lower-profile approach. Taiwan has reportedly already agreed to join but is unlikely to put its enormous electronics trade with China at risk. American tech companies will defend their own interests in consultations with the US Commerce Department”.
Jeff Pao notes “Chinese commentators said China’s chip sector could survive by focusing on making 28-55nm chips and mid-end memory semiconductors in the coming years while forming stronger ties with Japanese and South Korean chip companies. In August, Nikon announced that it planned to triple its sales of lithography tools by the year ending on March 31, 2026, from the average annual sales it earned in 2020-2022. At the time, before the US announced its latest chip ban, the Japanese company told the media that it would have a stronger focus on China’s markets. In April, Canon said it would start shipping its newly-developed 3D lithography tools, which can make wafers with multiple layers, in the first half of next year. Currently, semiconductor makers use stacking technologies to pile up wafers to increase the storage of a memory chip. Bai Yiwen…wrote in an article…that while ASML was facing pressure from the US to stop exporting DUV tools to China, Japanese lithography giants specializing in low and mid-level chip-making tools would welcome the new business opportunities. Bai said Nikon’s advanced packaging lithography tools, which are used to stack wafers to make memory chips, could help break ASML’s current monopoly on high-end chip-making equipment. That, however, likely explains why the US is now pressuring Japan to adhere to its export bans on China.
In his regard, Jeff Pao reports the Japanese government was now holding hearings with domestic semiconductor companies on the US chip export ban on China, according to Nikkei. Japan is the second country after the Netherlands to be requested by the US to follow its curbs including on DUV lithography chip-making equipment…Japan’s Nikon and Canon exported 29 and 140 DUV lithography units, respectively, last year. Although Nikkon sold less DUV lithography equipment than Canon, it shipped four units of ArFi and three units of ArF high-end chipmaking machines while Canon focused on low-end KrF and i-line units. ASML Holding NV, the world’s largest chip-making tool supplier, shipped 309 lithography tools last year, including 42 EUV machines that are used in making high-end chips smaller than 22nm…If the US succeeds in extending its curbs…China will not be able to produce most commonly-used semiconductors”. “China has warned Japan of undisclosed consequences if it follows the United States in imposing chip export bans against it, urging Tokyo instead to reject what it sees as Washington’s bullying behavior”.
Hideki Uno notes “the Ministry of Economy, Trade, and Industry (METI) has prepared a Semiconductor and Digital Industry Strategy to secure budget funding for greater public-private investment in the semiconductor industry”. Other Ministries are likely to “to submit additional funding requests for semiconductor-related projects” to support the achievement of “the strategic autonomy and strategic indispensability of Japan’s semiconductor industry. Scott Foster notes Japan’s “Ministry of Economy, Trade and Industry (METI)’s basic strategy for the revival of Japan’s semiconductor industry is: (Step 1) Urgent reinforcement of semiconductor production base for IoT; (Step 2) Develop next-generation semiconductor technology infrastructure through Japan-US collaboration; and (Step 3) Future technology infrastructure through global collaboration, including the realization of future technologies such as optoelectronic convergence. “Tokyo pouring money into chip-making capacity in last chance move to catch up with TSMC, Samsung, and Intel…Japan is not uniformly behind the curve. It has a significant presence in memory ICs and is a leader in image sensors and silicon carbide power devices; its production equipment and materials such as silicon wafers and photoresists are essential to the industry. But its weakness in advanced logic and dependence on TSMC make it as vulnerable as the US to a potential disruption of chip supplies from Taiwan”.
In terms of Japan-US collaboration, Dan Robinson notes there are plans to establish a joint research hub “with the goal of developing and putting in place the ability to mass produce 2nm semiconductors by the latter half of the decade”. “Participating institutions include the University of Tokyo, the National Institute of Advanced Industrial Science and Technology and science institute Riken, as well as research institutions from Europe and the US…According to Nikkei, the 450 billion yen will be spent on building production hubs for advanced semiconductors, with subsidies for companies such as TSMC, Kioxia, and Micron Technology to site semiconductor fabrication plants in Japan…370 billion yen will be spent on strengthening supply chains needed to bring to Japan essential materials such as those needed for producing silicon wafers”.
Scott Foster notes Japan has set up a company Rapidus which intends “to start mass production using 2nm process technology in 2025. Rapidus is aiming for 2027…Japan will join Taiwan, South Korea and the US as a provider of leading-edge foundry services by the end of the decade”. In addition, Tokyo-based Mercuria Investment reportedly plans to buy factories to address the shortage of semiconductor foundry capacity and recently announced it agreed to buy ON Semiconductor (onsemi)’s factory in Niigata. Onsemi Niigata – to be renamed JS Foundry – produces analog and power devices for the auto industry will be upgraded to provide contract manufacturing services to customers in the electric vehicle, renewable energy and other industries.
Mathieu Duchâtel points out “in semiconductor manufacturing, Europe’s share of global production capacity has declined from 24% in 2000 to 8% in 2021. Europe needs to reverse this downward trajectory…Europe is in the top six, with three large IDMs (STMicroelectronics, Infineon, and NXP), the world’s photolithography leader (ASML), a dynamic start-up scene, and leading research centers for nanoelectronics (IMEC, CEA-Leti, Fraunhofer). Yet, despite those strengths, Europe lacks world-leading fabless firms, and there is no foundry at the most advanced technological nodes in Europe, even if such a manufacturing process relies on a European technology – EUV lithography”. He explains that Europe is at a disadvantage because it lacks an industrial policy. “Europe is more restrictive than its competitors when providing public support for industrial production. The European Union (EU) has a traditional preference for supporting fundamental research and innovation rather than industrial production and imposes stringent competition rules to ensure the integrity of the single market”.
Mathieu Duchâtel points out “more fundamentally, Europe needs to face the reality of the US and Japan increasingly using controls as a tool in a global competition for technological supremacy…If the largest economic powerhouses use controls in combination with industrial policies and subsidies to compete with China, does Europe need to adjust? The Commission’s open strategic autonomy, which emphasizes the EU’s desire for openness and its strategic interest in multilateral norms of behavior, carries the opposite message. How to address that contradiction is a key question for Europe’s competitiveness…The main question for the EU is where to place limits on transatlantic cooperation, given in particular the uncertainties surrounding the future of US-China managed trade, the exceptions that the US often creates to its own rules to serve the interests of American companies at the expense of its competitors among its own allies and friends, and the use of extraterritorial legislation against Europe. The transatlantic partnership is an important horizon for the EU to conduct its own trade and technology policy towards China, but the risk of seeing the Biden administration further sacrificing European interests to other priorities should not disappear behind the promises of the TTC…For Europe, increased engagement with Japan will be an important complement to the transatlantic partnership. And for emerging technologies affecting the military balance, trilateral EU-Japan-US formats will be relevant… Europe should take a broader view of the strategic implications of technology transfers to China and deepen cooperation with the United States and Japan when cooperation can make European policy tools more efficient… Industrial policy is an essential tool, and a European strategy in the age of US-China technology competition should combine control instruments with industrial policy tools”.
