China Power

The State of China’s Autonomous Machine Computing Research

Recent Features

China Power | Economy | East Asia

The State of China’s Autonomous Machine Computing Research

A strong research sector is the foundation of the autonomy economy. How does China measure up?

The State of China’s Autonomous Machine Computing Research
Credit: Depositphotos

In a previous article, I explored the crucial role of autonomous machine computing (AMC) in China-U.S. tech competition, especially as these countries race to transition to the autonomy economy.  

In this article, I delve deeper into this topic and examine the state of AMC research in China. 

Note that the three factors that drive technology-driven economic growth are human capital, technological innovation, and financial capital. AMC research is the root of these factors in the autonomy economy – it cultivates AMC talents, generates technological innovations in the form of intellectual property, and attracts financial capital injections. 

Therefore, to assess China’s competence in the autonomy economy, it is imperative to understand the state of AMC research in the country by examining its current research output, talent pool development, and financial commitment.

Current Research Output

Research output is the most direct method to gauge the quality of research institutes. How does China’s AMC research output compare with that of the United States?

I followed the AMC research areas defined by the International Roadmap of Devices and Systems (IRDS), including computer architecture, electronic design automation, and circuit design. I then examined the AMC-related research articles published at leading semiconductor academic conferences and compared the number of articles published by Chinese institutions with those published by U.S. institutions. 

Figure 1 summarizes the results, using U.S. numbers to normalize the data (i.e., China’s research outlet is expressed as a percentage of the United States’). 

Made with Flourish

Compared with leading U.S. institutions, China still lags behind in all areas of AMC research. If we compare China’s AMC research output with that of the U.S. and its advanced semiconductor manufacturing allies such as Japan, South Korea, and the EU, the gap becomes even more staggering. 

However, note that a decade ago, semiconductor research from China’s research institutions rarely appeared at these leading conferences. In the past decade, China has leapfrogged in this field, and now Chinese research institutions, such as the Chinese Academy of Sciences, Tsinghua University, Peking University, Shanghai Jiao tong University, Alibaba, Baidu, Huawei, etc., have become major contributors in this field every year. 

With that in mind, if the Chinese government continues its intensive funding of semiconductor research, and if China’s research institutions continue their rapid rate of progress, I expect China’s AMC research output catch up with the United States’ within a decade.

Talent Pool Development

As the world transitions to the autonomy economy, the semiconductor industry has shifted its R&D focus to AMC. In particular, the new generation of semiconductor engineers is skilled in both traditional semiconductor technologies and artificial intelligence (AI), enabling them to address the emerging technical challenges of AMC. 

As summarized in Figure 2, I compare the AMC talent pools in China and the United States. I examine the size of the existing semiconductor talent pools, the number of incoming talents, and the quality of AMC talent in both countries.

Made with Flourish

Currently, there are more than 221,000 semiconductor engineers working in the semiconductor industry in China, compared to 277,000 in the United States. Due to the expansion of academic semiconductor engineering programs in China over the past decade, China supplies more than 210,000 new semiconductor engineering graduates to its semiconductor industry each year, while the U.S. produces only 90,000 semiconductor engineering graduates each year.

However, less than 15 percent  of such graduates in China enter the semiconductor industry each year, while 38 percent  of graduates in the U.S. enter the semiconductor industry. In addition, the average annual salary for semiconductor engineers in China is  $62,000, compared with $96,951 in the United States. This discrepancy reflects the significant difference in the quality of engineering education in China and the U.S.

As a result of better engineering education and higher salary, the U.S. semiconductor industry is able to attract more high-quality international semiconductor talent compared to China. In the U.S., more than 5,000 international students with advanced engineering degrees join the semiconductor industry each year, while in China, most of the semiconductor engineers are trained locally.  

Thus, although China has an AMC talent pool comparable in size to the United States, the U.S. is able to attract talent globally, while China must rely on its local talent supply.  This creates a wide quality gap unless China significantly improves the quality of its research and education.

Financial Resource Commitments

Now let us compare the financial commitment of China and the United States to their respective semiconductor industries, which are competing fiercely to develop advanced AMC chips. These financial commitments are essential catalysts for the growth of the AMC ecosystem in both countries. The size of the financial commitment, the strategy, and the efficiency of the use of the funds may determine which country will develop competitive advantages in the autonomy economy.

The United States has adopted a comprehensive strategy for funding its semiconductor industry. The CHIPS and Science Act provides up to $52 billion in federal funding to support semiconductor manufacturing and research initiatives. First, the CHIPS Act provides financial incentives to companies willing to build or upgrade semiconductor manufacturing facilities (fabs) in the United States. These incentives are designed to encourage investment in domestic semiconductor manufacturing.

Second, a significant portion of the funds will be used to support semiconductor-related research and development activities. This includes investments in cutting-edge technology, innovation and advanced manufacturing techniques to improve semiconductor production capabilities. 

Third, a portion of the funding is earmarked for workforce training programs in semiconductor manufacturing and related fields. The goal of this investment is to develop a skilled workforce capable of contributing to the growth and innovation of the semiconductor industry.

Compared to the more comprehensive U.S. CHIPS Act, China has decided to centralize its financial resources on advanced semiconductor manufacturing equipment, which is the weakest link in China’s semiconductor supply chain. It has been reported that China is set to launch a  $40 billion state-backed investment fund that focuses on advanced semiconductor manufacturing equipment, which is a direct response to Washington’s sweeping sanctions package aiming to cut China’s access to advanced chipmaking equipment. U.S. allies Japan and the Netherlands have taken similar steps. 

While the funds of China and the United States are comparable in size, the comprehensive efforts of the U.S. may enable its technology sector to quickly transition to the autonomy economy by leveraging its advantages in semiconductor manufacturing. In contrast, by making up for its weakest link in semiconductor manufacturing, China may lag behind the U.S. in its transition to the autonomy economy, but may wean itself off reliance on foreign computing solutions and become self-sufficient in its autonomy economy.

Although currently China still lags behind the United States in research capability and talent pool, I believe China can achieve AMC self-sufficiency in its transition to the autonomy economy.  Particularly, China is the largest manufacturer and the largest market of autonomous machines, and the rapidly growing domestic autonomous machine market in China will provide ample business opportunities for China’s AMC semiconductor industry. Hence it is likely that China can maintain the pace of progress in AMC research. Within a decade, China’s AMC research output can potentially catch up with the United States’.