虎嗅

Dialogue with Academician Chu Junhao of the Chinese Academy of Sciences: Moore's Law Has Its Limits; "Tao's Law" Opens Up New Possibilities. Chinese Technology Must Strive for Excellence in Fundamental Research

原文:对话中国科学院院士禇君浩:摩尔定律有天花板,“韬定律”开拓新思路,中国科技要在基础研究上做到极致

Summary of Key Points

Huawei's proposed "Tao Law" has recently gone viral, sparking discussions about a "paradigm shift" in the semiconductor industry. Chu Junhao, an academician at the Chinese Academy of Sciences and internationally recognized as the founder of the "CXT formula," was interviewed to explain its underlying logic: shifting from Moore's Law's focus on "dimensional reduction" to "time reduction"—that is, improving chip performance by reducing circuit resistance and capacitance and using three-dimensional architectures to shorten signal transmission times. He believes that Tao Law opens up new avenues for the post-Moore era, but whether it will become a global consensus depends on industry validation. He also emphasizes that China still needs to make breakthroughs in EUV lithography machines; combining both technologies can achieve more than the sum of their parts (1+1>2). Chu points out that for Chinese technology to move from "following" to "leading," it must address shortcomings in fundamental research and pursue ultimate innovation.

I. Tao Law: Not Replacing Moore's Law, but Finding a New Path

  • The Dilemma of Moore’s Law: For the past 60 years, chip performance has been improved by making them smaller (doubling transistor density every two years), but with silicon atoms having a diameter of only 0.2 nanometers, electron tunneling effects become significant when manufacturing processes approach 1 nanometer, severely reducing performance—this path is nearing its limit.
  • The Solution of Tao Law: Instead of focusing on size, the focus shifts to time. The time it takes for a signal to travel through a circuit (denoted as "time constant τ" in physics) is determined by resistance (R) and capacitance (C); Tao Law aims to reduce τ by using three-dimensional architectures and optimizing materials to decrease resistance and capacitance, thereby increasing speed. Huawei has achieved a 53.5% increase in transistor density with its first generation of products, reaching speeds of 4-5 GHz, demonstrating significant results.
  • Key Difference: Moore’s Law is an empirical rule established by the industry, while Tao Law is driven by fundamental physical principles and can guide the development of new technologies such as quantum and photonic chips.

II. Do We Still Need EUV Lithography Machines with Tao Law?

Many ask if we no longer need to invest in EUV technology since Tao Law can improve performance. Chu Junhao clarifies that the two are not substitutes but complementary:

  • The Role of EUV: It enables the production of even smaller chips (below 3 nanometers) and is essential for strengthening chip capabilities.
  • The Role of Tao Law: It enhances performance within existing dimensions through architectural optimizations, providing additional efficiency improvements.
  • Combined Effect: The combination of EUV (for smaller sizes) and Tao Law (for faster speeds) can lead to substantial performance gains (1+1>2), making chips both smaller and faster.

Therefore, China will not abandon efforts to develop EUV technology; instead, it aims to leverage both to gain a competitive advantage globally.

III. Can Tao Law Help China’s Semiconductor Industry Overtake?

Tao Law offers a new path for China’s semiconductor industry:

  • Huawei’s Success: Huawei has proven the feasibility of this approach, achieving a 53.5% increase in transistor density in one generation—far faster than the traditional Moore’s Law progression of three years.
  • Industry Impact: This could drive the entire supply chain (design, packaging, investment, etc.) towards "time reduction." If domestic companies follow Huawei's lead and deepen their research, they may gain a competitive edge in this new direction.
  • Uncertainty: Whether Tao Law will become a global standard depends on widespread industry adoption and more data supporting its effectiveness. Foreign companies like Intel and TSMC may continue to follow Moore’s Law but will also consider the advantages of Tao Law.

IV. What Does China Need to Move from “Following” to “Leading” in Technology?

As someone with firsthand experience, Chu Junhao identifies two critical shortcomings:

1. Lack of Perfection: Many technologies can be developed, but not always optimized to the highest standards. For example, laboratory equipment is often imported, and domestically produced equipment lacks precision and stability compared to foreign counterparts.

2. Insufficient Fundamental Research: Many technologies are understood in terms of implementation but not the underlying physical principles, limiting continuous improvement and innovation.

3. Lack of Disruptive Breakthroughs: While China is making progress in various fields, there are few truly groundbreaking innovations that can transform industries. Tao Law is a good start, but more fundamental innovations like the "CXT formula" are needed.

Chu emphasizes that the depth of foundational research determines the potential of a country's technological capabilities. In the next 10-20 years, China must continue to invest heavily in fundamental research to truly become a global technology powerhouse.

Conclusion

The significance of Tao Law goes beyond mere technological innovation; it represents a shift from following international rules to setting its own standards. However, achieving true leadership requires industry adoption, addressing foundational research gaps, and pursuing ultimate innovation. As Chu Junhao once wrote as a young scientist, "We need Chinese names in the ranks of those who establish scientific laws." Tao Law is a testament to this ambition. In the future, we need more such "Chinese laws" to secure a strong position in the global technology landscape.