Mathieu Duchâtel notes ‘the European Commission presented on February 8, 2022 its proposal for the Chips Act, a regulation that establishes a framework of measures for strengthening Europe’s semiconductor ecosystem. The first pillar of the Chips Act is the Chips for Europe Initiative. It explicitly aims at addressing several gaps in the European chips ecosystem”; particularly “design capacities, by focusing on Electronic Design Automation tools (IC design software) and RISC-V processor architecture. Production innovation, by supporting the development of pilot lines for FDSOI, leading edge nodes below 2 nm, 3D heterogeneous system integration and quantum chips”. “The real novelty…is the creation of a Chips Fund. The Commission has recognized the significant shortages in access to finance faced by start-ups and SMEs in the European semiconductor sector. In practice, the Chips Fund brings under the same umbrella various existing EU institutions and mechanisms”. “The cornerstone of the Chips Act is the creation of an exceptional competition regime to meet the needs of the semiconductor sector. The Act incorporates the new legal concept of first-of-a-kind facility…allows for approving State aid when a production facility goes beyond the Union’s state-of-the-art, for instance in terms of technology node, substrate material, such as silicon carbide and gallium nitride, and other product innovation that can offer better performance, process technology or energy and environmental performance”. “The second central concept to the Chips Act is the funding gap…Under the Act, it may be justified to cover with public resources up to 100% of a proven funding gap, if such facilities would otherwise not exist in Europe…The Chips Act removes a major legal roadblock to enable the EU’s approval of State aid”. “At the beginning of 2022, three projects are rumoured to seek a State aid package. TSMC in Dresden, Intel in Magdeburg, Germany and GlobalFoundries, in an undisclosed location. Intel’s foundry plan is the only fab that would bring 7 nm technology to Europe and seek production in volume”.
Mathieu Duchâtel adds “learning from the semiconductor shortages and the Covid-19 crisis, the Chips Act sets out measures to anticipate and respond to supply chain crises. Inspired by the Commission’s management of Covid-19 vaccine procurement, it first creates a legal basis for common purchasing and priority orders. It also institutes a European Semiconductor Board with representatives from the 27 Member States. This advisory body will be the main governance partner of the Commission to address and manage supply chain risks, and a lot will depend on its efficiency”. “It seems clear that the Chips Act will improve the EU’s overview of supply chain risks. Indeed, it requires regular monitoring and information sharing between the Commission and Member States. If managed efficiently, this mechanism will ensure early warning to supply crises…In case of crisis, the Chips Act entitles the Commission to impose an obligation for Integrated Production Facilities and Open EU Foundries to accept and prioritize an order of crisis-relevant products (‘priority rated order’), [an] obligation [that] shall take precedence over any performance obligation under private or public law. Article 21 of the proposed regulation makes clear that this obligation applies to other undertakings having benefited from public support”.
US firms are not yet convinced that its allied competitors will support the US bans. Dan Robinson reports “some US companies are alarmed at the amount of revenue they may lose from sales to China as part of complying with the new restrictions imposed by Washington, and also worried that the business may simply go to rivals in other countries instead….California-based Lam Research…warned that the new export controls could cost it as much as $2.5 billion in lost revenue in 2023. Applied Materials…that the new rules could lose the company up to half a billion dollars in sales” during its current quarter. “According to reports, US Secretary of Commerce Gina Raimondo told makers of chip manufacturing kit that they may need to wait as long as nine months before the Washington government can persuade allied nations to get on board with its program of measures to limit China’s access to advanced semiconductor technologies. The US is said “to be working on an agreement that would see companies in the Netherlands and Japan also subject to restrictions on sales of semiconductor equipment to China”.
Dan Robinson speculates the US could use the same strategies it has deployed against Huawei to “browbeat friendly governments into taking similar actions”. The UK and Germany initially resisted requests to ban Huawei. However, the US “began a campaign to compel other countries follow suit, heavily pressuring allies such as the UK with warnings that Washington might curtail the sharing of intelligence with Britain if Huawei continued to have a major presence in its telecoms networks…However, US sanctions that cut off the supply of western chip technology to Huawei forced the UK and others to fall in line, and the British government last month issued formal legal notices to telecoms operators instructing them to remove Huawei technology from the country’s 5G networks by the end of 2027. It isn’t clear what steps America will take to convince its allies to go along with its new export controls on China, but many chip firms have already been warning that the measures could have an adverse effect on their profits in a market already facing a downturn”.
Close collaboration between allies would isolate China but it would push China to move more aggressively on replacing OECD inputs. But allied unity is not yet achieved and it may be difficult to maintain due to intense competition among the major players[7]. Some OECD firms will commit to one sphere while others are still trying to straddle both spheres. While the rules of the game and landscape are evolving, nonetheless fragmentation appears inevitable. The wildcard in the semiconductor landscape is India[8]. India, possessing both the depth in talent and market scale, is bidding to replace China in the semiconductor supply chain and to become a leading player and could have a disruptive impact on the industry landscape.
Impact on China and potential retaliation
- China’s semiconductor industry pre-Oct 7
Earlier decoupling initiatives caused a setback but did not halt China’s progress. Scott Foster note IC Insights data shows “$31.2 billion worth of semiconductors were produced in China in 2021, of which $12.3 billion were made by Chinese companies. The rest came from the local fabs of TSMC, Samsung, Intel and other foreign companies. In comparison, China consumed $186.5 billion worth of semiconductors in 2021, accounting for 36.5% of the world market. Only 17% of Chinese semiconductor demand was met by production in China and only 7% by Chinese companies…A year earlier, in 2020, $22.7 billion worth of semiconductors were produced in China, of which $8.3 billion were made by Chinese companies…Total Chinese demand amounted to $143.4 billion and China accounted for 36% of the global market”. Overall, total semiconductor production in China increased by 37% in 2021 while production by Chinese companies increased by 48%. Over the past decade, semiconductor production in China has more than tripled”.
Xie Jun, Qi Xijia and Liu Yang reports “the country’s chip giant Semiconductor Manufacturing International Corp (SMIC) said it had made a breakthrough in the first generation FinFET technology and entered mass production in Q4 of 2019, while the tech’s second generation, rendered equivalent to the 7nm and 5nm manufacturing process of TSMC, is also in a period of pilot production”. Some experts think China is on course to overcome several technological bottlenecks after 3-5 years of stable development. As compared with “chips from Europe, the US and South Korea, China’s domestic chips are of good quality, priced about 60 percent lower than that of other countries…In a sense, Chinese companies already have the ability to produce high-end chips, and they just need time to achieve mass production”.
Qi Xijia notes “Shanghai-based firms have achieved mass production of semiconductors with 14-nm process and made breakthroughs in 90-nm lithography machines, 5-nm etching machines, 12-inch large silicon wafers, central processing units and 5G chips”. According to Chen Jia, “the large-scale production of 14-nm chips in Shanghai will greatly help the development of such sectors as new-energy cars, smart cities, intelligent manufacturing and the Internet of Things, which will help China consolidate its advantage as the world’s top manufacturing factory. “With the completion of Shanghai’s industry cluster for the 14-nm chips, more advanced projects in the 7- and 5-nm processes will be accelerated”. “Shanghai is the backbone of the nation’s semiconductors industry, with a market size reaching 250 billion yuan ($35.91 billion) in 2021, accounting for 25 percent of the nation’s total. More than 1,000 leading enterprises have settled in Shanghai, attracting 40 percent of the talent nationwide, officials said. At the beginning this year, the Shanghai municipal government announced a set of new policies to bolster China’s advanced chip-making capacity”. Global Times add “China is building an average of about 100 new chip production lines every year as part of the great industrial expansion to strengthen the safety and stability of its supply chain”.
Others have a different view on China’s progress. Sam Bresnick and Nathaniel Sher argue “SMIC’s achievement may not be quite as momentous as it seems…Instead, it points toward Beijing’s continued inability to achieve self-sufficiency in a highly specialized and globalized industry. Questions remain about the new chips’ yield, memory capacity, and energy efficiency, as well as whether the company can manufacture them at scale. This is largely because SMIC was forced to use suboptimal manufacturing techniques with decades-old DUV manufacturing equipment. In short, not all 7nm chips are created equal…SMIC’s latest breakthrough may very well be its last. More broadly, while Chinese companies increase their chip production, they remain unable to compete with the Taiwanese, South Korean, and American companies that produce most of the world’s cutting-edge semiconductors. This is because, due to effective export controls, China lacks the EUV lithography software and equipment that have fueled the latest leaps in chip technology. Beijing has some advanced technical know-how, but it does not have enough of the best design and manufacturing tools to become globally competitive”. However, they admit “broad bans on older technology may do more to hurt U.S. semiconductor firms than to protect U.S. national security…Historically, chipmakers have cut back on R&D spending during periods of declining revenue. Second, due to the globalized nature of the semiconductor industry, the United States will need to bring allies and partners on board before implementing more expansive controls. Otherwise, U.S. companies will be unable to sell to their largest market, while China continues to buy the tools it needs from other foreign suppliers. Coaxing South Korea, Japan, and the Netherlands into accepting new U.S. controls may prove difficult, as their semiconductor equipment manufacturers would lose billions in sales. If pushed too far, leading chipmakers could decide to establish new foundries without American technology to skirt expansive U.S. controls”. The conclude “despite investing billions in the semiconductor industry, China controls less than one percent of electronic design software and semiconductor tools and materials. All of this means that the U.S. should stop wasting its time on restricting the export of older, less critical technologies. Instead, Washington should focus on pushing the technological frontier forward”.
