As China-U.S. relations appear to deteriorate, we asked Ernst how his views on China have changed in recent years.
BARCELONA — Over the last few months, eyes of the media, the voices of politicians and the minds of the business and tech communities have been trained all on Huawei and, more broadly, on China.
The world is also watching closely the current U.S.-China trade talks and awaiting a ruling on March 1, to see if Canada grants the U.S. request to extradite Huawei CFO Meng Wanzhou.
Undoubtedly, the outcome of these political matters will be important. But being reactive to any immediate news item offers scant insight on big picture and the long arc of the future.
EE Times turned to Dieter Ernst, a long-time observer of China, economist and senior fellow at the Centre for International Governance Innovation (Waterloo, Canada) and the East-West Center (Honolulu). Ernst is known for research examining industrial and innovation policies in China, the U.S. and emerging economies, with a focus on standards and intellectual property rights.
As China-U.S. relations appear to deteriorate, we asked Ernst how his views on China have changed in recent years.
We also wanted to know if he, too, believes the world is pushing China into a corner, and if so, what might be the consequences. Ultimately, we asked how badly the global electronics industry might be damaged if we lose the world’s third largest buyer — Huawei — of semiconductors.
Your view on China in 2011
EE Times: Back in 2011, you testified in Washington, in front of the U.S.-China Economic and Security Review Commission Hearing. You said then that China’s innovation is not a threat to the U.S. leadership in science and technology. “Instead, China’s progress in innovation should be seen as a wake-up call for America.” Eight years later, what has changed in the state of the Chinese technology progress?
Dieter Ernst: Let me first explain what I meant by “wake-up call for America.”
In my June 15, 2011 testimony I stated unequivocally that “China’s innovation policy is not a threat to U.S. leadership in science and technology. …the US retains a strong lead in overall innovative capacity, and China still has a long way to go to close the innovation gap. …Rather than fearing China and blaming it for our problems, …both the US government and the private sector need to join forces to develop and implement:
In the discussion, I added
Don’t demonize or underestimate or overestimate China. It makes more sense for the U.S. to think about what we should do to strengthen our innovation system rather than expanding lots of energy trying to change Chinese policy.
So, you ask where does China stand today?
Since 2011, China has moved ahead in some key areas of the IT industry, but it also faces fundamental challenges to its manufacturing export-oriented development model.
After decades of rapid-fire growth, China has reached a level of development where catching up through an investment-driven “global factory” model based on low-wage production is no longer sufficient to create long-term economic growth and prosperity.
Serious constraints on environmental, human and financial resources imply that economic growth based on scale expansion is running out of steam, depressing China’s economic growth.
At the same time, international trade, a primary source of China’s rise, has fallen to its lowest level since 2009, and keeps languishing under the pressure of an intensifying trade war with the US. In addition, a declining labor force, rising wages, and skill bottlenecks are eroding China’s international competitiveness.
To break out of this growth impasse, China’s leadership has implemented three massively funded policy initiatives:
Western analysts typically see these plans as a Chinese ploy for world domination. American foreign policy, defense and corporate elites believe that any increase in China’s AI capacity poses a threat to U.S. leadership. Our research suggests that the danger of a Chinese threat may well be exaggerated.
Despite decades of effort to develop a robust domestic semiconductor industry, China remains weak in the design and fabrication of leading-edge memory and processors. This weakness is particularly grave for fabrication, where SMIC and other Chinese players continue to lag years behind in leading-edge process nodes.
Of particular concern is the persistent gap between semiconductor consumption and production. China has been the largest market for semiconductors since 2005. Yet only slightly more than 15% of China’s total semiconductor consumption is supplied by China-based production in 2018. And foreign companies with fabs in China may account for almost half of that domestic fab capacity.
While the U.S. semiconductor industry has consistently retained nearly half of the global market, China-based production has only around 5%. Other well-documented weaknesses are in leading-edge multi-core processors and memory devices, and China’s embryonic stage of development in semiconductor equipment and design tool services.
