扒车喵近期，美国正在试图拉拢日本、韩国以及中国台湾地区，组建一个类似于美日印澳“四方机制”的“芯片联盟”，意图以此来遏制中国半导体产业的发展，这让本趋于平稳的半导体江湖再起波澜。在海外问答论坛Quora上，美国网友提问道：中国的半导体芯片产业可以发展到欧美级先进水平吗？这引起各行业专家和网友的关注和热议，我们看看相关专业解答。
问题：
芯片专家卡兰•梅塔的回答
Compared with the United States, China has been difficult to keep up in the semiconductor field. Semiconductors are very different from other technical fields, because China is severely restricted in the supply of skilled technicians in this field. You know, it is much easier and cheaper to create a computer scientist than a semiconductor engineer.
与美国相比，半导体领域中国一直难以跟上。半导体与其他技术领域有很大不同，中国在该领域熟练技术人才的供应方面受到严重限制，要知道创造一个计算机科学家比创造一个半导体工程师要容易得多，也便宜得多。
I work in the world’s largest semiconductor company. I am a semiconductor device R & D designer. Doctors in various fields are everywhere. Chemical engineers and electrical engineers work together, mechanical and material science engineers discuss problems with each other, and some physics doctors.
我在世界上最大的半导体公司工作，是一名半导体器件研发设计师，在这里各领域的博士到处都是，化学工程师和电气工程师一起工作，机械和材料科学工程师相互讨论问题，还有一些物理博士。
To have a competitive semiconductor manufacturing plant requires a large number of highly skilled labor force. Compared with other industries, even technicians need to be extremely skilled. When you deal with things on the submicron scale, the complexity increases exponentially, the space for errors is very small, and the cost of each error may cost millions of dollars to make up for.
要想拥有一个有竞争力的半导体制造工厂，需要大量拥有高技能的劳动力，与其他行业相比，即使是技术人员也需要异常熟练。当你在亚微米尺度上处理事物时，复杂性会呈指数级增长，错误的空间非常小，每一个错误的代价都可能要花费数百万美元来弥补。
Therefore, a country needs a very strong technical education system to support its semiconductor industry. There are many top universities in China, so what’s the problem? Localized and highly specialized knowledge.
因此，一个国家需要一个非常强大的技术教育体系来支持本国的半导体产业。中国有很多顶尖大学，那么问题是什么呢？本地化和高度专业化的知识。
Most of the best semiconductor professors in the world live in the United States, so the best students receive relevant knowledge training here, work and experiment in Intel or NVIDIA foundries in the United States. So the question now is, can China attract the best talents from American academia and industry to China? I’m afraid it’s hard.
世界上最好的半导体专业教授大多住在美国，所以最好的学生在这里接受相关知识的培训，在美国的英特尔或英伟达铸造厂工作和实验。所以现在的问题是，中国能否吸引美国学术界和工业界的最佳人才来中国？这恐怕很难。
TSMC in Taiwan and Samsung in South Korea have also achieved this, but the values of these developed economies are similar to those of western countries, and people with Western-style higher education may prefer these places. In addition, both TSMC and Samsung have advanced R & D teams and offices in Silicon Valley, and R & D can only be carried out in places with rich talents.
台湾地区的台积电和韩国的三星也做到了这一点，但这些发达经济体的价值观与西方国家相似，受过西式高等教育的人可能更青睐这些地方，此外，台积电和三星都在硅谷拥有先进的研发团队和办事处，研发只能在人才富集的地方进行。
Earlier, China solved this problem by providing strong economic and professional incentives to Chinese students with advanced degrees in the United States to attract them back to China, but Trump’s new move to ban Chinese students from studying in American universities will prevent this path of talent return. In addition, U.S. intellectual property laws and intellectual property security have become more stringent and important in high-tech companies, making it difficult for China to obtain relevant technologies.