- Why now – motives for Oct 7
Why did the US choose this moment to take the tech battle to the next level – given its supply chain relocation initiatives are incomplete. One point is that past measures have not deterred China from pursuing self-sufficiency and this was probably the last opportunity to halt China’s advance before the US advantage evaporates.
Semiconductor Industry Association (SIA) notes China’s rapid progress. “If China’s semiconductor development continues its strong momentum – maintaining 30% CAGR over the next three years – and assuming growth rates of industries in other countries stay the same, the Chinese semiconductor industry could generate $116 billion in annual revenue by 2024, capturing upwards of 17.4% of global market share. This would place China behind only the United States and South Korea in global market share. Equally startling…nearly 15,000 Chinese firms registered as semiconductor enterprises in 2020. A large number of these new firms are fabless start-ups specializing in GPU, EDA, FPGA, AI computing, and other higher-end chip design. Many of these firms are developing advanced chips, designing and taping out devices on bleeding-edge process nodes. Sales of Chinese high-end logic devices are also accelerating, with the combined revenue of China’s CPU, GPU, and FPGA sectors growing at an annual rate of 128% to nearly $1 billion in revenue in 2020, up from a meager $60 million in 2015…Across all four subsegments of the Chinese semiconductor supply chain – fabless, IDM, foundry, and OSAT – Chinese firms recorded rapid increases in revenue last year, representing annual growth rates of 36%, 23%, 32%, 23%, respectively, based on an SIA analysis…28 additional fab construction projects totaling $26 billion in new planned funding announced in 2021…from September 2020 to November 2021, Chinese wafer manufacturers have added nearly 500K wafer per month (WPM) capacities in trailing nodes (>=14nm), and only an additional 10K in capacity for advanced nodes. China’s wafer capacity increase alone accounted for 26% of the worldwide total”.
A drastic approach was needed to halt China’s advance. The US is therefore drawing the line in the sand; getting MNCs declare their loyalties and stopping them dragging their feet on relocating their business out of China. Much was made of the “ready or not” nature of the Oct 7 measures where the lack of widespread consultation raised suspicions the intention was to circumvent industry lobbying. David P. Goldman notes the Oct 7 initiatives baffled chip industry experts who “concluded that the new policy was rushed into effect in panic mode, without weighing its civilian or military implications”. Experts point out China already has the technical capability to produce advanced chips to satisfy military needs, just not on a commercial basis. In any case, “military systems use older chips that China makes at home, according to a 2022 RAND Corporation report. But if China requires the most advanced 5nm chips for AI-driven military applications – for example, drone swarms controlled by a 5G broadband network and guided by artificial intelligence – China can make enough of them, although at high cost. The US ban won’t affect weapons systems, but it will delay China’s rollout of autonomous vehicles, data centers and other civilian applications”. Experts note “the shoddiness of the Commerce Department’s technical specifications and the exclusion of US industry from the policy loop suggests a sudden onset of panic in the Biden administration over China’s technological advancement” and setting the controls at levels that China can already produce (14-16nm logic chips) indicate “a shift away from trying to freeze or hold back China’s advances in chips. I see this as an attempt to roll back and degrade China’s existing capabilities”.
My big picture view is that the Oct 7 export bans sends a powerful message: Whatever the costs, OECD allies and MNCs are expected to align with the US geopolitical goals.
There is considerable guideline ambiguity and the US has not yet made clear how broadly it would enforce the new controls, just the appearance of intend that restrictions could be further tightened will suffice enough to steer MNCs away from China. The impact could extend well beyond semiconductors and affect all industries and products that rely on semiconductors. This would give a greater push to OECD MNCs relocate the whole semiconductor ecosystem – including end products – to outside of China. Bifurcation is starting to emerge as OECD and China travel on separate paths for components, networks and operating systems.
- October 7 impact on China.
Views on the impact of the Oct 7 measures on China diverge. Jeff Pao notes some industry observers think the chip export ban “has widened the technology gap between Chinese chip makers and global industry players…China’s plan to self-supply 70% of its semiconductors by 2025 and become a hub of global chip foundries by 2030 has basically vanished due to the US curbs. The columnist said China’s SMIC and Hua Hong Semiconductor can now produce only 28-nanometer (nm) chips, instead of the more advanced 14-16 nm chips”. Some think the ban “can hurt China’s ability to develop supercomputers capable of being used to direct hypersonic missiles and develop nuclear weapons”.
Ben Thompson calls it “a daunting challenge: it’s not just that China needs to re-create TSMC, but also ASML, Lam Research, Applied Materials, Tokyo Electronic, and all of the other pieces of the foundry supply chain. And, to go one layer deeper, not only does China need to re-create ASML, but also Zeiss, and TRUMPF, and Access Laser, and all of the other pieces of the global supply chain, much of which is not located in China. China’s manufacturing prowess is centered on traditionally labor-centric components; even though Chinese labor is now much more expensive than it was, and automation much more common, path dependency matters, and China’s capability is massive but in some respects limited…China is going to need to build up these capabilities from the ground up, and it’s going to be a long hard road. Moreover, China will not have the benefit of partnership and distributed expertise that have driven the last decade of innovation: in some respects China is going to need to be Intel, doing too much on its own”.
However, Ben Thompson notes China has three big advantages. “First, it is much easier to follow a path than to forge a new one. China may not be able to make EUV machines, but at least they know they can be made…Second, China has benefited from all of the technological sharing to date: SMIC has successfully manufactured 7nm chips (using ASML’s immersion lithography machines), and Shanghai Micro Electronics Equipment (SMEE) has built its own immersion lithography machines. Granted, those 7nm chips almost certainly had poor yields, and the trick is for SMIC to use SMEE on the cutting edge, but that leads to the third point: China has unlimited money and infinite motivation to figure this out…Money does, though, pay for processes that don’t have great yields: the problem for Intel at 7 nanometer, for example, wasn’t that they couldn’t make chips, but that they couldn’t get yields high enough to make them economically. That won’t be a concern for China when it comes to chips for military applications. What is more meaningful, though, will be the alignment of China’s private sector behind China’s chip companies: TSMC didn’t only need ASML, it also needed Apple and AMD and Nvidia, end users who were both willing to pay for performance and also work deeply with TSMC to figure out generation after generation of faster chips. Tencent and Alibaba and Baidu will now join Huawei in being the China chip industry’s most demanding customers, in the best possible sense”.