Noteworthy achievements of China’s semiconductor industry include optical devices (especially LED), low-power embedded processors, sensors, and discrete devices, with China now approaching self-sufficiency. Equally important is the surge of China’s semiconductor assembly, packaging, and testing (APT) industry, which has moved ahead of Taiwan and Japan.
China’s leading OEMs
China's leading OEMs are among the top 10 chip buyers, significantly increasing their bargaining power relative to U.S. semiconductor firms. Increasing market power is as much important as technological capacity in shaping China’s position in the global semiconductor industry.
Since 2011, China’s leading OEMs (Huawei, Lenovo, BBK Electronics and Xiaomi) have significantly improved their position as chip buyers. In 2018, these four leading Chinese OEMs are in the top ten chip buyers. Huawei in fact increased its chip purchases by 45%, moving ahead of Dell into the third spot. This will make it harder for US chip vendors to maintain high margins.
In addition, Huawei’s telecom equipment revenue during 2018 has become nearly as large as Nokia and Ericsson combined. This again indicates that Huawei will continue to play an important role as a leading chip buyer.
What’s happening now in AI chips?
As for AI chips, our research finds that China is still way behind in these rapidly evolving markets where CPUs, GPUs, FPGAs intensely compete with new special-purpose AI chips for algorithms used in deep learning neural networks. US firms are ahead in all of these fields. For China, entry barriers remain very high, in terms of access to experienced talent and intangible knowledge both for IC design and fabrication. In addition, indirect forms of knowledge sourcing through M&A and technology licensing are increasingly restricted.
Huawei’s subsidiary HiSilicon in fact is the only company with sufficient design and engineering talent to compete with the large international firms. However, Huawei itself remains heavily dependent on core ICs from U.S. and other foreign industry leaders. According to industry sources, Huawei is still very reliant on foreign suppliers for CPUs, GPUs, FPGAs, and high-end memory.
Western observers typically argue that China’s best bet would be to continue using off-the-shelf AI chips from global semiconductor industry leaders, and then focus more on the application of AI. In this view, the immense challenges of developing an integrated AI chip value chain would make it impossible to catch up with industry leaders at reasonable cost. This argument assumes largely free and open markets for leading-edge AI chips — an extraordinarily brave but misleading assumption!
Critical voices within China however suggest that “slightly-behind-the-leading-edge” chip architectures and process nodes would be good enough for diffusing AI technologies across China’s manufacturing and service industries. At the same time, there are expectations that big changes in mainstream chip architectures may open up new opportunities for Chinese firms to leap-frog into at least some niches of the AI chip market, especially if this is done through strategic partnerships with leading U.S. and other foreign semiconductor companies.
It will be interesting to see whether the U.S. government will be able to effectively block U.S. companies like Intel, Qualcomm, Xilinx and Nvidia from continuing to cooperate with Chinese companies. After all, these U.S. companies critically depend on China, the largest worldwide semiconductor market. In addition, will the visible hand of the U.S. Commerce Department be strong enough to stop Arm-Softbank, TSMC or HonHai/Sharp from continuing highly profitable links with Chinese companies?
China bashing begins
EE Times: Obviously, the U.S. has woken up. Do you think we’ve pushed China too far into a corner? If yes, what are the expected outcome? What options are we leaving Huawei at this point? Give us your worst-case scenario.
Ernst: The current “China bashing” mood in Washington reflects the rise of Techno-Nationalism in the U.S. A consensus is taking shape among U.S. defense, foreign policy and economic policy-making elites that China’s rise in information technology poses a serious threat to America’s leadership in science and advanced technology. Hence, trade and investment restrictions are needed to contain China’s technological and geopolitical ambitions. This consensus embraces the political spectrum. Both Republicans and Democrats share it, and it will persist even once Trump is gone.
A negotiated settlement cannot be assumed. Trade diplomacy is now giving way to a “Regime Change” agenda. According to an influential think tank, China must converge to U.S. “advanced economic norms”, or else “the United States will take a series of self-protective steps far more hawkish than transitional safeguards, including restrictions on visas, educational and professional exchanges, technology sharing, and cooperation arrangements.”
