Should the world be scared of China in semiconductors?

As per the Yole Group, Mainland China is rapidly becoming a central player in semiconductors. In 2024, China held 21% of foundry capacity, despite generating 5% of the global wafer demand. Much of this excess capacity is either foreign-owned, or offered as open foundry services, although utilization rates remain below the global average. 

By 2030, China is projected to dominate the global foundry landscape, accounting for 30% of installed capacity, outpacing Taiwan, South Korea, and Japan. Now, that's something to ponder about! 

As of 2024, Taiwan led the foundry ecosystem share at 23%. It is followed by China 21%, Korea 19%, Japan 13%, USA 10%, Europe 8%, and others 6%. Southeast Asian region, especially Singapore and Malaysia, owns 6% of the global foundry capacity, although it lacks domestic players, and operates on foreign-owned foundries. 

I am not even talking about India, as we have a very long way to go!

Foundry capacity is forecasted to grow at 4.3% CAGR from 2024 to 2030. Mainland China will lead in foundry capacity by 2030, with 30% of global share. US players account for 57% of wafer demand, but own only 10% of foundry capacity. Global foundry utilization will remain at about 70%, with growth heavily dependent on demand.

The semiconductor equipment industry has consolidated over the last 20 years, except in China. It has gone the opposite -- from nine companies to 27. Should that be a worrying aspect? Maybe, experts can throw more light on this!

China is also a large consumer of fabrication equipment. Last year, China consumed 35%. In contrast, USA consumed 13%, and Japan, 9%. Growth in 2019-24 saw four US companies grew from 1-18 percent CAGR. In China, for two companies, its 50-60% growth. 

China tends to measure success in decades, and not in quarters. They are willing to do what is needed. Do look up China Integrated Circuit Industry Investment Fund, if you wish to know more.

There has been an increase in fabrication across South Korea and Taiwan. Their philosophy has been: we have to be the very best. There are partnerships with US and European companies. However, China also wants the best tools, and will use Chinese products for as long as possible. They are indigenizing tools and equipment, as much as possible.

Extremely precise production
Semiconductor production is extremely precise. You have to get 99.9999 percent yield at each step. Therefore, you simply cannot afford to have any substandard equipment in your fab!

Chip fabricators also need to first qualify the specific tools that are to be used for various processes. Two to three tools generally qualify for any step. Any delay here in qualification of tools can be massive for any fab. Chinese semiconductor manufacturing equipment (SME) folks also know this. They are now trying to qualify their tools. 

Next, deep ultraviolet (DUV) technology is more mature, and seeing steady incremental improvements. Extreme ultraviolet (EUV) lithography faces some significant ongoing challenges, despite potential for higher resolution. We can expect to see some progress in DUV across China, but EUV is going to be much tougher. In fact, developing EUV machinery is going to be much tougher for anyone!

ASML remains the only company today, producing EUV machines. Nikon is another player, but they seem to focus more on DUV. Using EUV light, the ASML NXE and EXE systems deliver high-resolution lithography, and enable the mass production of world’s most advanced microchips possible.

Georgetown’s Center for Security and Emerging Technology (CSET), USA, recently came up with a report by Jacob Feldgoise and Hanna Dohmen on China's chip-making market. A look at some categories:

CMP: Hwatsing Technology helped increase China’s market share by eight percentage points (pp) over the five-year period (2019-24).
Dry etch and clean: China gained parity with Japan in dry stripping tool market and achieved an eight pp rise in the dry etch market.
Deposition: NAURA, Piotech, and AMEC have played a key role in increasing China’s market share in deposition from 2% to 7% over the five-year period.
Lithography: ASML and Nikon continue to control the market. China’s Shanghai Micro Electronics Equipment (SMEE) has only reached a 4% market share in i-line lithography tools, which are used for legacy chip fabrication.

However, lithography, the most complex tool category, remains a critical constraint for China.

In test, since 2019, Hangzhou Changchuan and AccoTEST have driven China’s market share to almost 70% in linear and discrete test tools. They have contributed to solid gains in burn-in and system-on-a-chip (SoC) test tools, as US suppliers ceded ground.

Mainland China is strengthening local development and manufacturing of chip-making equipment. That's the main purpose of the third installment of the National Integrated Circuit Industry Investment Fund, also called Big Fund III. This is to address most notorious gaps in China’s semiconductor industry — lithography tools and EDA.

Further, China’s 28nm-65nm market share surged from 18% in 2020 to 32% in 2023. Many fear that China may get into mature-node dumping push that could destabilize the semiconductor industry growth.

Focus on one nation, or losing focus
I have always been wondering why the world is so 'seemingly focused' on just one nation, and how can they stop it from moving ahead, rather than paying attention to what they themselves are doing in semiconductors? Or, are they all losing focus? If so, they all need to rethink their own strategies separately, and very quickly.

A case in point is working on developing EUV! So far, no one else, barring Nikon, is even trying to take on ASML. And, they are some distance behind. Where is the thought: if one can do it, so can I?

As I mentioned earlier, India still has a very long way to go. Do not be swayed by any current report that you may have come across. India is currently working on developing 28nm chips. And, China has the potential of flooding the market with its own mature node chips. 

Six months is a very long time in semiconductors, and one year is considered ancient history. Also, India is still not learning from countries like Taiwan and South Korea. If India really wants to be known as a significant contributor in semiconductors, India really needs to start pushing hard. Once India is able to build a strong local supply chain, we can definitely progress. 

Maybe, I cannot do anything much, except probably, write. And, I may be wrong in my assessments too. If so, kindly pardon me, as my knowledge is very limited. However, there are much, much more able folks within the Indian government, and within the Indian semiconductor industry, who can take note, and make things happen.

Nevertheless, the world has to scale up semiconductors in the future. We also need better semiconductor manufacturing equipment and materials to tackle future challenges.