Ben Thompson argues “China that picked up a lot of the slack…building more basic chips using older technologies. China’s share of >45 nanometer chips was 23% in 2019, and probably over 35% today; its share of 28-45 nanometer chips was 19% in 2019 and is probably approaching 30% today. Moreover, these chips still make up most of the volume for the industry…The Biden administration’s sanctions are designed to not touch this part of the industry: the limitations are on high end fabs and the equipment and people that go into them, not trailing edge fabs that make up most of this volume. There is good reason for this: these trailing edge factories are still using a lot of U.S. equipment; for most equipment makers China is responsible for around a third of their revenue. That means cutting off trailing edge fabs would have two deleterious effects on the U.S.: a huge number of the products U.S. consumers buy would falter for lack of chips, even as the same U.S. companies that have built the advantage the administration is seeking to exploit would have their revenue (and future ability to invest in R&D) impaired. It’s worth pointing out, though, that this is producing a new kind of liability for the U.S., and potentially more danger for Taiwan. Go back to Intel’s strategy of selling off and/or reusing its old fabs, which again, made sense given the path Intel started on decades ago: that means that Intel, unlike TSMC, doesn’t have any trailing edge capacity (outside of what it acquired in the Tower Semiconductor deal). GlobalFoundries, the U.S.’s other foundry, had the same model as Intel while it was the manufacturing arm of AMD; GlobalFoundries acquired trailing edge capacity with its acquisition of Chartered Semiconductor, but there is a reason why the U.S. >45 nanometer market share was only 9% in 2019 (and likely lower today), and 28-45 nanometer market share was a mere 6% (and again, likely lower today). Again, these aren’t difficult chips to make, but that is precisely why it makes little sense to build new trailing edge foundries in the U.S.: Taiwan already has it covered (with the largest marketshare in both categories), and China has the motivation to build more just so it can learn….What would be much more difficult to replace are, paradoxically, trailing node chips, made in fabs that Intel long ago abandoned. China meanwhile, has had good reason to keep TSMC around, even as it built up its own trailing edge fabs: the country needs cutting edge chips, and TSMC makes them. However, if those chips are cut off, then what use is TSMC to China…the more that China builds up its chip capabilities – even if that is only at trailing nodes – the more motivation there is to make TSMC a target, not only to deny the U.S. its advanced capabilities, but also the basic chips that are more integral to everyday life than we ever realized. So is this chip ban the right move? In the medium term, the impacts will be significant, particularly in terms of the stated target of these sanctions – AI…In the long run, meanwhile, the U.S. may have given up what would have been, thanks to the sheer amount of cost and learning curve distance involved, a permanent economic advantage. Absent politics there simply is no reason to compete with TSMC or ASML or any of the other specialized parts of the supply chain; it would simply be easier to buy instead of build. Now, though, it is possible to envision a future where China undercuts U.S. companies in chips just like they once did in more labor-intensive industries, even as its own AI capabilities catch up and, given China’s demonstrated willingness to use technology in deeply intrusive ways, potentially surpass the West with its concerns about privacy and property rights”.
The Oct 7 measures seems to be speeding up relocation. Dong-Joon Kwon notes the industry is reporting “the demand for proof of origin of semiconductors to exclude Chinese-made products is spreading in the fabless industry since July to August. It seems like most of US customers are trying to avoid various sanctions that may arise from the US government checking on Chinese semiconductors. Some of the companies are proving the country of origin by changing their Taiwanese foundry production site to Taiwan instead of Republic of China. Supply to the US may be blocked if China is indicated and this might be connected to US customers requesting for the origin of the products… It will affect the semiconductor production supply chain due it being a De-Sinicization foundry….It is highly likely that a reorganization in foundry supply chain will occur in order to respond to American customers. It can be an opportunity for Korean fabless, which has a low proportion of mass produced products in China”. On this note, “the proportion of Asia, Europe, and North America including Korea shows decrease in sales when looking SMIC’s sales by region in the second quarter. It is interpreted that the US’s move to exclude Chinese products as well as restrictions on exports of materials, parts, and equipment has had an impact. On the other hand, SMIC’s share of China and Hong Kong continued to increase”. While foreign plants in China obtained 1-year exemptions from the US export controls, but future renewals are uncertain.
Some MNCs are reluctant to abandon the China market. Laura Dobberstein reports “systems that once contained Nvidia and TSMC chips, which are now restricted by the US government, are popping up this week with slower specs to meet US export controls to China and evade the hassles of obtaining special licenses. Companies in China are swapping out A100 Graphic Processing Unit (GPU) in its NF5688MG [PDF] for an A800 in servers….The Nvidia A800 GPU, which went into production in Q3 2022, is another alternative product to the Nvidia A100 GPU for customers in China. The A800 meets the US government’s clear test for reduced export control and cannot be programmed to exceed it”. But the overall picture is murky. “Nvidia warned in a late August SEC filing that new US export license requirements could hinder the development of the H100 and support for A100 customers and potentially require some operations be moved out of China…Meanwhile, GPUs still in development and produced by Taiwan Semiconductor Manufacturing Company (TSMC) under contract for Alibaba and startup Biren Technology also reportedly have reduced processor speeds. China’s domestic chip facilities still have a long way to go to be globally competitive. However, Biren did have a frontrunner in the making, but amid uncertainty about its potential legality, TSMC had reportedly put the brakes on the silicon”.
Jonathan Goldberg is sceptical of the efficacy of export bans. Firstly, “the biggest force working to dilute past efforts was US companies themselves. Nvidia’s shares fell on the latest news, and they warned of a $400 million revenue shortfall resulting from the ban…Secondly, how will this work in a modern supply chain? Let’s say a US company wants to buy a few million dollars worth of banned Nvidia chips. Who will then assemble the chips into working systems, inserting the chips onto motherboards and installing those in server racks? Today, most of that work is done by Taiwanese companies’ factories in China. Can Nvidia ship those chips to China? Technically, we think the answer is yes, but it is easy to see how this process can get easily derailed. Third, this is a short-term move, but it has long term consequences…every Chinese company buying parts from US vendors is now busy looking for domestic alternatives…The US government’s actions are directly leading to interest in a sea of aspiring Nvidia competitors. For companies like Biren, the US government’s actions are a major boost. And those are just the first order effects…one of Nvidia’s biggest advantages in the market for AI chips is its CUDA software. It is now highly unclear what the status of this software is in China. This will certainly increase interest in open source alternatives to CUDA. So not only is Nvidia losing out in direct sales, it now risks seeing its competitive advantage being worn away. This also extends beyond the GPU/AI semiconductor space, it applies to all US semis, what we would call third order effects. All Chinese companies, even those with absolutely no ties to China’s military now have to look for alternative vendors. This is not patriotism, it is just rational commercial contingency planning. The effects are most likely to be felt first in industrial and automotive semis — i.e. far away from leading edge GPUs, as this is the area where China’s aspiring chip companies look most competitive today…So while we understand the US government’s interest in curbing the supply of high technology to a potential long-term adversary, we have to wonder if over that long-term this works to China’s advantage. There are no simple answers to this problem. That said, the US government needs to think very strategically here. Is the goal to limit specific Chinese military projects? Is it to cripple the entire Chinese semis complex? Is there even an end goal? From what we can tell right now, that does not seem to be the case”.
Long term, geopolitical force will likely prove insufficient to dictate economic outcomes. Economic outcomes will be largely determined by the reactions of chip industry players – old and new – to government policies and, more importantly, commercial imperatives. David P. Goldman thinks “the damage to capital investment and R&D in the Western semiconductor industry will exceed Washington’s modest subsidies for the chip industry by a factor of five or more…The industry had already cut capital investment plans from about US$200 billion to $160 billion for 2022. US restrictions on exports of semiconductor equipment, design tools and high-end chips to China will shrink revenues further, putting an air pocket into R&D and capital expansion. The world’s dominant chip fabricator, Taiwan’s TSMC, planned $44 billion in CapEx just six months ago but on Wednesday announced a cut to $36 billion”. “The US measures won’t affect China’s sensors, satellite surveillance, military guidance and other strategic systems because the vast majority of military applications use older chips that China can produce at home. But it may postpone autonomous driving, cloud computing and other efforts to digitize China’s economy. It will also elicit an all-out Chinese effort to replace American chip-making and design technology. CapEx and R&D will shrink drastically in the US semiconductor industry while China allocates a massive budget to the sector…On a five- or ten-year horizon, America’s technological edge in semiconductor design and fabrication is likely to vanish. As capital budgets collapse in the Western semiconductor industry, the damage to the US and other Western economies is likely to be greater than the harm inflicted on China”.
