Chinese semiconductor industry

Wangxi said:

Please, Log in or Register to view URLs content!

Click to expand...

Excellent news. This basically confirms what we have been hearing all along. This SMEE 28nm Litho machine is equivalent to the latest ASML Duv machine which is used by the likes of TSMC etc.

weig2000 said:

Yes, these are the two key areas of strategic and foundational technology that China is still quite behind the world's best. There are areas China is also behind, but they're either niche or not as strategic and foundational as these two.

Neither of them is easy for China to catch up in a short period of time with the world's best. But if you insist, I'd say it's semiconductor. Two of the more important reasons:

First is the size and scope of the related markets and industries in China. China has the world's largest electronic manufacturing and ICT industries and is the world's largest technology product exporter. Yes, a lot of them are designed outside of China, but still. Think of Apple, which has created a thriving indigenous and increasingly sophisticated smartphone supply chain, upon which Huawei, Xiaomi, Oppo and Vivo have developed and evolved. There are also other industries other than smartphone, including consumer electronics, industrial electronics, telecommunication. Contrary to what a lot of people in the West like to believe, China is much, much more than just assembling the final products. It has a lot of its own brands and OEMs. Semiconductor industry is the upper stream industry and will benefit from all these large, vibrant and sophisticated downstream industries. Also, if you look at the semiconductor industry itself, China is probably the only country/economy other than the US to have players in equipment, material, design, manufacturing, testing & packaging of semiconductor. In other words, it's an all around player. Sure, most of them are quite behind the latest generation of technologies and have rather small market share. That's mostly because they're late to the fast-changing game. Now, thanks to the US sanctions, ...

China also has a very large domestic aviation market, the second largest and soon will be the largest in the world. That is why China may still pull off to create a viable aero engine industry. But civil aircraft industry is a mature and relative stable market. It's a B2B industry and customers are risk averse. It's hard for China to break into the industry globally other than its domestic industry having coming to the game so late. Other than China, I can't think of any other country having even a chance.

The second is the diffusion of the related semiconductor technology and industry. The US has the strongest semiconductor industry, all considered. But it hardly dominates in everything semiconductor. South Korea, Japan, Taiwan, and Europe all have strong sectors or niche within the semiconductor industry. It is difficult for the US or anyone else to corner any particular technologies for long. China has been recruiting talents from Taiwan, South Korea and Japan and these outside helps have made a big difference.

Compare to the aero engine industry, which has only a few strong players (GE, HP, R&R) with very high entry-barriers. It is a very mature industry and is difficult to be disrupted by new technologies or new players. The know-how has been accumulated over many decades and are closely guarded by fewer players.

China has come a long way in designing and manufacturing military aero engines, but it still has a long way to go to produce civilian aircraft engines - its indigenous civilian aircraft industry has barely started.

Click to expand...

I. agree with every thing you said. I think necessity is the most important thing that drives progress and development. From a technical perspective both semis and jet engines are difficult and expensive technology to master.

As you noted China has come a long way in developing military jet engines but has basically no civil jet engine. The only reason this is the case is because we got sanctioned by the US and that prevents us from buying military jet engines. So we had to put in the money and effort to develop it ourselves. There isn't that motivation on the civil engine front. Since it's always easier buying boeing or Airbus jets or GE or RR engines for the C919.

Same goes for IC equipment. No body in China wanted to be in the IC equipment industry 5 years ago. There was no demand for domestic IC equipment, so basically no domestic company invested seriously in developing cutting edge IC equipment. Basically it was easier and more reliable to import IC equipment from Holland or Japan or the US. Nor was there any market demand for domestic fabs. Hence the deplorable state of Chinese IC equipment industry.

My cousin studied electrical engineering and was super envious of software engineers who were getting high paid jobs in the Bay Area and Shenzhen 4 years ago. But now at least in China, things have changed. Billions of VC capital is being channeled to any firm that is in the semiconductor industry. Universities, private companies, government big finds, private investors are all coming together in a Manhattan Project type effort to achieve decoupling from the US in semiconductors.

This simply isnt the case for civil jet engines. Sure we would all really be proud of a domestic engine for civil airlines. But it isn't absolutely necessary right now. Even if the US stops selling us GE engines we can still buy it from RR or last resort we could team up with Russia or Ukraine to find a solution.

