面对美国的科技脱钩,中国必须建立集成电路的产业基础
Return to the Library

Faced with Technology Decoupling by the United States, China Must Establish an Industrial Base for Integrated Circuits

面对美国的科技脱钩,中国必须建立集成电路的产业基础

Lu Feng, a Peking University professor, argues a closed-loop domestic integrated circuit (IC) supply chain is urgently needed in the face of U.S. and allied technology controls. He suggests Beijing advance this goal by encouraging Chinese enterprises in the field to buy from and sell to each other – decisions that, Lu argues, will be made easier by U.S. technology controls. Lu also suggests China play to its strengths and use its expansive market as a source of leverage to influence the scope of such controls.


FacebookTwitterLinkedInEmailPrintCopy Link
Original text
PDF
English text
PDF
See an error? Drop us a line at
View the translated and original text side-by-side

In 2023, the United States is continuing to escalate its technology war against China. In the face of the United States’ aggressive technological decoupling, as well as the ensuing “de-Chinaization” of the United States and its allies, China should play to its strength as the world’s number one demand market, strike back, and pursue cooperation in the midst of struggle. At the same time, we should concentrate resources as soon as possible on promoting the formation of the local integrated circuit (IC) production chain, fully leverage the advantages of our national system to create an independent manufacturing base, and change the passive situation in which our leading science and technology-based enterprises are repeatedly embargoed.

进入2023年,美国对华科技战愈演愈烈。面对美国咄咄逼人的技术脱钩,以及随之而来的美国及其盟友的“去中国化”,中国要发挥在全球第一需求市场的优势,予以反击,在斗争中求合作;与此同时,要尽快集中资源促进集成电路本土产业链的形成;发挥举国体制的优势,打造自主制造基地,改变科技龙头企业屡遭封锁的被动局面。

Advantage of being the world’s No. 1 chip market

全球第一芯片市场的优势

Early in 2018, the Trump administration launched a trade war and a technology war against China. After Biden took office, the momentum of U.S. containment of China not only did not ease, but intensified. The United States has put 36 Chinese chipmakers on its ban list and has encouraged TSMC to relocate to the United States The U.S. bans will undoubtedly have a negative impact on China’s technological and economic development.

从2018年开始,特朗普政府对中国发动了贸易战和科技战。拜登上台后,美国遏制中国的势头不仅没有缓和,反而变本加厉。美国已将36个中国芯片制造商列入禁令名单,并鼓动台积电迁美。美国的禁令,毫无疑问将对中国的技术和经济发展造成影响。

Semiconductor technology originated in the United States, and the United States has a strong technological advantage. Viewed from the supply side of the global market, U.S. semiconductor companies account for 46.3% of the overall semiconductor market share (2021 Global Semiconductor Industry Association (SIA) data) From the demand side, according to SIA data, global semiconductor sales in 2021 were $555.9 billion, a record high. Sales in the China market were $192.5 billion, accounting for 34.6%, a year-on-year increase of 27.1%, and it remains the world’s largest consumer market for semiconductors. This is a huge advantage for China.

半导体技术起源于美国,美国具有强大的技术优势。从全球市场供给来看,美国半导体公司占据了整个半导体市场份额的46.3%(2021全球半导体行业协会SIA数据)。从需求方面看,据SIA数据显示,2021年全球半导体销售额为5559亿美元,创历史新高;中国市场的销售额为1925亿美元,占比34.6%,同比增长27.1%,仍是全球最大的半导体消费市场。这也是中国一个巨大的优势。

Michael Porter’s Five Forces model is commonly used in management science to analyze the five internal and external determining factors of enterprises. One of these forces is the power of buyers. That is to say, the bargaining power of the demand side of the market can affect the profitability of an enterprise. In the case of the semiconductor industry, the industry pattern in which the United States dominates the upstream (supply) and China dominates the downstream (demand) reflects the relationship between the U.S. and Chinese industries: the United States is starting to decline in many downstream industries, including even semiconductor manufacturing, while China’s downstream industry is gradually developing, its upstream industry is still a weakness. This has formed a unique “nuclear terror” equilibrium in the semiconductor industry, with the United States and China each having their own “nuclear weapons.”

管理学常用迈克尔•波特的五力理论模型,分析企业内外部五种决定性因素。其中的“一力”是购买者的力量,也就是说市场需求方的议价能力可以影响企业的盈利。就半导体产业来说,美国主导上游(供应)、中国主导下游(需求)的产业格局,体现了中美工业此消彼长的关系:美国很多下游工业开始衰落,甚至包括半导体的制造,而中国的下游工业逐渐发展起来,但是上游产业仍是短板。这就形成了半导体产业特有的“核恐怖”平衡状态,中美各自拥有自己的“核武器”。

China is the world’s largest chip market. If China were to not import U.S. chips at all, the U.S. semiconductor industry would be hit hard. Over the past five years, the United States’ suppression of China’s S&T enterprises has been a case of throwing stones in a glass house. They know that once China becomes completely throttled, their own enterprises will be throttled at the same time. For example, U.S. companies, including Qualcomm and Nvidia, rely on the Chinese market for more than 70% of their operating income. If they lost the Chinese market, these companies would reduce investment and lay off staff, and their stock prices would plummet, which in turn would lead to a panic on Wall Street and other chain reactions. In this situation, the United States continues to exert a stranglehold on China in some key technological areas on the one hand, carrying out focused attacks on Chinese enterprises (such as Huawei) that challenge U.S. technological hegemony; on the other hand, it continues to expand sales of its products in the Chinese market. At the same time, the United States sought to bring in a TSMC plant, first to revitalize the U.S. IC manufacturing industry, and second because the U.S. strategic community is worried that the U.S. IC production chain will be broken after China recovers Taiwan. Therefore, this is an important move of the United States in using re-industrialization against China.