China can explore different paths to bypass the US chokeholds. Jean-Yves Larguier points out “China is taking full advantage of the incredible dynamic around RISC-V, its open-source approach and growing ecosystem…is becoming a credible alternative to the dominant x86 and ARM architectures…While RISC-V will not be enough for China to achieve semiconductor independence, it allows China to be less vulnerable to intellectual property, without isolating itself in a proprietary ecosystem restricted to China and a few partners”. RISC-V International had already moved from US to Switzerland in 2019, “a clear sign of its wish to remain independent from the US…As of 2022, Chinese players hold half of the seats on the board of directors of RISC-V International and numerous technical steering committee seats. These positions give China extensive decision-making powers regarding the evolution of the RISC-V standard and its extensions, as well as the RISC-V technical programs and roadmap, that best serve their future application needs”. China has also set up local RISC-V organizations such as the China Open Command Ecosystem (RISC-V) Alliance (CRVA) and the China RISC-V Industry Consortium (CRVIC). Chinese companies such as Alibaba is leading many projects around RISC-V. “On the chip design side, its subsidiary T-Head Semiconductor develops high-performance RISC-V cores (XuanTie) and RISC-V chips for artificial intelligence, and actively contributes to the open-source community by sharing some of its design of XuanTie processor cores and related software. Alibaba Cloud is also working on porting Android to the RISC-V architecture in collaboration with Google. This major contribution to the RISC-V ecosystem paves the way for the Android world and will place RISC-V in even more direct competition with ARM”.
Jean-Yves Larguier cautions “sovereign withdrawal trends in this segment of the microelectronics industry could ultimately break the momentum of RISC-V. The authorities of certain countries might be tempted to limit the contribution of their companies and universities to the open-source community, to the point that decoupling of open-source data is not an unlikely scenario. However, China is surely preparing itself for the worst scenario. Hasn’t it already set up its own open-source sharing platform, Gitee, like the American GitHub for fear of sanctions? In the event of a split within the RISC-V community, China would certainly have the means to continue to innovate, but the main challenge would be to bring together enough international players so as not to find itself isolated and therefore fall behind a mainstream community”.
OECD MNCs have to figure out their China strategies and what to do with their stranded China assets. Already the industry faces a market environment of falling global demand, expanding capacity, the emergence of new competitors, higher financing costs, a slowdown in the pace of innovation (i.e. can’t buy each other’s patents, exchange knowledge, do partnerships or takeovers) while supply chains are unreliable (don’t know where the next blow is coming from). The advantage of China’s MNCs is that they are less concerned with profits and have strong support from their government.
Estimates vary on how quickly China can make up lost ground. Sujai Shivakumar, Charles Wessner and Thomas Howell report Peter Wennink (CEO of ASML) thinks “in 15 years’ time, they’ll be able to do it all by themselves – and [Western equipment suppliers’] market…will be gone”. “Tudor Brown, a former independent director at SMIC, concurs: The US is being naive if it thinks this is going to slow them down for any length of time. I think it will slow them down for two to five years, not 10”. “Chinese chip firms are also reported to be developing technological workarounds that help bridge the gap with more advanced Western chips, including clever packaging and maximum use of DUV ArF (argon fluoride) immersion lithography…Chinese semiconductor equipment maker AMEC is making etching equipment of sufficient quality that it is reportedly being tested by the world’s most advanced semiconductor device makers, including TSMC, Samsung, Intel, and Micron Technologies…And a Chinese lithography equipment maker, SMEE, is reportedly working on an ArF immersion lithography tool that can be used to produce semiconductor devices at the 5 nm node”.
Sujai Shivakumar, Charles Wessner and Thomas Howell note though “even as Chinese chipmakers work hard to catch up, Intel, Samsung, and TSMC are still racing ahead…Reflecting these realities, China to date has not adopted a go-it-alone developmental strategy. Rather, the Chinese effort emphasizes acquisition and absorption of foreign chip technologies through purchase of foreign semiconductor enterprises or theft of their technology. Their approach also includes attracting talent from foreign chip firms, bargaining with foreign firms for technology transfer in exchange for access to the Chinese market, and a policy of encouraging indigenous innovation by Chinese enterprises – defined as enhancing original innovation through co-innovation and re-innovation based on the assimilation of imported technologies. To drive technology transfers, Chinese antimonopoly law enforcers have levied massive fines against foreign semiconductor firms for alleged excessively high technology licensing fees and unreasonable terms for patent licensing”.
- When rather than how China will retaliate
There is speculation on the troubling question of potential China retaliation. Many think China’s options are limited to either leveraging on its rare earth monopoly or harassing foreign MNCs. Some suggest such measures have adverse effects on China’s already struggling economy and that China has little to gain from escalation.
I think these immediate and direct responses are off the mark and easily gives up China’s strategic advantage. On rare earth, China would probably follow the OPEC/Russian playbook of maintaining the adversary’s dependency and opting instead to ration supply and increase prices. In the meantime, China would establish new price discovery centers and seek to price raw earth in yuan. Why should China pick a fight with foreign MNCs? The US is the one doing all the hard running and, in the process, expanding a lot of goodwill to force OECD firms and individuals to sever business dealings with China. China might as well be a passive observer and allow the consequences to sink in on US allies and MNCs, let them fight among themselves as they figure how to survive if they aren’t able to upgrade, expand their plants and sell their latest technology in the China market. China can afford to play the “victim” and file formal complaints that the US is not following international rules. In any case, the supply chain hasn’t relocated and China has yet to achieve self-sufficiency. Therefore China would keep its door open to foreign talent and MNCs willing to “reorganise” to do business with China. After all, it is probably more annoying to the US government that OECD MNCs continue to do business with China.
In other words, China’s non-response isn’t because it is avoiding escalation but because it is not yet time to retaliate. China is following a familiar guerrilla warfare strategy: “The enemy advances, we retreat; the enemy camps, we harass; the enemy tires, we attack; the enemy retreats, we pursue”[9]. Hence, “to defeat a formidable enemy, revolutionary armed forces should not fight with a reckless disregard for the consequences when there is a great disparity between their own strength and their enemy’s…that we should fight no battle unprepared and fight no battle we are not sure of winning; and that in any battle we fight we should develop our army’s strong points and its excellent style of fighting. These are the major principles of fighting a war of annihilation”.
My view is that China is likely to stick to its long-term path of gradually opening its economy while prioritising the goal of self-sufficiency across the entire semiconductor supply chain. There are two possible responses. One is an asymmetric response – striking in in an area where the US is vulnerable such as in finance. The other is waiting for the right time. Direct retaliation will be more effective once the new “allied” factories are operational and after OECD MNCs exit from China. The semiconductor supply chain is quite long and there are many vulnerable interdependencies to target.
China is anyway pressing ahead with the removal of “OECD inputs” and encouraging support of local production. Nuying Huang and Jessie Shen reports “Chinese car vendors have asked chip suppliers, including foreign ones, to place wafer starts for their demand with China-based foundries as China accelerates its push for IC self-sufficiency…Foreign foundries with fab operations in China are also being required by their Chinese automotive customers to allocate more of their fab capacities in China for automotive chips…According to sources at Taiwan-based IC design houses engaged in the supply chain of China’s auto companies, fabless chipmakers are being asked to place orders with local foundries in China. Automotive ICs are built mainly using mature process technologies, which are available at China-based foundries. For China-based foundries, gaining more orders shifted for automotive ICs may help them better compete with international IDMs, such as Infineon, NXP, Onsemi, Renesas, STMicroelectronics and TI…Nevertheless, the observers said, it is far from being enough to break these IDMs’ dominance in the automotive IC sector. The self-sufficiency rate of China’s automotive chips is about 5%, whereas Europe-, Japan- and US-based IDMs collectively hold an around 80% share of the global automotive IC market”.