These words are now followed by action. In fact, the U.S. government is drastically tightening the export control of a broad portfolio of information technologies, with a focus on China. Since August 2018, the Export Control Reform Act (ECRA) requires the Commerce Department to create lists of “emerging” and “foundational” technologies that are essential to US national security.
In October 2018, U.S. Treasury released a pilot program as part of the Foreign Investment Risk Review Modernization Act of 2018 (FIRRMA) that drastically expands the mandate of the Committee on Foreign Investment in the U,S. (CFIUS).
Of particular importance are restrictions on so-called “deemed exports” which limits any information about a controlled technology to a foreign national. Once these restrictions are implemented, they will throw a wrench into the existing pattern of knowledge sharing that is the lifeblood of the global IT industry. This unprecedented arsenal of technology export restrictions is bound to constrain China’s options to use off-the-shelf chips from leading U.S. and other foreign vendors.
Huawei is at the center of this conflict. A bipartisan group of US Senators introduced a bill in January to require the President to deny U.S. technology to Huawei. The campaign against Huawei was intensified by the large U.S. government delegation from the FCC, the state department, and the intelligence Services sent to the recent Mobile World Congress in Barcelona. This indicates that the goal really is to weaken China’s most successful company, if not to push it out of business.
It is worthwhile, however, considering some of the hurdles that the US government may face in this effort. This campaign may end up hurting the U.S. IT industry, and its broader economy, as well as key partners of the U.S. Supply chains may be disrupted and jobs destroyed on a significant scale in China, but also in the US, and customers would have less choice.
Here is why pushing Huawei into a corner may not be that easy …
Size matters, as do well-established relationships. With a 2019 sales target of $12 billion, Huawei would generate more revenue than Boeing. And Huawei has established a dense international network with key telecom operators, and with suppliers across the semiconductor value chains for telecom equipment, smartphones, and servers.
Huawei’s revenue share gains since 2015 have been most pronounced in key telecom equipment markets — core, router and optical transport. There are no significant U.S. companies which could fill the vacuum left by Huawei’s absence. And it would take quite a while for Nokia and Ericsson (a company in deep trouble) to take over. And telecom operators would not be happy.
Economies of scale mean that Huawei can make equipment for 5G base stations for 20% to 30% less than its competitors.
China’s TDD standard has advantages, especially for 5G. China has focused on developing TDD technology as a matter of national policy since the 3G era, while U.S. and European players have generally opted for FDD. The two perform much the same when it comes to 4G, but TDD is expected to be the main choice for 5G, as FDD cannot manage the necessary transfer speeds.
Huawei’s extensive experience in standard-setting and its significant investments in the critically important standard-essential patents (SEPs) for 5G technologies has enabled Huawei to sign cross-licensing agreements with all major IPR holders in the wireless industry, including Ericsson, Nokia-Siemens, Alcatel-Lucent, Qualcomm, Nokia, Sony-Ericsson, Sisvel and other leading players. Given its huge portfolio for 5G relevant SEPs, it won’t be easy to push Huawei to the sidelines, not even for the powerful U.S. government.
Huawei’s multiple defense positions
It would seem to be unrealistic to expect that a company with a street-fighter mentality like Huawei will simply sit and wait until the U.S. Commerce Department will cut off its supplies of leading-edge semiconductors from U.S. industry leaders.
In fact, Huawei appears to have prepared multiple defense positions:
a. In its quest to become the world’s largest smartphone vendor, Huawei has asked its Taiwan supply chain partners (IC packaging, testing, chip probing, and optical components) to relocate their operations to China, although Taiwan’s regulatory restrictions may make this quite difficult.
b. Huawei is one of the three largest foundry service customers of TSMC. It wouldn’t be easy for TSMC to lose them.
c. As for Huawei’s recently announced Arm-based CPU Kunpeng 920, this is important as it seems to indicate a fairly close relationship with Softbank’s Arm. Hi-Silicon has designed the CPU for Kunpeng 920, but that CPU appears to be slower than the Arm Cortex-A76 or Marvell ThunderX2. But it seems that for the China server market that performance is considered to be sufficient.