中国是全球最大的芯片市场,假如中国完全不进口美国的芯片,美国半导体工业也会遭到重创。5年来,美国打压中国科技企业一直都是投鼠忌器,他们知道一旦彻底把中国卡死,同时也会把自己的企业卡死。比如美国的高通、英伟达等公司70%以上的营业收入依靠中国市场。失去中国市场,这些企业就会减少投资、裁员、股价大跌,进而导致华尔街恐慌等连锁反应。在这样的情况下,美国一方面在一些关键技术领域继续卡中国脖子,对中国对美国技术霸权构成挑战的企业进行重点打击(比如华为);另一方面继续在中国市场上拓展其产品销售。与此同时,美国把台积电工厂拉拢过去,一是想重振美国的集成电路制造业,二是美国战略界人士担心中国收复台湾之后,美国的集成电路产业链就会断掉。因此,这是美国以再工业化对付中国的一个重要举动。

Why hasn’t China formed a local IC production chain?

中国为什么没有形成集成电路的本土产业链

Although China faces weakness in the field of integrated circuits, there are Chinese enterprises in every segment of the global semiconductor production chain, and they are actually very strong in some fields, such as chip design. The key reason for being in a U.S. “stranglehold” is that the supply and demand cycle of the production chain has not been formed domestically.

虽然中国在集成电路领域面临短板,但在全球半导体产业链上的每个环节都有中国企业,有的领域还很强,比如芯片设计领域。被美国“卡脖子”的关键原因,在于国内还没有形成产业链的国内供需循环。

Why hasn’t it developed? It is not that we have not tried to do it. Rather, we gave up halfway several times, lacking the tenacity to stick with it to the end. Although China’s IC industry is still backward relative to the United States, it has a feature rarely seen globally, namely, there are Chinese enterprises in almost every segment of the semiconductor production chain. Thus far, however, the Chinese enterprises in the different semiconductor production chain segments have failed to form relatively strong supply and demand relationships with each other. Instead, each has formed supply and demand relationships with foreign enterprises, causing the fundamental problem of China’s IC industry. For example, ten years ago, because domestic chip design companies did not consider SMIC advanced enough, most of the orders it received came from overseas (though this has now improved). The chips designed by HiSilicon are ordered from TSMC. Until recently, it imported all of its equipment and did not use domestic equipment. Advanced Micro-Fabrication Equipment (AMEC) produced China’s first chip etching machines, but mainland companies did not buy them, and industrialization was not achieved until after they were sold to Taiwanese enterprises. It was only after China’s chip manufacturing enterprises were sanctioned by the U.S. that companies started to order domestically and procure domestically produced equipment and materials. Domestically produced equipment and materials are not advanced enough in terms of quality and performance, and are in a state of marginal substitution. Another aspect of why companies like SMIC do not dare to take orders from Huawei and HiSilicon is that they are afraid of coming under U.S. sanctions on technology and equipment.

为什么没有发展起来?并不是我们没有去做,而是数次半途而废,缺乏坚持到底的韧性。与美国相比,虽然中国集成电路工业还处于落后位置,但它具有一个全球罕见的特点,就是在半导体产业链几乎每一个环节都有中国企业存在。不过迄今为止,在半导体产业链各个环节的中国企业,彼此之间并没有形成比较强的供需关系,而是各自分别与国外企业形成供需关系,这就造成了我国集成电路工业的根本性问题。例如,10年前,中芯国际接的订单大部分都来自海外(现在已有所改善),因为国内芯片设计企业认为它不够先进。海思设计的芯片是在台积电下订单,直到不久前,它的设备还全部进口,不用国产设备。上海中微半导体设备生产出中国第一批芯片刻蚀机,但大陆企业不买,卖给台湾企业后才实现产业化。直到中国芯片制造企业遭到美国制裁封锁,才开始在国内下单,采购国产设备、材料。目前国产设备、材料的质量性能还不够先进,处于边际性替代的状态。还有一方面原因,如中芯国际不敢接华为、海思的订单,就是怕受到美国在技术和设备方面的制裁。

Why has China’s semiconductor industry not formed a local production chain? The root cause lies in our model of following others. China began to develop the semiconductor industry in the 1950s, and made integrated circuits in the 1960s. It always insisted on independent R&D in technology, and products and technologies were constantly updated. Under the planning system of the time, products from China’s semiconductor industry were mainly used in military industries and scientific research, and were not combined with commercial applications, so development was severely constrained.

中国半导体工业为什么没有形成本土的产业链?其根源在于我们的跟随模式。20世纪50年代,我国就开始发展半导体工业,60年代做出了集成电路,在技术上一直坚持自主研发,产品和技术不断更新。在当时计划体制下,我国半导体工业的产品主要用于军工和科研,没有与商业化应用结合,发展受到很大制约。

In the early 1980s, amid the wave of opening up to the outside, the domestic IC industry was overwhelmed by imports. Semiconductor enterprises that were mainly supporting military industries could not adapt to the mass market, and so they passively relied on imported equipment and technology. Research and development of IC technology were almost abandoned domestically.