One overlooked aspect is China gaining market leverage over older semiconductor technologies. Gregory C. Allen points out “legacy semiconductor designs made with older technology nodes (>28 nm) still play an important role in the global economy. There are two large markets where older chip technologies remain competitive: (1) chips for devices such as washing machines that do not require the latest technologies and (2) chips for safety critical systems such as aerospace, automotive, and infrastructure systems that are reluctant to change designs that have already made it through lengthy integrated system tests”.
Iain Morris notes the spillovers from “the impact of anti-China rules on US and other Western kit vendors. US policymakers see the companies backing this concept as potential alternatives to a neutered Huawei”. Yet, “overall unintended harm from the Huawei geopolitics that weakened the burgeoning ORAN movement…by blocking Chinese radios the entire investment in building [an] ORAN end-to-end supply chain had to be repeated in the worst supply chain crisis of our generation causing not only significant cost (and now in a capital markets-constrained environment) but material delays in ORAN progress”.
Overall, China’s weakness in chip manufacturing technology can be offset by its advantage in using information and its IOT infrastructure. In my view, data use offers a more enduring advantage than software and hardware which is subject to rapid obsolescence and substitutable. Chinese firms will move forward on a separate path and this will lead to technology bifurcation and ecosystem fragmentation.
The endgame
The semiconductor technology war is now in full blast. Reva Goujon, Lauren Dudley, Jan-Peter Kleinhans and Agatha Kratz point out “some US policymakers have argued that their intent is not to destroy US market share in China. On the contrary, if China’s indigenous development remains crippled and thus dependent on the import of US-origin chips for everything from data centers to smartphone devices, then the US can build-up its strategic leverage over China. However, this argumentation only holds so long as the United States avoids excessive application of FDPR. Just as Beijing is trying to steer its most dynamic private tech giants toward hard tech industrial priorities, including semiconductors, sprawling tech conglomerates developing AI chips alongside IoT devices may run a higher risk of getting blacklisted by the US. A proliferation of Huawei-like targets would inhibit the ability of US companies to sell to major Chinese tech consumers and undercut the strategic aim of building one-way dependencies”.
Yet, the October 7 measures are hard-hitting with irreversible effects in the sense that once doubts are planted, the damage is likely to be permanent. Similar to the pressure that led to Western corporate exodus from Russia, similar Western “forces” are clamouring for complete fragmentation; cheering on new export bans and blacklists, rejection of Chinese investments and technology in OECD, and MNC withdrawals from the China market.
In relation to this, there is a deafening industry silence on the future outlook. It is hard to contain the fall-out and the damage could be more extensive than many industry players would care to admit. This is because semiconductors are ubiquitous. Complete decoupling means that the bans could apply not only to semiconductor manufacturing technology but be extended to cover chips and products. At the product level, China has an outsized presence. The Semiconductor Industry Association (SIA) notes that although China accounts for only 7.6% of total global semiconductor sales, “in 2020, China imported a whopping $378 billion in semiconductors; assembled 35% of the world’s electronic devices; accounted for 30% to 70% (depending on the product) of the global TV, PC, and mobile phone exports; and consumed one-quarter of all semiconductor-enabled electronics. Access to this massive market is essential to the success of any globally competitive chip firm today and in the future”.
As an example, consider an extrapolated scenario for Apple. Like most MNCs, Apple seemed to have adopted a China+1 strategy, to relocate production facilities for OECD markets to outside China while relying on domestic plants and suppliers for the China market. Even then, immense pressure caused it to reverse its decision to purchase chips from YTMC. In the future, the new iphone models will be powered by technologically advanced chips. But it is unclear whether products incorporating advanced chips will be subject to the US export ban. Even if the US government agrees, it is unclear if the Chinese government and/or consumers would find it acceptable. This leads to the dilemma that Apple’s presence on both the production and consumer sides would be increasingly constrained over time and Apple should perhaps plan for a scenario whereby it may need to exit from China if geopolitical tensions worsen.
What about the European and North Asian players? Will they buckle under US pressure to comply with the sanctions or China pressure not to comply? Due to the adverse corporate consequences, allied support for the US October 7 measures has so far been lukewarm. Europe, Japanese, Korean and Taiwanese players are concerned by the collateral damage they have already suffered from tech decoupling. Yet, even if they don’t support the US sanctions, as China makes progress towards self-sufficiency, then they will lose out anyway.
While US and European MNCs can fall back on support from their governments, it is the Asian MNCs that are caught in a bigger dilemma. Global Times warns that with US and EU locked into a looming subsidy war, “the global economy will be characterized by more fragmentation and protectionist measures”. “The chip industry is highly globalized, and over the past few decades, as a result of the division of labor through globalization and the huge market demand, Asia has become the region with the most concentrated chip production capacity. Now as the US and the EU are both aiming to reconstruct the current pattern of the global semiconductor industry, the old balance of the industrial chain will be sabotaged, which will bringing shocks to the Asian semiconductor industrial chains. Indeed, semiconductors could only be one example of Asian industrial chain facing various risks. High-tech sectors such as new-energy vehicle manufacturing are also likely to bear the brunt of the rising anti-globalization sentiment. Under such circumstances, while adhering to support for globalization and free trade, Asian economies need not only to guard against the potential artificial shift of manufacturing sector to the West, but also to further strengthen regional industrial chain cooperation to fend off external risks. From another perspective, changes in external environment will offer opportunities for further integration of Asian industrial chains in the high-tech sector…for the sake of regional industries, major economies like China, Japan and South Korea should increase communication so as to enhance the regional competitiveness in the tech-intensive industrial chain. Moreover, China and ASEAN have highly complementary industrial chains and resource distribution structures. Therefore, as China seeks to upgrade its free trade zone with ASEAN, it is possible to make industrial chain cooperation and the development of a common market a focus to promote industrial upgrading and unleash economic potential”.
The trend towards regionalisation threatens to undermine the dominance of the Asian supply chain. As North Asian firms relocate production to Europe and North America, they undermine the significance of their home and Asian economies. The losses of the China market and Asian integrated supply chain could threaten their future survival. TSMC and Taiwan are an example. At the moment, TSMC’s, and Taiwan, global foundry dominance is being threatened on sides. It is likely that over the next five years, their global market share is likely to shrink considerably. Asian competitors need to cope with not only the macro effects of relocation out of China but also with global over-capacity as more countries seek self-sufficiency.
The endgame is thus becoming visible. China will “go-it-alone” to pursue self-sufficiency on a separate pathway to reduce the criticality of US technology chokeholds. In this regard, the advantage of US chokeholds is transient as technology becomes obsolete quickly and chokeholds can be easily by-passed or leapfrogged in peer-to-peer networks[10] and in IOT. Thus, fragmenting ecosystems appear inevitable. If OECD firms are not allowed to expand in China and Chinese firms are not allowed to expand in OECD, it implies that OECD firms will be eventually forced to, by one government or the other, to pull out of China’s markets. Chinese firms are already (forced) to re-direct their expansion into the Global South. Nonetheless, they benefit from whole-of-government support offering both investments and reciprocal market access. In contrast, OECD is at a disadvantage as their governments tend to threaten to cut investments and market access via secondary sanctions while firms are distracted by the operationalisation of their ex-China supply chains.
The competition for the Global South is in effect a contest of dependencies. A “win” for the OECD would mean success at containing China’s expansion. But a China “win” would pose a dilemma for OECD governments and MNCs. They may need to retreat from China containment or if the result is unacceptable, there may be an escalation in the global geopolitical conflict. There is a need to factor in the impact of India’s rise as a market and competitor on this global contest. Developments on the semiconductor battlefront will set the tone for conflicts across a broad front such as AI, biotechnology, EVs, medical supplies and rare earth.
