Wu Hanming: Moore's Law may end, but innovation will never end

On September 17, the 2020 23rd China Integrated Circuit Manufacturing Annual Conference and Guangdong Integrated Circuit Industry Development Forum opened grandly. Wu Hanming, an academician of the Chinese Academy of Engineering and dean of the School of Micro-Nano Electronics at Zhejiang University, attended the conference and gave a keynote speech entitled "New Opportunities and New Challenges Facing the Development of my country's Integrated Circuit Industry", proposing suggestions for the development of my country's integrated circuits.

Wu Hanming said that in addition to the huge amount of capital and talent required for the development of my country's integrated circuit industry, there are two major barriers. The first is strategic barriers, including the Paris Coordinating Committee and the Wassenaar Agreement. We need to have a relatively controllable industrial chain, focusing on the layout of processes, equipment and materials, design IP cores and EDA. The second is industrial barriers, including the intellectual property rights deployed by the world's leading companies in the early stage. We need to have a patent library and master core technologies.

"From the evolution of the gap between the domestic integrated circuit industry and the world's advanced level, in the 1950s, when the first silicon single crystal was born, China was six years behind the United States and two years ahead of Japan. However, with the passage of time, the gap between the domestic integrated circuit industry and the world's advanced level has widened." Wu Hanming pointed out that we have understood the relationship between research and industry. Research is a means and industry is the goal. If there is a lot of research but no industry, it is still equal to zero. This is a lesson learned. Industrial orientation is very important, especially for our industry.

From the perspective of Moore's Law, Wu Hanming believes that Moore's Law has three major bottlenecks. The first is the bottleneck of materials that needs to be broken through. The second bottleneck is the physical limitations of the device. We need to use some new tertiary materials to replace the original tertiary silicon oxide. The third bottleneck is a breakthrough in technology. The photolithography process has always been the first bottleneck in our process and is a fundamental challenge. These challenges seem to be challenges encountered by the industry. In fact, these challenges must be supported by very solid basic research results. The challenges we see in the industry are all paved and supported by the results of some basic research 10 or even 20 years ago, which enables us to break through these bottlenecks.

It is worth noting that China's investment in basic research is far from enough. Wu Hanming pointed out that from the distribution of global research funds, the total amount of our R&D funds is not less than that of the United Kingdom, the United States, France and Japan, and is more than some countries, but our proportion is seriously unbalanced, mainly reflected in the fact that 5% of our funds are invested in basic research and development, while 84% of the funds are used for experimental development research and development, which is generally called trial and error. This research approach and method is actually very worrying, especially for some of our chip manufacturing companies, where the process is trial and error, and the cost of trial and error accounts for 84% of the total R&D funds. This is a very bad situation. We must have a basic guidance to turn this so-called trial and error research and development into experimental research and development.

Regarding how to measure the level of domestic integrated circuit manufacturing, Wu Hanming believes that the complete set of processes is a concentrated expression of the level of the integrated circuit industry, because it requires the complete process integration of many key technologies including lithography technology and thin film technology to form a commercially available complete set of processes to support a certain product. This can truly measure the level of integrated circuit manufacturing.

"The industry situation is rather bad now, and we have encountered very serious challenges, but there is no need to worry too much. We have come this far, and various blockades have enabled us to make various breakthroughs, including aerospace and supercomputing, all of which were achieved under blockade conditions. So if we blockade for another five or ten years, I think all problems will be solved." Wu Hanming pointed out that in 2019, the local market share was nearly 30%, the highest in our history. Although the data for 2020 has not yet come out, I am confident that the proportion of domestically produced chips will definitely go up, because many chip-using companies are actively using domestically produced chips, so this is both an opportunity and a challenge.

Wu Hanming believes that the biggest challenges in technology are actually lithography and new materials, but a complete set of technology is a landmark demand. The fragmentation in the post-Moore era may be an opportunity for our numerous small and medium-sized enterprises, especially since our localization efforts are going much smoother than before. Although Moore's Law may end, our innovation will never end.