20世纪80年代初,在开放浪潮中,国内集成电路产业被进口品冲垮。主要为军工配套的半导体企业不能适应大众市场,被动依赖引进设备和技术。国内几乎放弃了对集成电路技术的研发。

At that time, the central government was not unaware of the importance of developing integrated circuits, and had organized efforts to introduce several more advanced production lines through the joint venture approach, such as the 908 and 909 projects and the Shougang [-NEC] semiconductor project. However, reliance solely on introduced production lines could not enable Chinese enterprises to develop their own capabilities, and they were unable to keep up with the rapid changes in technology and markets. After accepting the lessons learned, Project 909 turned to independent development and evolved into today’s Huahong.

当时,中央并非没有意识到发展集成电路的重要性,也曾组织力量,以合资方式引进了数条较先进的生产线,如908、909工程和首钢的半导体项目。但是,仅靠引进生产线,不能使中国企业发展出自己的能力,也跟不上技术和市场的快速变化。接受了经验教训后,909工程转向自主开发,演变成为今天的华虹。

China entered a third phase of semiconductor industry development around 2000, marked by the establishment of SMIC in Shanghai. The characteristics of this round can be summarized as China’s enterprises adopting an internationalized development approach and participating in the international production chain cycle. This coincided with China’s economy entering a high-growth phase, and expanding market demand and investment capacity enabled the semiconductor industry to achieve a great deal of development.

2000年前后,以中芯国际在上海成立为标志,中国进入发展半导体工业的第三个阶段。这一轮的特点可以概括为,我国企业采用国际化的发展方式,参与国际产业链的循环。此时恰逢中国经济进入一个高增长阶段,不断扩大的市场需求和投资能力,使半导体工业获得很大的发展。

Since China has a long history of developing the semiconductor industry, there are Chinese enterprises in almost every segment of the semiconductor production chain. However, the internationalized development approach has not led to the formation of relatively strong supply and demand links between these Chinese enterprises distributed in the upstream and downstream, resulting in a situation where China’s leading chip enterprises can be easily held back by the United States.

由于中国发展半导体工业的历史很长,半导体产业链上的每个环节几乎都有中国企业。但是,国际化发展方式没有使这些分布在上游、下游和中游的中国企业之间形成比较强的供给和需求联系,造成了中国龙头芯片企业能够轻易被美国扼制的局面。

Forming an industrial base is the primary goal and task

形成产业基础是首要目标和任务

The key to dealing with the U.S. technology blockade is to form a production chain capable of domestic self-circulation.

应对美国技术封锁的关键,是要形成能够国内自我循环的产业链。

Under U.S. suppression, the technological progress of every Chinese IC enterprise must rely on the technological progress of the entire domestic production chain. If Chinese enterprises in the various segments cannot form strong supply and demand links with each other, then the technological progress of each enterprise will be subject to U.S. suppression. Therefore, connecting and opening up the domestic IC production chain is equivalent to establishing the industrial base of this industry. Once the local production chain is formed, we will not need to fear U.S. technological embargoes, because China’s market is large enough.

在美国的打压下,我国每一个集成电路企业的技术进步,都必须依托于国内整个产业链的技术进步。如果各个环节的中国企业彼此之间不能形成较强的供应和需求联系,那么每个企业的技术进步都会受制于美国的压制。因此,打通本国集成电路的产业链,就等同于建立起这个工业的产业基础。一旦形成本土的产业链,我们就不怕美国的技术封锁,因为中国的市场足够大。

Summarizing the past lessons learned and accelerating the formation of an industrial base should become the primary goal and task of China’s development of the IC industry.

总结以往的经验教训,加快形成产业基础,应该成为中国发展集成电路工业的首要目标和任务。

Since the release in 2000 of State Council Document No. 18 encouraging the development of the software and IC industries, the State Council has issued documents every few years to support the development of integrated circuits, but these document’s objectives and content do not include the development of an independent industrial base. Major special projects are supported and financed by the state, with technical indicators as the criteria for project establishment, but those indicators take internationally advanced technology as the reference system. The projects under the “following others” system support individual technologies and take the existing foreign technologies as the targets to follow. They only follow the frontier of international technological progress, and most of them are undertaken by universities or research institutions.

2000年,鼓励软件产业和集成电路产业发展的国务院18号文件发布以来,每隔几年,国务院就会发布一个支持集成电路发展的文件,但其内容未有发展自主产业基础的目标和内容。重大专项由国家支持、资助,以技术指标作为项目立项的标准,而这个技术指标是以国际上已有的先进技术为参照系的。“跟随体制”下的项目是支持单项技术,并以国外已有的技术为跟随目标,仅仅是跟随国际技术进步的前沿,而且大多由大学或研究机构承担。

China’s major special projects have been implemented for three five-year plans. We have seen the achievements of China’s new energy vehicles, which originated in the independent innovation movement of China’s automobile industry more than 20 years ago, thereby providing a basis for the national program. At the time, the relevant national authorities formed the concept of using new energy technology to promote the implementation of “overtaking around the turns,” but only enterprises with self-developed products would be able to try new technologies, implement “overtaking around the turns,” and prompt more enterprises to enter the new production chain. The new energy vehicle production chain or foundation formed from numerous independent development-based enterprises produced the achievement of China’s new energy vehicles, being number one in the world in production and sales. Contrast this with the IC industry. It has received no less national support than the new energy automobile industry, but so far, it is still as lacking in cohesion as scattered sand. We must draw a lesson from this.