References
Alicia García-Herrero, Pauline Weil (18 November 2022) “Lessons for Europe from China’s quest for semiconductor self-reliance”. Bruegel. https://www.bruegel.org/policy-brief/lessons-europe-chinas-quest-semiconductor-self-reliance
Alexandra Alper, Karen Freifeld (2 August 2022) “US considers crackdown on memory chip makers in China”. Reuters. https://www.theedgemarkets.com/article/us-considers-crackdown-memory-chip-makers-china-0
Anton Shilov (13 September 2022) “U.S.-proposed chip 4 alliance faces opposition from partners”. Toms Hardware. https://www.tomshardware.com/news/us-proposed-chip-4-alliance-faces-opposition-from-partners
Anton Shilov (12 September 2022) “US lawmakers concerned about Apple’s decision to use YMTC’s 3D NAND”. Toms Hardware. https://www.tomshardware.com/news/us-lawmakers-concerned-with-apple-decision-to-use-ymtc-3d-nand
Antonio Varas, Raj Varadarajan, Jimmy Goodrich, Falan Yinug (April 2021) “Strengthening the global semiconductor supply chain in an uncertain era”. Boston Consulting Group (BCG), Semiconductor Industry Association (SIA). https://www.semiconductors.org/wp-content/uploads/2021/05/BCG-x-SIA-Strengthening-the-Global-Semiconductor-Value-Chain-April-2021_1.pdf
Arjun Gargeyas (25 August 2022) “The Chip 4 Alliance might work on paper, but problems will persist”. The Diplomat. https://thediplomat.com/2022/08/the-chip4-alliance-might-work-on-paper-but-problems-will-persist/
Bureau of Industry and Security (BIS) (7 October 2022) “Commerce implements new export controls on advanced computing and semiconductor manufacturing items to the People’s Republic of China (PRC)”. www.bis.doc.gov/index.php/documents/about-bis/newsroom/press-releases/3158-2022-10-07-bis-press-release-advanced-computing-and-semiconductor-manufacturing-controls-final/file
Ben Thompson (25 October 2022) “Chips and China”. Stratechery. https://stratechery.com/
Chad P. Brown (2 November 2022) “National security, semiconductors, and the US move to cut off China”. Roundtable transcript. Petersen Institute for International Economics (PIIE). https://tradetalkspodcast.com/wp-content/uploads/2022/11/Episode-170-Transcript-Complete.pdf
Cheng Ting-Fang (11 October 2022) “China’s sea turtle tech executives stranded by U.S. crackdown”. Nikkei Asia. https://asia.nikkei.com/Business/Tech/Semiconductors/China-s-sea-turtle-tech-executives-stranded-by-U.S.-crackdown
Congressional Research Service (2021) “China’s new semiconductor policies: Issues for Congress”. https://crsreports.congress.gov/product/pdf/R/R46767
Dan Robinson (2 Nov 2022) “Uncle Sam wants allies to join its anti-China chip crusade”. The Register. https://www.theregister.com/2022/11/02/us_allies_china_chip_sanctions/?td=keepreading
Dan Robinson (4 November 2022) “US chip industry worried it may lose out to rivals over China ban”. The Register. https://www.theregister.com/2022/11/04/us_chip_industry_china_ban/?td=keepreading
Dan Robinson (7 November 2022) “US to Japan: We’ll help you make chips. Now about that China ban…”. The Register. https://www.theregister.com/2022/11/07/us_japan_2nm_chips/?td=keepreading
David P. Goldman (13 October 2022) “China chip ban a US exercise in extreme self-harm”. Asia Times. https://asiatimes.com/2022/10/china-chip-ban-a-us-exercise-in-extreme-self-harm/
David P. Goldman (5 November 2022) “US godfather makes a chip offer you can’t understand”. Asia Times. https://asiatimes.com/2022/11/us-godfather-makes-a-chip-offer-you-cant-understand/
Dong-Joon Kwon (29 September 2022) “US semiconductor customer not accepting Made in China products”. Korea IT News. https://english.etnews.com/20220929200001
Gary Clyde Hufbauer, Megan Hogan (October 2022) “CHIPS Act will spur US production but not foreclose China”. Peterson Institute for International Economics (PIIE). https://www.piie.com/publications/policy-briefs/chips-act-will-spur-us-production-not-foreclose-china
Gregory C. Allen (11 October 2022) “Choking off China’s access to the future of AI”. Center for Strategic & International Studies (CSIS). https://www.csis.org/analysis/choking-chinas-access-future-ai
Global Times (24 November 2022) “GT Voice: Asia must jointly tackle external risks amid looming US-EU subsidy war”. https://www.globaltimes.cn/page/202211/1280373.shtml
Hideki Uno (19 September 2022) “Japanese semiconductor industrial policymaking in the twenty-first century”. Center for Strategic and International Studies (CSIS). https://www.csis.org/blogs/perspectives-innovation/japanese-semiconductor-industrial-policymaking-twenty-first-century
Iain Morris (3 August 2022) “The West’s RAN supply chain and the goal of freedom from China”. Light Reading. https://www.lightreading.com/open-ran/the-wests-ran-supply-chain-and-goal-of-freedom-from-china/d/d-id/779446
Jean-Yves Larguier (9 September 2022) “China’s semiconductor industry: The promises of RISC-V open source architecture”. Institut Montaigne. https://www.institutmontaigne.org/en/analysis/chinas-semiconductor-industry-promises-risc-v-open-source-architecture
Jeff Pao (3 November 2022) “Chinese chip makers seek way out of US curbs”. Asia Times. https://asiatimes.com/2022/11/chinese-chip-makers-seek-way-out-of-us-curbs/
Jeff Pao (22 October 2022) “US mulling bans to stunt China’s quantum computing”. Asia Times. https://asiatimes.com/2022/10/us-mulling-bans-to-stunt-chinas-quantum-computing/
Jeff Pao (4 November 2022) “US pressing Japan to follow its China chip ban”. Asia Times. https://asiatimes.com/2022/11/us-pressing-japan-to-follow-its-china-chip-ban/
John Lee, Meia Nouwens, Kai Lin Tay (12 August 2022) “Strategic settings for 6G: Pathways for China and the US”. International Institute for Strategic Studies (IISS). https://www.iiss.org/blogs/research-paper/2022/08/strategic-settings-for-6g-pathways-for-china-and-the-us
Jon Bateman (6 May 2022) “U.S. sanctions on Hikvision would dangerously escalate China tech tensions.” Carnegie Endowment for International Peace. https://carnegieendowment.org/2022/05/06/u.s.-sanctions-on-hikvision-would-dangerously-escalate-china-tech-tensions-pub-87089
Jonathan Goldberg (19 September 2022) “US-China semiconductor battle: Second and third order consequences”. Techspot. https://www.techspot.com/news/96021-us-china-semiconductor-battle-second-third-order-consequences.html
Kati Suominen (22 November 2022) “Implications of the European Union’s digital regulations on U.S. and EU economic and strategic interests”. https://www.csis.org/analysis/implications-european-unions-digital-regulations-us-and-eu-economic-and-strategic-interests
Laura Dobberstein (8 November 2022) “Chipmakers cripple products to dodge US China ban”. The Register. https://www.theregister.com/2022/11/08/chipmakers_alter_products_china_ban/
Lin Biao (1965) “Long live the victory of people’s war!” Marxist.org. https://www.marxists.org/reference/archive/lin-biao/1965/09/peoples_war/ch05.htm
Mathieu Duchâtel (March 2022) “Semiconductors in Europe: The return of industrial policy”. Institut Montaigne. https://www.institutmontaigne.org/ressources/pdfs/publications/europe-new-geopolitics-technology-1.pdf
Mathieu Duchâtel (March 2022) “Technology transfers: the case for an EU-Japan-US cooperation framework”. Institut Montaigne. https://www.institutmontaigne.org/ressources/pdfs/publications/europe-new-geopolitics-technology2.pdf
Mathieu Duchâtel (26 October 2022) “An alternative supply chain in India? Taiwan’s experience”. Institut Montaigne. https://www.institutmontaigne.org/en/analysis/alternative-supply-chain-india-taiwans-experience
Matthew Reynolds (3 November 2022) “Assessing the new semiconductor export controls”. Center for Strategic & International Studies (CSIS). https://www.csis.org/analysis/assessing-new-semiconductor-export-controls
Nuying Huang, Jessie Shen (4 October 2022) “Chinese car vendors asking for China-made ICs”. Digitimes Asia. https://www.digitimes.com/news/a20221003PD208/automotive-ic-china-foundry.html?chid=13
Phuah Eng Chye (2015) Policy paradigms for the anorexic and financialised economy: Managing the transition to an information society.
http://www.amazon.com/dp/B01AWRAKJG
Phuah Eng Chye (5 June 2021) “Global reset – Two whales in a pond”.
http://economicsofinformationsociety.com/global-reset-two-whales-in-a-pond/
Phuah Eng Chye (12 March 2022) “Global reset – Technology decoupling (Part 1: Challenges, checkpoints, chokepoints and IOT)”.