早些时候，中国通过向获得美国高等学位的中国学生提供强有力的经济和职业激励，吸引他们回国，从而解决了这个问题，但特朗普禁止中国学生在美国大学学习的新举措将阻止这条人才回流的道路。此外，美国的知识产权法律和知识产权安全在高科技公司中变得愈加严格和重要让中国难以获得相关技术。
A modern cutting-edge semiconductor manufacturing plant costs more than $10 billion. Do you think they will bother to invest in semiconductor related high-tech industries in China? Some of these tools cost more than $10 million and require highly skilled (even doctoral) personnel to look after, operate and modify them.
一座现代尖端半导体制造厂的成本超过100亿美元，你认为他们会费心在中国投资半导体相关的高科技行业吗？其中一些工具的价格甚至都超过了1000万美元，需要高技能（甚至是博士水平）的人员来照看、操作和修改它们。
In addition, people working in factories are often limited to narrow areas, so few people know the whole process of manufacturing chips, which reduces the risk of employees leaving the company to go to competitors, because few people know the overall technology and drawings of the whole industry. All these mean that it will be extremely difficult to establish a semiconductor manufacturing plant from scratch in China. This technology is highly sensitive, strictly regulated and labor is scarce. Unlike computer science, semiconductor manufacturing is an industry that cannot be created by money alone.
此外，在工厂工作的人往往被限制在狭窄的区域，所以很少有人知道制造芯片的整个流程，这使得员工离开公司去竞争对手那里的风险降低，因为很少有人知道整个产业全局的技术和图纸，所有这些都意味着，要想在中国从头开始建立一个半导体制造厂将是极其困难的。这项技术高度敏感，监管严格，劳动力稀缺，与计算机科学不同，半导体制造业是一个单靠金钱无法创造的行业。
In other words, in terms of all key indicators – time, cost and manpower, breakthroughs in artificial intelligence, biotechnology or chip design are several orders of magnitude easier than semiconductor manufacturing.
换言之，在所有关键指标——时间、成本和人力等因素上，人工智能、生物技术或芯片设计领域实现突破的可能性都要比半导体制造业容易几个数量级。
美国工程技术专家西蒙•奎纳汗的回答
The semiconductor supply chain network itself is complex, including design companies focusing on chip R & D, foundries manufacturing chips (Korean and Taiwan enterprises have the largest share), and semiconductor capital equipment suppliers providing machinery (semcap), which are currently located in Europe, the United States and Japan.
半导体供应链网络本身是复杂的，包括专注于芯片研发的设计公司、制造芯片的铸造厂（韩国和台湾地区的企业拥有最大份额），以及提供机械的半导体资本设备供应商（SEMCAP），目前位于欧洲、美国和日本。
Chinese companies such as Huawei and SMIC are already designing cutting-edge semiconductor chips, which have reached the same technical level as their counterparts in Europe and the United States. However, in order to manufacture these advanced semiconductor chips, they need to rely on Taiwan foundry factories such as TSMC to manufacture semiconductors. Some of the equipment used by these foundry factories or the design patents of these equipment are controlled by the United States or designed with American software. Therefore, in the Sino US trade war, Americans are bent on destroying successful Chinese high-tech companies. They regard these devices or software as weapons against China.
华为、中芯国际等中国公司已经在设计尖端半导体芯片，而且已经达到欧美同业同等技术水平，然而，为了制造这些先进的半导体芯片，他们需要依赖台积电等台湾代工厂来制造半导体，这些铸造厂使用的一些设备或这些设备的设计专利是由美国掌控的，或是用美国软件设计的，所以在中美贸易战中，美国人一心想摧毁成功的中国高科技公司，他们将这些设备或软件视为武器对付中国。
The United States first aims at chip manufacturing tools. From design to manufacturing and even testing, no company has all the advantages in manufacturing the best chips, even TSMC, Samsung or Intel. The United States uses patent trumps to force all countries and enterprises to comply with U.S. laws and regulations. In the future, China will have to pursue a “clean supply chain” that is not affected by the United States.