我国重大专项实施了三个五年规划,我们看到了中国新能源汽车取得的成就,它源于20多年前中国汽车工业的自主创新运动,从而使国家计划有了依托。当时国家有关部门形成了利用新能源技术推动实施“弯道超车”的构想,而只有自主开发产品的企业才会尝试新技术,才会进行“弯道超车”,才会引发更多的企业进入新的产业链。众多的自主开发企业形成新能源汽车的产业链或基础,造就了中国新能源汽车全球产销量第一的成就。反观集成电路工业,得到的国家支持绝不亚于新能源汽车行业,但至今仍是一盘散沙,其中的教训是需要总结的。

The feasibility of fully independent manufacturing

全自主制造的可行性

In response to U.S. technology sanctions, China needs policy-level support to build an industrial base for ICs. China’s policy focus must concentrate resources on promoting the formation of China’s indigenous production chain, and change the status quo of enterprises going their own way, with decentralized pursuit of single projects or single technical indicators. The government should resolutely develop key links to promote the formation of the production chain. At present, the key links include independent R&D of the underlying technology and, especially, grasping fully independent manufacturing and using it to connect and open up the upstream and downstream production chains.

应对美国的技术制裁,中国要建起集成电路的产业基础,需要政策层面支持。中国的政策重点要集中资源促进中国本土产业链的形成,改变企业各自为政,分散地追求单个项目或单项技术指标的现状。政府要下狠心去发展促进产业链形成的关键环节。目前看,关键环节包括底层技术的自主研发,特别是抓全自主制造,以其打通上下游的产业链。

How can fully independent manufacturing be achieved? This can be divided into two steps:

如何实现全自主制造?可以分两步:

The first step is to achieve the non-Americanization of production lines, that is, using domestic and Japanese, South Korean, European, and other non-U.S. equipment and materials in production lines. China’s IC manufacturing enterprises are already working hard on this. The second step is to replace all foreign equipment and materials with domestic ones. The development of fully independent manufacturing can go upstream in the production chain to provide traction for domestic equipment and materials, and downstream to strengthen the interaction between manufacturing and design enterprises. The principle is that there cannot be any risk of being put in a “stranglehold.” Whether there is 100% domestic production or not depends on the specific situation, however. The development of fully independent manufacturing needs to rely on the cooperation of Chinese enterprises in all segments of the production chain, as the technologies of each segment interact with those of other segments. As long as you seize the manufacturing link, you have seized the bull by the horns for forming the whole production chain.

第一步是实现生产线的非美国化,即在生产线上使用国产和日本、韩国、欧洲等非美国的设备及材料。目前中国的集成电路制造企业已经在努力。第二步是以国产设备、材料全部替代国外设备、材料。发展全自主制造可以向产业链上游拉动国产设备和材料,向下游强化制造与设计企业之间的互动,原则是不能出现任何被“卡脖子”的风险。而是不是百分之百国产化,要看具体情况。发展全自主制造需要依靠产业链上各个环节的中国企业的合作,每个环节的技术都是与其他环节互动的。只要抓住制造环节,就抓住了形成整个产业链的牛鼻子。

When it comes to developing fully independent manufacturing, the world’s advanced level cannot be attained at present, but fully independent manufacturing starting from the 28 nm level is achievable. Some people ask, if Huawei’s chips are 7 nm, isn’t doing 28 nm fully independent manufacturing backward? This involves a fundamental perspective for viewing technological progress, which in turn involves two fundamental strategic issues:

目前发展全自主制造,还达不到世界先进水平,但是从28纳米的全自主制造开始,这是可以实现的。有人问,华为的芯片用的是7纳米的,做28纳米的全自主制造不是落后了吗?这涉及看待技术进步的根本性立场,它涉及两个根本性的战略问题:

First, when it comes to innovation, the foundation of technical capability is more important than the current level of technology. You must have that capability in order to grasp technological progress and innovate. At present, China’s IC industry has not formed an independent industrial base, that is, a capability base. This is also the real reason why most individual enterprises do not do deep-level technology R&D. However, relying on foreign technology will not allow the formation of a capability base, so the technological progress of individual enterprises depends more than ever on the technological progress of the entire production chain, i.e., the progress of the industrial base. The lack of a foundation is our biggest shortcoming in the China-U.S. technology war, which we have to overcome.

第一,对于创新来说,技术能力基础比当下的技术水平更重要,有能力才能把握技术进步,才能创新。目前中国集成电路行业没有形成自主的产业基础,即能力基础,这也是大部分单个企业都不做深层技术研发的真实原因。但是,依靠国外技术是构不成能力基础的,因此单个企业的技术进步比以往任何时候都更加取决于整个产业链的技术进步,即产业基础的进步。缺乏基础是我们在中美技术战中最大的短板,我们要把短板的部分补齐。

Given the current state of the industry, we can build fully independent 28-nanometer production lines, thereby opening up a connected production chain and forming the industrial base for the nation’s integrated circuits. From there, we will be able to build fully independent 14 nm production lines, and so on. Technological progress is iterative, and capabilities are developed cumulatively through product platforms.