More important for Huawei’s survival may be that Arm is working with Huawei and several others to build a server ecosystem around Arm under the Neoverse brand. Huawei can be an effective representative of this effort in China.
Even more than the ZTE shock before, the campaign against Huawei will likely force China to accelerate its drive for self-reliance. By strengthening the hardliners in China’s leadership, U.S. restrictions on technology exports may also strengthen the current drive towards re-centralization of political control and the spread of mass surveillance. In fact, in January President XI announced a new 15-year Science and Technology Innovation Plan to speed up its indigenous innovation campaign.
As for the semiconductor industry, China continues to advance its foundry industry with huge investments in new fabs and technology, despite trade tensions and a slowdown in the IC market.
China has the most fab projects in the world, with 30 new facilities or lines in construction or on the drawing board, according to data from SEMI’s World Fab Forecast Report. Of those, 13 fabs are targeted for the foundry market, according to SEMI. The remaining facilities are geared toward LEDs, memory and other technologies.
In the end, technology export restrictions imposed by the U.S. government on Huawei and other Chinese companies may disrupt China’s access to key technologies, at least for some time, and it may disentangle the close economic interdependence between the US and Chinese technology sectors.
But technology export restrictions are a crude and often self-destructive policy instrument. Collateral damage is guaranteed, not only to U.S. industry but also to the IT industry in America’s partner countries. Most importantly, such restrictions on the free flow of knowledge, technology and brains will stifle innovation in an industry that is unrivaled in its degree of globalization.
In the age of economic nationalism…
EE Times: In the same 2011 testimony, you talked about “proactive and smart trade diplomacy.” Obviously, under the current administration, nobody appears talking about being “proactive” or “smart diplomacy.” If you had to suggest, what would be the 2019 version of your “proactive and smart trade diplomacy recommendations” to the U.S. government?
Ernst: In 2011, I was reasonably optimistic that trade and investment conflicts with China could be gradually reduced if U.S. government and the private sector would join forces in reforming key aspects of U.S. trade policy on China.
“Asymmetric Interdependence” implies that China needs the U.S. even more, both as a market and as a source of technology. As China continues to lag behind the U.S. in innovation capacity, the U.S. can still play an important role in shaping the scope and the speed of this cooperation process. The paper emphasized that implementing such cooperation between countries at different stages of development would only work if both countries accept that their economic and political systems are different.
I no longer share this optimism.
There is no doubt that both the U.S. and China have squandered this unique opportunity. As the rise of economic nationalism in the U.S. interacts with the recentralization of state control in China, it is now much more difficult to identify and mobilize stakeholders in both countries who would be willing to compromise and to find areas for selective cooperation.
Today’s global economy is not promising. Intensifying trade, investment and technology conflicts are likely to dominate U.S.-China economic relations for quite some time. It will take many years simply to repair the damage.
But more fundamental forces are at work. In both countries, ideology shapes industrial and trade policies. In the U.S., industrial policy remains a taboo, denying the important role played by the defense department and especially DARPA in creating America’s IT industry. Obama’s Advanced Manufacturing Partnership (AMP) program remained half-hearted and did not provide the big push in education, basic research and innovation infrastructure necessary to upgrade the US innovation system.
By contrast, China’s leadership is eager to use all the tools of industrial, trade and competition policy to co-shape international standardization, and to catch up and forge ahead in advanced manufacturing and services. While U.S. analysts typically see these policies as a ploy for world domination, in China they are viewed as unavoidable if the country wants to move beyond the outdated “Global Factory” model based on low-wage mass production. In essence, moving up the value chain through innovation is China’s response to its slowing economy and the increasingly severe economic, social and environmental costs of its outdated development model.