Phuah Eng Chye (26 March 2022) “Global reset – Technology decoupling (Part 2: Decoupling race and scenarios)”. http://economicsofinformationsociety.com/global-reset-technology-decoupling-part-2-decoupling-race-and-scenarios/
Phuah Eng Chye (2 July 2022) “Theories on war and diplomacy (Part 3: Conflict in the information realm)”. http://economicsofinformationsociety.com/theories-on-war-and-diplomacy-part-3-conflict-in-the-information-realm/
Phuah Eng Chye (30 July 2022) “The Great Economic War (GEW) (Part 1: The beginning)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-1-the-beginning/
Phuah Eng Chye (13 August 2022) “The Great Economic War (GEW) (Part 2: Strategic concepts and implications)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-2-strategic-concepts-and-implications/
Phuah Eng Chye (27 August 2022) “The Great Economic War (GEW) (Part 3: The financial nuclear bomb)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-3-the-financial-nuclear-bomb/
Phuah Eng Chye (10 September 2022) “The Great Economic War (GEW) (Part 4: Battles reshaping the global monetary order)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-4-battles-reshaping-the-global-monetary-order/
Phuah Eng Chye (24 September 2022) “The Great Economic War (GEW) (Part 5: Global economic breakdown and monetary disorder)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-5-global-economic-breakdown-and-monetary-disorder/
Phuah Eng Chye (8 October 2022) “The Great Economic War (GEW) (Part 6: Geopolitics, monetary policy challenges and the collapsing tripod)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-6-geopolitics-monetary-policy-challenges-and-the-collapsing-tripod/
Phuah Eng Chye (22 October 2022) “The Great Economic War (GEW) (Part 7: Global depression and deglobalisation)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-7-global-depression-and-deglobalisation/
Phuah Eng Chye (5 November 2022) “The Great Economic War (GEW) (Part 8: Lawfare)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-8-lawfare/
Phuah Eng Chye (19 November 2022) “The Great Economic War (GEW) (Part 9: Geopoliticisation of MNCs)”. http://economicsofinformationsociety.com/the-great-economic-war-gew-part-9-geopoliticisation-of-mncs/
Qi Xijia (14 September 2022) “Shanghai confirms 14-nm chips being mass produced now”. Global Times. https://www.globaltimes.cn/page/202209/1275240.shtml
Reva Goujon, Lauren Dudley, Jan-Peter Kleinhans, Agatha Kratz (21 October 2022) “Freeze-in-place: The impact of US tech controls on China”. Rhodium Group.
Roslyn Layton, Jeff Ferry (June 2022) “Silicon sellout: How Apple’s partnership with Chinese military chip maker YMTC threatens American national security”. China Tech Threat and Coalition For A Prosperous America. https://chinatechthreat.com/wp-content/uploads/2022/06/Silicon-Sellout.-How-Apples-Partnership-With-Chinese-Military-Chip-Maker-YMTC-Threatens-National-Security.pdf
Sam Bresnick, Nathaniel Sher (16 September 2022) “New export controls on Chinese semiconductors may prove self-defeating”. China File. https://www.chinafile.com/reporting-opinion/viewpoint/new-export-controls-chinese-semiconductors-may-prove-self-defeating
Scott Foster (14 September 2022) “Gallium oxide fuelling new chapter chip war”. Asia Times. https://asiatimes.com/2022/09/gallium-oxide-fuelling-new-chapter-chip-war/
Scott Foster (19 September 2022) “S Korea seeks neutral ground in US-China chip war”. Asia Times. https://asiatimes.com/2022/09/s-korea-seeks-neutral-ground-in-us-china-chip-war/
Scott Foster (14 November 2022) “Japan’s grand plan to home grow 2nm chips”. Asia Times. https://asiatimes.com/2022/11/japans-grand-plan-to-home-grow-2nm-chips/
Semiconductor Industry Association (SIA) (2021) “Taking stock of China’s semiconductor industry”. White Paper. https://www.semiconductors.org/taking-stock-of-chinas-semiconductor-industry/
Semiconductor Industry Association (SIA) (10 January 2022) “China’s share of global chip sales now surpasses Taiwan’s, closing in on Europe’s and Japan’s”. Blog Post. https://www.semiconductors.org/chinas-share-of-global-chip-sales-now-surpasses-taiwan-closing-in-on-europe-and-japan/
South China Morning Post (SCMP) (23 April 2021) “Explainer | US-China tech war: Everything you need to know about the US-China tech war and its impact”. https://www.scmp.com/news/china/article/3198542/amid-chill-china-ties-india-and-taiwan-look-deeply-engage-electronics-chips-and-green-tech?module=perpetual_scroll_2&pgtype=article&campaign=3198542
Special Competitive Studies Project (Sept 12, 2022) “Mid-decade challenges to national competitiveness”. https://www.scsp.ai/wp-content/uploads/2022/09/SCSP-Mid-Decade-Challenges-to-National-Competitiveness.pdf
Sujai Shivakumar, Charles Wessner, Thomas Howell (14 November 2022) “A seismic shift: The new U.S. semiconductor export controls and the implications for U.S. firms, allies, and the innovation ecosystem”. Center for Strategic and International Studies (CSIS). https://www.csis.org/analysis/seismic-shift
U.S. Department of Commerce (6 September 2022) “A strategy for the chips for America Fund”. https://www.nist.gov/system/files/documents/2022/09/07/CHIPS-for-America-Strategy.pdf
Vishnu Kannan, Jacob Feldgoise (22 November 2022) “After the CHIPS Act: The limits of reshoring and next steps for U.S. semiconductor policy”. Carnegie Endowment. https://carnegieendowment.org/2022/11/22/after-chips-act-limits-of-reshoring-and-next-steps-for-u.s.-semiconductor-policy-pub-88439
Xie Jun, Qi Xijia, Liu Yang (27 July 2022) “US pushes chip bill to encircle China, but unable to lure firms to decouple with mainland”. Global Times. https://www.globaltimes.cn/page/202207/1271588.shtml
[1] See U.S. Department of Commerce and Congressional Research Service reports; See Antonio Varas, Raj Varadarajan, Jimmy Goodrich and Falan Yinug for industry background.
[2] Alicia García-Herrero and Pauline Weil provide an overview of China’s semiconductor policies.
[3] See Chad Bown.
[4] See Bureau of Industry and Security (BIS).
[5] See Jeff Pao.
[6] See Vishnu Kannan and Jacob Feldgoise for more details on the CHIPs Act.
[7] See Kati Suominen on the “Implications of the European Union’s digital regulations on U.S. and EU economic and strategic interests”.
[8] See Mathieu Duchâtel on Taiwan’s experience in India.
[9] Mao Tse-tung enumeration of the basic tactics of guerrilla warfare. See Lin Biao.
[10] See John Lee, Meia Nouwens and Kai Lin Tay on strategic issues related to 6G.