根据目前的产业状况,我们可以建起28纳米全自主的生产线,以此打通产业链,形成国家集成电路的产业基础。由此,我们就有能力建14纳米的全自主生产线,以此类推。技术的进步是逐级迭代的,能力是通过产品平台而累积性发展的。

Second, in terms of demand and usage, the global IC market is dominated by mature process chips. Advanced process chips are only a very small slice of the market. As of 2021, advanced processes below 10 nm accounted for only 2% of global IC capacity, while chips with processes from 28-180 nm and above accounted for nearly 61% of global capacity.1 In 2021, TSMC also significantly expanded production of mature 28 nm process chips to cope with a shortage of products in the market. Advanced process chips only represented 50% of TSMC’s annual revenue that year, while the key products supporting TSMC’s capacity utilization rate were still mature process chips of 10 nm and above.2 Currently, the automotive chip world is basically one of mature 28 nm, 45 nm, and 65 nm processes, and only a few automotive chips, such as self-driving chips, require the use of advanced processes. Engineering fields such as aerospace are still using even micron-level chips, albeit in limited quantities. If China can truly form a production chain at the 28-60 nm technology level that is not subject to intervention by external forces, it will have an industrial foundation for continuous technological advancement, and at the same time, can form another kind of competitive advantage. Chinese industry has one of the world’s strongest capabilities: As soon as it learns to make a certain product, it can quickly make the cost of that product the lowest in the world and capture a large share of the market. If China takes a major share of the world market for mature process chips, it will gain a bargaining position: If the United States sanctions 20% of advanced process products, we will reciprocate by sanctioning 80% of mature process products, and from a cost perspective, U.S. automobile factories will not switch to advanced process chips. In addition, because all the enterprises that monopolize advanced process products (including TSMC) are heavily dependent on mature process products to maintain profitability, the loss of this market would seriously shake their self-confidence in sanctioning China.

第二,全球集成电路市场的需求量和使用量最大的是成熟制程的芯片。先进制程的芯片只占据了很低的市场份额:直至2021年,10纳米以下的先进制程在全球集成电路产能的占比只有2%,而28-180纳米及以上制程的芯片则占据了近61%的全球产能。2021年台积电还大幅扩产28纳米成熟制程的芯片以应对市场产品的紧缺,当年先进制程芯片在台积电全年营收中的占比只有50%,而支撑台积电产能利用率的关键产品仍然是10纳米以上的成熟制程芯片。目前,车用芯片基本上是28纳米、45纳米以及65纳米成熟制程的天下,只有自动驾驶芯片等少数汽车芯片才需要用到先进制程。航天等工程领域的芯片甚至还在使用微米级别的,虽然使用量有限。如果中国能够在28-60纳米技术级别真正形成不受外部力量干预的产业链,就有了技术不断进步的产业基础,同时形成另一种竞争优势。中国工业有一项世界上最强的能力:只要学会做某种产品,就可以把这种产品的成本迅速做到世界最低,占据很大的市场份额。如果中国在成熟制程芯片的世界市场上占据主要份额,就会获得“讨价还价”的地位:如果美国封锁20%的先进制程产品,我们就对等封锁80%的成熟制程产品,从成本的角度看,美国的汽车厂不会去用先进制程的芯片。此外,由于所有垄断了先进制程产品的企业(包括台积电)都严重地依赖成熟制程产品来保持赢利,所以失去这个市场会严重动摇它们封锁中国的自信心。

Utilizing the advantages of the national system to respond to the chip sanctions

发挥举国体制优势应对芯片封锁

How can the new national system play an effective role in independent innovation? The development of fully independent manufacturing calls for the use of entirely domestically produced equipment and materials. This requires a process, and will not be accomplished overnight. However, we need to build up our IC industry base in accordance with this goal. Once this foundation is formed, China’s upstream equipment enterprises and materials enterprises can join the technological progress chain, and downstream chip design development will not be subject to U.S. restrictions. Firmly grasping this key point will enable the connection and opening up of the whole production chain.

新型举国体制如何在自主创新上发挥有效作用?发展全自主制造要求全部采用国产设备和国产材料,这需要一个过程,不会一夜之间达成。但是我们要按照这个目标建立起我们的集成电路产业基础。一旦形成这个基础,中国上游的设备企业和材料企业就可以加入技术进步的链条,下游的芯片设计发展就不会受美国的限制。抓住这个关键点就可以把整个产业链打通。

In China’s earlier success with “two bombs, one boat” [两弹一艇] (atomic bombs, guided missiles, and submarines), the “Central Special Committee” played an important role as a special institution. In discussing the new national system, our research has found that the fundamental feature of the national system historically has been that a special decision-making and executive body directly responsible to the top leadership must be set up to execute and complete the country’s major strategic tasks; otherwise, it is difficult to mobilize the power of the whole nation. With regard to China’s industrial system today, following the success of our own large aircraft project, we need to focus on addressing the industrial-level weakness in integrated circuits. This is a major task that China must resolve. The “two bombs, one boat” program has set a good example for the construction of the IC industrial base today. The Party Central Committee can set up an organization similar to the Central Special Committee, directly responsible to the Party Central Committee, and directly take charge of this project below it, because it is difficult for any department to cope with this major task. The overall goal of this major task must be to incubate a production chain in which demand and supply relationships form between the various segments, and to cultivate the production chain’s self-circulation capability. Achieving this goal will require relying on market mechanisms, but not entirely on market mechanisms. In other words, since the development of China’s IC industry must rely on enterprises that grow through market competition, the development of this industry must rely both on market mechanisms as well as the unified coordination of various forces, including market mechanisms. Otherwise, it will be difficult to make breakthroughs. Therefore, IC industry development needs a “Central Special Committee” for the new era to directly lead the major task of integrated circuits.