美国首先瞄准的是芯片制造工具，从设计到制造，甚至测试，没有哪家公司在制造最佳芯片方面拥有全部优势，即使是台积电（TSMC）、三星（Samsung）或英特尔（Intel），美国使用专利王牌迫使所有国家和企业都要遵守美国法律规定。中国今后将不得不追求一个不受美国影响的“清洁供应链”。
The Chinese have been looking forward to and planning for years. At present, many Chinese foundries like Yangzi can produce 12NM semiconductors, and it is only a few months away from mass production.
中国人已经期待并计划了几年了。现在有很多像扬子这样的中国铸造厂可以生产12nm半导体，而且距离批量生产只有几个月的时间。
In addition, foreign foundries such as Samsung are completing advanced foundries in China and expanding foundries in South Korea under the expectation / commitment of Chinese enterprises, as are European companies such as St micro. These companies are stripping or have stripped American components from their equipment because they hope to make considerable profits by selling products to the Chinese. In addition, China is designing new manufacturing processes and using new materials for the semiconductor industry.
此外，三星（Samsung）等外国铸造厂正在完成在中国的先进铸造厂，并在中国企业的预期/承诺下扩大在韩国的铸造厂，ST Micro等欧洲公司也是如此。这些公司正在剥离或已经剥离其设备中的美国部件，因为他们希望通过向中国人出售产品来赚取可观的利润，此外，中国正在为半导体行业设计新的制造工艺和使用新材料。
This will be a complex task. It may take 5 to 10 years for China to develop chips with a 5-nm process, but eventually China’s economy will be self-sufficient in this field, because it is about the future. The future is digital, and the digital economy needs chips.
这将是一项复杂的工作，中国距离研发出5纳米制程的芯片可能需要5到10年的时间，但最终中国经济将在这一领域实现自给自足，因为这事关未来，未来是数字的，数字经济需要芯片。
The United States is not far ahead in the semiconductor market. In terms of manufacturing, South Korea and Taiwan have 81% market share. In terms of mechanical lithography, only ASML in the Netherlands can produce the extreme ultraviolet (EUV) lithography equipment required to manufacture the most advanced chips.
美国在半导体市场也并非远远处于领先地位，就制造业而言，韩国和台湾拥有81%的市场份额，就机械光刻机而言，只有荷兰的ASML才能生产制造最先进芯片所需的极紫外（EUV）光刻机设备。
Time will tell us how long it will take for China to catch up with its neighbors and Europe, Japan and the United States. It is we who have left China with no choice but independent research and development.
时间会告诉我们，中国最终需要多长时间才能赶上邻国以及欧洲、日本和美国，是我们让中国除了独立研发之外别无选择。
海外华裔工程师蒋家伟的回答
I can’t predict when China will achieve a breakthrough in the chip field，but you can see China’s determination and execution from the following historical events.
我不能预测什么时候中国在芯片领域可以，但你可以从以下历史事件中看到中国的决心和执行力。
China’s nuclear weapons program was launched on January 15, 1955. Later, there was a bad relationship between China and the Soviet Union. On June 20, 1959, Nikita Khrushchev decided to withdraw Soviet experts who assisted China, and China turned to a comprehensive and independent policy of self-reliance.
中国的核武器计划于1955年1月15日启动，后来中苏交恶，1959年6月20日，尼基塔·赫鲁晓夫（Nikita Khrushchev）决定撤回援助中国的苏联专家，中国转向了自力更生的全面自主政策。
Then, the 596 project was officially established, which means that China’s first independent nuclear weapons project was launched. By January 14, 1964, Lanzhou nuclear power station had successfully enriched enough fissile uranium 2351. On October 16, 1964, a 1550 kg U-235 fission implosion device was detonated.
紧接着，596项目正式组建，这意味着中国首个独立的核武项目启动。到1964年1月14日，兰州核电站成功地富集了足够多的可裂变铀235，1964年10月16日，一个重达1550公斤的铀-235裂变内爆装置被引爆。
In 32 months, China successfully detonated its first atomic bomb (October 16, 1964), launched its first nuclear missile (October 25, 1966), and detonated its first hydrogen bomb (June 14, 1967).