WTAN said:

Excellent news. This basically confirms what we have been hearing all along. This SMEE 28nm Litho machine is equivalent to the latest ASML Duv machine which is used by the likes of TSMC etc.

Click to expand...

Noob question, whats the main difference between DUV and EUV, besides the light source and wave length?

free_6ix9ine said:

Noob question, whats the main difference between DUV and EUV, besides the light source and wave length?

Click to expand...

The main difference is really the light source. Another difference is that the EUV machine most likely uses a more advanced type of projection lense as it has to handle a greater amount of plasma produced when the CO2 Laser hits the tin droplets. Apart from that the other features of the 2 types of machines should be the same.

WTAN said:

The main difference is really the light source. Another difference is that the EUV machine most likely uses a more advanced type of projection lense as it has to handle a greater amount of plasma produced when the CO2 Laser hits the tin droplets. Apart from that the other features of the 2 types of machines should be the same.

Click to expand...

Got it ok thanks. So in terms of the light source, which component or process is the most difficult part? Ie. a powerful enough laser is needed, the shape of the tin droplets, timing the laser and droplets, collecting the plasma that is generated?

AndrewS said:

Saw that a while ago.

It's a hard 2 year deadline for Huawei to de-Americanise.

The worse case scenario is that Huawei shrink for a while.
The Chinese government isn't going to let them go under, and Huawei do have the capacity to operate in many segments without US technology.

Then they will start to grow again, based on non-US technology.

---

But let's say Huawei did go bankrupt for argument's sake.

They would be taken over by the Chinese government, and then be tasked with developing Chinese technology self-sufficiency.

Click to expand...

Huawei was really naive to believe that it could retain telecommunication dominance with a reliance on foreign and especially US tech. The Americans have for decades telegraphed their hostility towards China attaining state of the art technological capabilities and for long Hawks have urged a policy of blockade of advanced chips and other tech to China and have advocated persuading their allies in other countries to join in it. It was predictable that the present embargo would come and that American companies making bumper sales to China would be unable to prevent their government to enact an embargo against their interests of retain access to the lucrative Chinese market and that the United States would strongarm, if necessary, all other countries to do the same.

Huawei is one of the culprits with regards to China lagging with regards to IC chip and IC chip making equipment, because it refused to patronize domestic chip chip manufacturers and did not engage in chip manufacturing of its own.

Weaasel said:

Huawei was really naive to believe that it could retain telecommunication dominance with a reliance on foreign and especially US tech. The Americans have for decades telegraphed their hostility towards China attaining state of the art technological capabilities and for long Hawks have urged a policy of blockade of advanced chips and other tech to China and have advocated persuading their allies in other countries to join in it. It was predictable that the present embargo would come and that American companies making bumper sales to China would be unable to prevent their government to enact an embargo against their interests of retain access to the lucrative Chinese market and that the United States would strongarm, if necessary, all other countries to do the same.

Huawei is one of the culprits with regards to China lagging with regards to IC chip and IC chip making equipment, because it refused to patronize domestic chip chip manufacturers and did not engage in chip manufacturing of its own.

Click to expand...

It's not Huaweis fault, no one in China bothered with developing IC equipment or Fab capilities, so Huawei had no choice. It's not economical, if Huawei wanted to scale up quickly they had to rely on US technology. But I agree with the gist of your post, instead of blindly worshiping foreign brands or technology or education etc. We should learn to give business to our domestic companies to help them improve their technology.

free_6ix9ine said:

Noob question, whats the main difference between DUV and EUV, besides the light source and wave length?

Click to expand...

Besides the EUV light source, you need a different lens system to adapt to the new wavelength, and a better wafer stage system (not directly related to EUV, but due to the increased precision requirements of sub-7nm nodes). These are basically the 3 core modules of a lithography machine.