当年“两弹一艇”(原子弹、导弹、核潜艇)的成功,“中央专委”这个特殊机构发挥了重要作用。在讨论新型举国体制时,我们的研究发现,历史上的举国体制的根本特点,是必须成立一个直接对最高领导层负责的特殊决策和执行机构,去执行和完成国家的重大战略任务,否则很难动员举国的力量。对于今天的中国工业体系来说,在我们自己的大飞机项目做成之后,要集中面对集成电路这个工业级的短板,这是中国必须要解决的重大任务。“两弹一艇”项目为今天集成电路产业基础的建设,树立了一个很好的榜样。党中央可以成立一个类似于中央专委的机构,向上对党中央直接负责,向下直接抓这个项目,因为任何一个部门都难以应对这个重大任务。重大任务一定要以培育各个环节之间形成需求和供应关系的产业链及其自主循环的能力为总目标。实现这个目标既要依靠市场机制,又不能完全依赖市场机制。换句话说,由于中国集成电路工业的发展必须依靠通过市场竞争成长的企业,所以发展这个工业既要依托市场机制,又要对包括市场机制在内的各种力量进行统一协调,否则难以突破。因此,发展集成电路工业就需要新时代的“中央专委”来直接领导集成电路的重大任务。

The development of independent manufacturing capacity is a key link for connecting and opening up China’s IC production chain, and for the project to succeed, market principles must be followed. Adherence to market principles has two meanings: First, there is complete reliance on enterprises to do projects, whether they are existing enterprises or new ones, and the number of enterprises can be greater than one. The ultimate guarantee of project success is the growth of the capacity of enterprises. Second, the standard of success of projects is whether products can be produced that meet the market’s needs and are competitive in terms of price and performance. For equipment and materials enterprises, it is whether the production lines work. Moreover, all product output must be products that can be mass-produced, and it must be on an industrial scale. Implementation of the fully independent manufacturing project must also rely on coordination beyond market mechanisms. The goals of the independent manufacturing project include using production lines to provide traction for the independent development of equipment and materials, and to provide manufacturing services for independently designed chips. Therefore, implementation of the project must be accompanied by cooperation among numerous companies in the production chain. Under existing structural conditions, such cooperation cannot be formed quickly (at least there will be financial risks that exceed the capacity of enterprises to bear), and it must be coordinated directly by the state institution carrying out the major task. The fundamental principles of such coordination are that production lines must use domestically produced equipment and materials, that the products of equipment and materials enterprises must meet the requirements of production lines, and that chip design enterprises must support the manufacturing and testing of production lines. Moreover, these are the only reasons for sponsoring enterprises. Of course, certain coordination work can be carried out partly through market mechanisms, but the fundamental nature of this major task is to achieve fully independent manufacturing.

发展自主的制造能力是打通中国集成电路产业链的关键环节,项目的成败要遵循市场化原则。坚持市场化原则有两个含义:第一,项目完全依托企业去做,无论是现有企业还是新建企业,企业的数量也可以不止一个。项目成功的最终保证是企业的能力成长。第二,项目成败的标准是能不能生产出在价格和性能上符合市场需要、有市场竞争力的产品。对设备和材料企业来说,就是生产线能不能用,而且所有的产品产出必须是可以大批量生产的产品,必须是有工业规模的。实施全自主制造项目还必须依靠市场机制之外的协调。自主制造项目的目标包括以生产线来拉动设备和材料的自主开发,并为自主设计的芯片提供制造服务,所以项目的实施必须伴随产业链上众多企业的合作。在现有的结构条件下,是无法迅速形成这种合作的(至少会出现超过企业承受能力的财务风险),必须由执行重大任务的国家机构来直接协调。这种协调的根本原则就是,生产线必须使用国产设备和材料,设备和材料企业的产品必须符合生产线的要求,芯片设计企业必须支持生产线的制造和试验,并以此作为资助企业的唯一理由。当然,某些协调工作可以部分地通过市场机制进行,但实现全自主制造是这个重大任务的根本性质。

By grasping fully independent manufacturing, the state will grasp the leverage for turning around the market structure. And by using policies to support the sale of fully independently manufactured chips, the state will provide equipment and materials enterprises with sales outlets and opportunities for technological progress. When all Chinese enterprises in the production chain are able to form supply and demand linkages between each other, and when all domestic technologies can gain application opportunities, China’s industrial base for integrated circuits will have been formed, and the conditions will be met for relying mainly on the power of market competition to promote the industry’s development.