In addition, fundamental changes in the nature of global competition further increase the tensions between these two countries. There is a fundamental shift in the dynamics of global competition. Until recently, the main rivalry was in manufactures trade.
Today, the contest is for dominance of the data-driven economy and AI. The current AI and Big Data boom has deepened U.S.-China rivalry.
According to my colleague Dan Ciuriak, Canadian economist, “the emerging data-driven economy features large economies of scale and scope, often accompanied by network externalities, and pervasive information asymmetry. These characteristics tend to result in ‘winner take most’ economics, with the prize to the winner being the capture of international rents; these rents promise to be very large and thus serve as an inducement for strategic trade and investment policy.”
Both in the U.S. and China, massive sums have been invested in development of technology in the digital space to capture these rents. It is this new data-centered competition that explains the proliferation of U.S. technology export restrictions. At present, there is little that can stop this contest about data rents.
In the meantime, it is up to the global IT communities — both in companies and labs — to search for ways to move things forward on their own in places like the World Semiconductor Council (WSC), standard development organizations, important conferences, and technology and business publications. No doubt such efforts will remain constrained for quite some time by the powerful headwinds of mutual distrust.
How badly will the electronics industry be damaged?
EE Times: How badly do you believe that the global electronics industry will be ultimately damaged by losing the world’s third largest buyer — Huawei — of electronics products?
Ernst: Much depends on how extensive and radical the restrictions codified in ECRA, FIRRMA and CFIUS will be in practice. I hope that SEMI, SIA or WSC will commission research on this important question.
I hear that losses are already significant in companies with heavy exposure to China, due to uncertainty and preventive adjustments in contracts and location decisions. Specifically for Huawei, the following U.S. lead suppliers will badly suffer, i.e. Flex, Broadcom, Qualcomm, Seagate, Micron, Qorvo, Intel, and of course the suppliers of design tools and production equipment.
Of particular interest will be how non-U.S. companies like ASML and TSMC will be affected. ASML already stopped serving Fujian Jinhua, after the Commerce Department denied US technology to Fujian Jinhua.
But the nuclear option really would be if TSMC could no longer provide leading-edge node process technology foundry services to Hi-Silicon and other aspiring Chinese IC design companies.
It is unclear at this stage whether the WTO dispute resolution system could be called upon to limit such extraterritorial US technology export restrictions. Nor do we know at this stage of any pending litigation or other legal battles in response to.
My gut feeling is that, given the considerable lobbying power of the US semiconductor industry, US tech export restrictions may be implemented incrementally, rather than in a massively destructive big push.
 On MIC 2025, see Ernst, D.,2018, “Advanced Manufacturing and China’s Future for Jobs”, chapter in, Eva Paus, editor, Confronting Dystopia. The New Technological Revolution and the Future of Work, Cornell University Press Paperback, Ithaca and London. On AIDP, see Ernst, D., forthcoming, China’s Challenges in Artificial Intelligence – Data Governance and Semiconductors, Centre for International Governance Innovation/CIGI, Waterloo, Canada. See also the China AI Development Report 2018, published by our research partner, the China Institute for Science and Technology Policy at Tsinghua University/CISTP,
 Ernst, D., 2016, China’s Bold Strategy for Semiconductors—Zero-Sum Game or Catalyst for Cooperation?, East-West Center Working Papers: Innovation and Economic Growth Series, no. 9, September, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2836331; and Ernst, D., 2015, From Catching Up to Forging Ahead: China’s Policies for Semiconductors, East-West Center Special Study, September, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2744974 .
3] Rosen, D.H. and S. Kennedy, 2019, Building a Better Deal with China, Center for Strategic and International Studies /CSIS, Washington, D.C.
5] Ernst, D., 2017, China’s Standard-Essential Patents Challenge From Latecomer to (Almost) Equal Player?, A CIGI Special Report, The Centre for International Governance Innovation/CIGI, Waterloo, Canada
 Ciuriak, D., 2019, “A Trade War Fueled by Technology”, CIGI, Trade and Finance, January 11: page 4.