Weaasel said:

Huawei was really naive to believe that it could retain telecommunication dominance with a reliance on foreign and especially US tech. The Americans have for decades telegraphed their hostility towards China attaining state of the art technological capabilities and for long Hawks have urged a policy of blockade of advanced chips and other tech to China and have advocated persuading their allies in other countries to join in it. It was predictable that the present embargo would come and that American companies making bumper sales to China would be unable to prevent their government to enact an embargo against their interests of retain access to the lucrative Chinese market and that the United States would strongarm, if necessary, all other countries to do the same.

Huawei is one of the culprits with regards to China lagging with regards to IC chip and IC chip making equipment, because it refused to patronize domestic chip chip manufacturers and did not engage in chip manufacturing of its own.

Click to expand...

Quite the opposite. Huawei has been the least naive civilian sector company in China on recognizing the true political nature of global competition. They experienced this since as early as 2003, when Cisco filed a major lawsuit against them. Since 2007, Huawei has been subjected to constant attacks from the US government and media. Huawei's founder Mr RenZhengfei is famous for his survivalist and crisis-centric philosophy, constantly thinking about how Huawei could collapse even during the exceptionally good years. This was well known back in the early 2000s, and has become an important element in Huawei's corporate culture.

The book "Will Huawei be the Next to Fail" was a book about the philosophy and culture inside Huawei, and it explained this crisis-centric philosophy very well, even with just the title. Huawei's 2012 lab for fundamental research was also named 2012 for a reason. and one of its sub-division is called the Noah's Ark Lab. Basically it was well known in China's tech circles that Huawei has been preparing for some kind of "end-of-the-world" scenario in the past decade. All people see now is that Huawei still can't make its own SoC, but SoC is far from the only problem, if not for these preparation work that started 10+ years ago, Huawei would be dead long before the US has to resort to pressuring TSMC with ridiculous x% tech restrictions. I can tell you that no other high-tech hardware companies in this world, including US companies, could do as well as Huawei did under the same level of restrictions.

Yet you should know that another important element of Huawei's culture is openness. You don't shut your door and hide in the basement once you realize there are many enemies out there, NO, the more enemies there are, the more friends you'd need to go out and make. This has been a key to Huawei's success. That's why you'd sometimes hear Ren stresses the importance of learning from foreign companies and don't let nationalism or China-US animosity cloud your judgment. This doesn't mean he blindly worships western companies, just that it is even more important to stay open and make connections during the worse attacks. In fact this is not just about Huawei, do you see the resemblance to what China is doing now?

This also has nothing to do with ignoring domestic companies. I think it's quite silly to accuse Huawei of refusing to support the domestic IC industry, when Huawei smartphones contain the highest percentage of domestic components. Huawei is the biggest consumer of high-end BoE screens, it is the exclusive consumer to SMIC's 14nm node and a major customer to the 28nm node, it is one of the few (or the only? haven't check) brand that included Omnivision camera sensor on flagship-tier smartphones, and its de-Americanization effort has pushed an entire group of Chinese suppliers making single processing components (e.g. power amplifiers, antenna switches, filters, low noise amplifiers etc.) to rapidly expand and upgrade their technology. Not to mention Hisilicon itself is by far the biggest IC design company in China. Each and every one of these is a boost to IC manufacturing in China even if some high end chips are still fabricated outside. The mere existence of a high end domestic IC supply chain will naturally necessitate and nutriate the development of high end domestic IC manufacturing, which is fabrication+probing+packaging. By the way, Huawei also invested in SMIC research back in 2015, it is a shareholder of the SMIC Advanced Technology R&D (Shanghai) Corporation.

As for why Huawei didn't engage in chip manufacturing of its own, assuming the aim is to circumvent US sanction on its high end SoC, then it means building its own TSMC+ASML+LAM/Applied Materials etc.. This is the job for a country, not the job for any single company. Even for countries, this would be out of reach for 99% of the countries on this world.

superdog said:

Besides the EUV light source, you need a different lens system to adapt to the new wavelength, and a better wafer stage system (not directly related to EUV, but due to the increased precision requirements of sub-7nm nodes). These are basically the 3 core modules of a lithography machine.

Click to expand...

Ok so the two machines have some similararities in design. So it's not like starting from complete scratch to design an EUV machine. I read some of Harbin Institute's research on EUV, it looks like they have been researching this technology for a while, even before Trump was in office. So hopefully the development cycle can be faster.