抓住全自主制造,国家就抓住了扭转市场结构的杠杆。国家从政策上支持全自主制造的芯片销售,就为设备和材料企业提供了销路和技术进步的机会。当产业链上所有的中国企业都能够形成彼此之间的供应和需求联系,当所有的国产技术都能够得到应用的机会,中国集成电路的产业基础就形成了,就有条件主要依靠市场竞争的力量来推动产业的发展。

Choking demand to give China’s local chip enterprise development opportunities

卡住需求,给中国本土芯片企业发展机会

If the United States controls the supply of semiconductors as a means of suppressing China, then China can and should control the demand for semiconductors. Over the past few years, the United States has wanted to both keep China in a stranglehold, and also make money in the Chinese market. In that case, the right response for China is: Since you want to strangle me, I will not let you make money. The United States’ “nuclear weapon” is technology, and China’s “nuclear weapon” is the market. If you have the market but not the technology, you can develop the technology; if you have the technology but not the market, the technology is ultimately a dead end. In short, China must not let the United States have its cake and eat it too. Dutch lithography giant ASML’s revenue is mainly from mature process DUV lithography, rather than the most advanced. The United States is now demanding that its own enterprises stop supplying China with high-end chip manufacturing equipment, and that its allies participate in the siege of Chinese industry. If the enterprises of the United States and U.S. allies do this, it will be tantamount to choking off supply to Chinese chip production chain companies, and we can require any enterprises that carry out U.S. sanctions orders against China to submit to the Chinese government’s scrutiny when they sell in the Chinese market. These companies would then need to weigh the fact that it was the other side that started the sanctions against China first, not China that was violating any rules of free trade.

美国打压中国的手段是管制半导体的供应,那么中国就应该而且可以管制半导体的需求。几年来,美国既想卡住中国的脖子,又想在中国市场上赚钱;那么中国的应对之道就是既然你想卡我,我就不让你赚钱。美国的“核武器”是技术,中国的“核武器”是市场。有市场、没技术,可以发展技术;有技术、没市场,最后技术也是死路一条。总之,中国绝不能让美国鱼和熊掌兼得。荷兰光刻机巨头阿斯麦(ASML)的营收主要是成熟制程的DUV光刻机,而不是最先进的。现在美国要求本国的企业停止对中国提供高端芯片制造设备,要求其同盟国参与对中国工业的围堵。如果美国和其他美国同盟国的企业这么做,相当于卡住了中国芯片产业链企业的供给,我们可以要求任何一个执行美国对华制裁令的企业在中国市场销售时,接受中国政府的审查。这些企业就需要掂量,是对方先动手制裁中国,而不是中国在违反什么自由贸易的规则。

If ASML wants to follow U.S. policy and stop exporting state-of-the-art lithography machines to China, we can block its sales of ordinary lithography machines in the market after we enforce reciprocal sanctions (and actually, the number and dollar amount of ordinary lithography machines are much larger). By doing so, we can make it easier for Chinese companies to open up the domestic market for their lithography machines. By the same reasoning, China’s enterprises in the low-end chips field may have more opportunities for development, as no one will be able to stop enterprises that can make low-end products from continuing to advance to high-end ones.

如果阿斯麦要追随美国的政策停止对华出口最先进的光刻机,我们执行对等的制裁之后,可以在市场上封杀它的普通光刻机的销售(其实普通光刻机的销售数量和金额更大)。这样做,可以使中国企业的光刻机更容易打开国内市场。同理,中国中低端芯片领域的企业可能就有了更多的发展机会,谁也阻挡不了能做低端产品的企业继续向高端挺进。

In short, it is necessary for China to make a decision: In response to the U.S. government’s overall strategy of launching a technological blockade against China at all levels, we should control their access to the Chinese market, for example, by implementing controls on the orders of all foreign enterprises that sanction China in technological fields.

总之,我国有必要做出抉择,针对美国政府以通盘策略从各个层面对中国展开技术封锁,我们应该对他们进入中国市场进行管制,比如凡是在技术领域制裁中国的外国企业,对其订单实施管制。

At the same time, we must unswervingly develop China’s indigenous IC production chain, and, in particular, develop weak areas such as chip manufacturing equipment and chip materials. Centralized coordination by a national-level authority will define the development of the IC industry as a major task. This task is more complex than “two bombs, one boat, one satellite” because it involves more enterprise growth, market competition, and indirect policy coordination. This complexity places higher demands on the special institution to accomplish the task.

同时,我们必须坚定不移发展中国集成电路的本土产业链,特别是把芯片制造设备和芯片材料领域等短板发展起来。通过国家层面的权威机构,集中协调,把发展集成电路工业定义为重大任务。这个任务比“两弹一艇一星”更复杂,因为它涉及更多的企业成长、市场竞争和间接政策协调的内容。这种复杂性对完成重大任务的特殊机构提出了更高的要求。

In the face of the technological blockade, we must seek cooperation in the midst of the struggle

面对技术封锁,要在斗争中求合作

There are some who worry that if China goes for fully independent manufacturing, the relationship between China and the United States and its allies may become increasingly tense, or even that the United States and the West will accelerate their decoupling from China in the short term.

有一些人担心,如果中国搞全自主制造,中国和美国及其同盟国的关系可能会越来越僵,甚至短期内美西方会加快与中国的脱钩。

In fact, decoupling is not good for anyone, and this is our starting point. But if the other side forcefully decouples from us, we must fight back. At a time when the United States and its enterprises are enforcing technology bans on China, we cannot allow U.S. enterprises to make any more money in China. The United States has its advantages, but we have to see our own advantages. We must see that a very comprehensive industrial system is China’s strategic asset, the source of China’s strength, and China’s advantage. This industrial system includes the low end as well as the high end. Whether it is technological R&D services or labor-intensive types, they are all important. They are not divided into superior and inferior, and cannot substitute for each other.