在32个月的时间里，中国成功地引爆了第一颗原子弹（1964年10月16日），发射了第一枚核导弹（1966年10月25日），并引爆了第一颗氢弹（1967年6月14日）。
In order to ensure national defense security, the Chinese military tried to reduce its dependence on the global positioning system (GPS) operated by the US air force. The idea of developing Beidou was formed in the 1990s. When the first Beidou satellite was launched in 2000, its coverage was limited to China.
为了保障国防安全，中国军方试图减少对美国空军运营的全球定位系统（GPS）的依赖，开发北斗的想法在20世纪90年代形成。2000年发射第一颗北斗卫星时，覆盖范围仅限于中国。
With the expansion of the use of mobile devices, in September 2003, China plans to join the European Galileo positioning system project and plans to invest 230 million euros in Galileo in the next few years. China can invest a huge amount of money, but it cannot be treated commensurate with it. China is not only unable to enter the decision-making body of the Galileo plan, but also blocked by the deliberate obstacles set up by the European Space Agency in the development of technical cooperation. China is very dissatisfied.
随着移动设备使用的扩大，2003年9月，中国打算加入欧洲伽利略定位系统项目，并计划在未来几年向伽利略投资2.3亿欧元。可投入巨额资金，却得不到与之相称的对待，中国不但进不到“伽利略”计划的决策机构，甚至在技术合作开发上也被欧洲航天局故意设置的障碍所阻挡，中方十分不满。
Later, China resolutely withdrew and focused on the research and development of Beidou navigation system. At that time, people thought that China’s “Beidou” navigation system would only be used by its army. Surprisingly, China built a Beidou navigation system comparable to GPS in the field of global navigation.
后来中国毅然退出，专注于北斗导航系统的研发。当时，人们认为中国的“北斗”导航系统将只供其军队使用，可令人惊喜的是，中国在全球导航领域建成了可与GPS匹敌的北斗导航系统。
The US government made a decision in 2015 to restrict China’s use of Intel’s fastest microprocessor. As a last resort, China can only build its supercomputing capacity independently and steadily, including efforts to develop its own microprocessors.
美国政府在2015年做出决定，限制中国使用英特尔最快的微处理器。不得已之下，中国只能独立稳步建设其超级计算能力，其中包括努力开发自己的微处理器。
In 2011, China produced a relatively small supercomputer, which tentatively used domestic processors, but its large-scale computing system relied on American processors until 2016. In 2016, China released the latest supercomputer, which is a single-chip system with 10.65 million computing cores. It is completely constructed by Chinese microprocessors, and its computing power ranks first in the world.
2011年，中国生产了一台相对较小的超级计算机，该计算机尝试性地使用了国产处理器，但其大型计算系统在2016年之前一直依赖于美国处理器。2016年，中国发布了最新的超级计算机，这是一个单片系统，拥有1065万个计算核心，完全由中国微处理器构建，其计算能力跃居世界第一。
In just 35 years, China has developed from a backward agricultural society to an industrial power. 35 years ago, China’s per capita income was only one third of that of sub Saharan Africa. Today, China is the largest manufacturing power in the world.
在短短35年内从落后的农业社会发展成为工业强国，35年前，中国的人均收入仅为撒哈拉以南非洲的三分之一，如今，中国是世界上最大的制造业强国。
it produces nearly 50% of the world’s major industrial products, including crude steel (800% of the level of the United States, 50% of the global supply), cement (60% of the world output) and coal (50% of the world output), China is also the world’s largest producer of ships, high-speed trains, robots, tunnels, bridges, highways, chemical fibers, machine tools, computers and mobile phones. I believe China is a country good at creating miracles.
它生产了世界上近50%的主要工业产品，包括粗钢（美国水平的800%，全球供应的50%）、水泥（占世界产量的60%）、煤炭（占世界产量的50%），汽车（超过全球供应量的25%）和工业专利申请（约为美国水平的150%），中国也是世界上最大的船舶、高速列车、机器人、隧道、桥梁、公路、化学纤维、机床、计算机、手机等生产国，我相信中国是一个擅长创造奇迹的国家。