事实上,脱钩对谁都不好,这是我们的出发点。但是如果对方强行跟我们脱钩,我们必须还手。当美国及其企业正在执行对华技术禁令的同时,不能让美国企业再赚中国的钱。美国有美国的优势,但是我们要看到自己的优势。我们要看到,非常齐全的工业体系是中国的战略资产,是中国力量的源泉,是中国的优势。这个工业体系包括低端,也包括高端,无论技术研发服务还是劳动密集型都很重要,不分高下,不能相互替代。

A few years ago, when China was in the process of reducing excess capacity and transformation and upgrading, a large amount of low-end production capacity was forced to shut down, suspend operations, merge with others, or shift to different lines of production. That had a big impact on China’s industrial system. Because the traditional industry is the largest customer of the high-tech industry, when there is large-scale compression of traditional industry, the high-tech industry will also be affected. As far as the IC industry is concerned, the stranglehold exerted against the high-end technologies of Huawei and other enterprises has sparked a nationwide discussion, but we have to see that there are a large number of low-end enterprises behind these high-end technologies, and this provides the basic conditions for technological breakthroughs in this industry. The reason why China has such a large demand for integrated circuits is because China’s downstream industry is well developed, which in turn highlights the shortcomings of the upstream. This makes even more urgent the need for breakthroughs in the upstream of China’s industry. If someone thinks that upstream industry development must come at the expense of eliminating downstream industry, that is a fruitless “climbing trees to catch fish” approach. Imagine that in 2020, Chinese industry had switched to the production of masks, and after the outbreak of the COVID-19 epidemic, masks quickly became a global public good, making an important contribution to the global fight against the epidemic. Except for materials such as melt-blown cloth, which is high-tech, mask manufacturing is low-end and labor-intensive, but nonetheless indispensable. Therefore, we must adhere to the overall development of industry, and make key breakthroughs under this precondition.

前几年中国在去产能和转型升级的过程中,大量低端产能被迫关停并转,这对中国工业体系造成很大冲击。因为传统工业是高技术工业最大的客户,大规模压缩传统工业,高技术工业也会受到影响。就集成电路行业来说,华为等企业的高端技术被卡脖子引起全民讨论,但是我们要看到这些高端技术背后有大量的低端企业为基础,为这个行业的技术突破提供了基础条件。中国之所以对集成电路有这么大的需求量,就是因为中国的下游工业发展得好,反倒把上游的短板凸显出来,这就更加紧迫地要求中国工业在上游突破。如果有人认为应该以消灭下游产业为代价来发展上游产业,那是“缘木求鱼”的做法。试想2020年中国工业转产口罩,在新冠肺炎疫情暴发后,口罩很快成为全球公共品,为全球抗疫做出重要贡献。口罩除了熔喷布等材料是高技术,生产环节都是低端劳动密集型的制造,但是它不可或缺。因此,我们必须坚持工业的全面发展,在这个前提下进行重点突破。

How can China both not decouple itself from the rest of the world and, at the same time, establish our independent manufacturing in the IC sector? Independent innovation is not about shutting the door on technology. Rather, it is insisting on doing the technology ourselves, but also learning from others. In that case, how do we go about achieving both cooperation with the United States and other Western allies on an equal footing and, at the same time, achieve independent innovation? Our strategy should be to seek cooperation in the midst of struggle, and to insist on doing technology and industry independently under open conditions. If we give up the struggle, we will be unilaterally “strangled” by the United States.

中国如何做到既不脱钩,又能在集成电路领域建立我们的自主制造?自主创新并不是关起门来搞技术,而是坚持自己做技术,但也要向别人学习。那么,怎么去实现既能与美国和其他西方同盟国平等合作,又可以实现自主创新?我们的战略应该是在斗争中求合作,在开放的条件下坚持自主做技术、做产业。如果我们放弃斗争,就会被美国单方面“卡脖子”。

We have to play to our strengths and do what we must. We do not expect that only Chinese companies will be the strongest in all areas of the semiconductor industry. We want to coexist with the world and establish equal trade relations, with each having its own advantages, but we do not accept unequal relationships where a country like the United States can strangle others with impunity. The more China backs down, the more blows it will receive. Therefore, at this time, China’s fist must be hard, and it must have the power to choke its opponent’s “throat.”

我们要发挥自己的优势,做我们自己应该做的事。我们并不指望在半导体工业的所有领域都只有中国企业最强。我们希望与世界共存,建立平等的贸易关系,彼此都有各自的优势,但我们不接受像美国那样可以肆无忌惮地把别人卡死的不平等关系。中国越退让,招致的打击就会越多。因此,这时候中国的拳头要硬,要有可以扼住对方“咽喉”的功力。

To top

Cite This Page

路风 (Lu Feng). "Faced with Technology Decoupling by the United States, China Must Establish an Industrial Base for Integrated Circuits [面对美国的科技脱钩,中国必须建立集成电路的产业基础]". CSIS Interpret: China, original work published in Economic Herald [经济导刊], December 5, 2022

FacebookTwitterLinkedInEmailPrintCopy Link