Semiwiki on LinkedIn Semiwiki on Facebook Semiwiki on Twitter Register / Log in 
Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/techanalye-assesses-china-as-3-years-behind-tsmc.20865/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

TechanaLye assesses China as 3 years behind TSMC

F

Fred Chen

Moderator
TechanaLye, a Japanese reverse engineering company, estimated SMIC to be 3 years behind TSMC in technology advancement, after comparing Huawei's Kirin 9010 (SMIC 7nm) with its Kirin 9000 (5nm TSMC). The area of SMIC's 7nm mass-produced chip is 118.4 square millimeters, while TSMC's 5nm chip is 107.8 square millimeters; less than 10% area difference, but the processing performance is basically the same.

https://zh.cn.nikkei.com/china/ccompany/56519-2024-08-27-05-00-00.html
 
Last edited:
Fred Chen said:
https://zh.cn.nikkei.com/china/ccompany/56519-2024-08-27-05-00-00.html

TechanaLye, a Japanese reverse engineering company, estimated SMIC to be 3 years behind TSMC in technology advancement, after comparing Huawei's Kirin 9010 (SMIC 7nm) with its Kirin 9000 (5nm TSMC). The area of SMIC's 7nm mass-produced chip is 118.4 square millimeters, while TSMC's 5nm chip is 107.8 square millimeters; less than 10% area difference, but the processing performance is basically the same.
Click to expand...
not surprising. CN has headcount of 1.4b. vs TW's 0.03 b.
 
Fred Chen said:
https://zh.cn.nikkei.com/china/ccompany/56519-2024-08-27-05-00-00.html

TechanaLye, a Japanese reverse engineering company, estimated SMIC to be 3 years behind TSMC in technology advancement, after comparing Huawei's Kirin 9010 (SMIC 7nm) with its Kirin 9000 (5nm TSMC). The area of SMIC's 7nm mass-produced chip is 118.4 square millimeters, while TSMC's 5nm chip is 107.8 square millimeters; less than 10% area difference, but the processing performance is basically the same.
Click to expand...

To compare Huawei's Kirin 9010 (SMIC 7nm) released in 2024 with its Kirin 9000 (5nm TSMC) released in 2021 then suggest that Huawei is three years behind TSMC is probably not logical and meaningful.

Semiconductor technology progress and manufacturing improvement are not linear. We can't just assume Huawei will be able to catch TSMC in three years for the advancement TSMC has achieved between 2021 and 2024 without considering all the factors.
 
hist78 said:
To compare Huawei's Kirin 9010 (SMIC 7nm) released in 2024 with its Kirin 9000 (5nm TSMC) released in 2021 then suggest that Huawei is three years behind TSMC is probably not logical and meaningful.

Semiconductor technology progress and manufacturing improvement are not linear. We can't just assume Huawei will be able to catch TSMC in three years for the advancement TSMC has achieved between 2021 and 2024 without considering all the factors.
Click to expand...
Agreed. TSMC had so much momentum during those 3 years enabling it to go to 3nm. Hard if not impossible to say the same for SMIC or other Chinese entity.
 
Fred Chen said:
Agreed. TSMC had so much momentum during those 3 years enabling it to go to 3nm. Hard if not impossible to say the same for SMIC or other Chinese entity.
Click to expand...
Don't forget the EUV ban and other tool restrictions. It's not an apples-to-apples comparison. In general, without any restrictions, the mainland side could catch up really fast due to the headcount advantage.
 
Fred Chen said:
Agreed. TSMC had so much momentum during those 3 years enabling it to go to 3nm. Hard if not impossible to say the same for SMIC or other Chinese entity.
Click to expand...
I think more impressive than the technical side of N3 is how well TSMC has scaled out production.
 
China will catch up eventually. Progress on the leading edge is slowing down, and China will eventually build their own EUV machines. It's a matter of when, not if.
 
Daniel Nenni said:
Intel and Samsung cannot catch TSMC, China has even less of a chance. It is not just about technology, it is about collaboration inside the ecosystem and customer learning.
Click to expand...
China's market is enormous. There is a lot of collaboration in many internal Chinese markets, look at the Chinese auto industry, and battery industry. People were caught off guard that SMIC was able to even release a 7nm chip, I wasn't. Manufacturing is a culture, and no-one does it better than China.
 
Daniel Nenni said:
Intel and Samsung cannot catch TSMC, China has even less of a chance. It is not just about technology, it is about collaboration inside the ecosystem and customer learning.
Click to expand...
The Chinese have already learned all the tricks to boost collaboration. They are more spiritual than most people here and collaborate like brothers.

Location/Time - TEF-Thanksgiving English Fellowship

www.tefshanghai.org www.tefshanghai.org

The US has an integrity issue by treating certain groups of engineers unfairly, leading to its decline.
 
count said:
China's market is enormous. There is a lot of collaboration in many internal Chinese markets, look at the Chinese auto industry, and battery industry. People were caught off guard that SMIC was able to even release a 7nm chip, I wasn't. Manufacturing is a culture, and no-one does it better than China.
Click to expand...

I spent 20 or so years working with the China semiconductor ecosystem. You can't really compare it to Silicon Valley or Taiwan. I also worked with Japan, South Korea, and Europe. I have the least amount of confidence in China in regards to semiconductor investment and innovation. So much money was said to be invested, more so than any other country or even all of Asia? I have not been to China since the pandemic but from what I can see through the ecosystem progress is very slow and fraught with corruption. Brute force progress has been made in China but if you consider the investment made it is not good at all.
 
Daniel Nenni said:
I spent 20 or so years working with the China semiconductor ecosystem. You can't really compare it to Silicon Valley or Taiwan. I also worked with Japan, South Korea, and Europe. I have the least amount of confidence in China in regards to semiconductor investment and innovation. So much money was said to be invested, more so than any other country or even all of Asia? I have not been to China since the pandemic but from what I can see through the ecosystem progress is very slow and fraught with corruption. Brute force progress has been made in China but if you consider the investment made it is not good at all.
Click to expand...
bad samples maybe, there are better samples and you missed.
tictok has the capability of challenging Meta/google.
 
DepthFirst said:
Not really. China had the rather foolish assumption that it can block foreign firms, but still compete with those foreign firms abroad. China’s mercantilist policy making has guaranteed that it will be on the unpleasant end of reciprocity.
Click to expand...
are you talking about Apple and US govt?
 
Daniel Nenni said:
I spent 20 or so years working with the China semiconductor ecosystem. You can't really compare it to Silicon Valley or Taiwan. I also worked with Japan, South Korea, and Europe. I have the least amount of confidence in China in regards to semiconductor investment and innovation. So much money was said to be invested, more so than any other country or even all of Asia? I have not been to China since the pandemic but from what I can see through the ecosystem progress is very slow and fraught with corruption. Brute force progress has been made in China but if you consider the investment made it is not good at all.
Click to expand...
I'm currently working with some Chinese companies on battery development and manufacturing so I have a different perspective. I can say at least for batteries, after seeing how advanced they are in this area, it's not surprising that the leading battery companies in the world are from China. In some ways it's similar to semi manufacturing (precision, yield learning), but in other important ways it's different (integration with EDA tools, IP libraries, ect). So maybe it's the EDA/IP side of things that will be harder for China to master. I still have a lot of confidence that they will figure it out.
 
count said:
China's market is enormous. There is a lot of collaboration in many internal Chinese markets, look at the Chinese auto industry, and battery industry. People were caught off guard that SMIC was able to even release a 7nm chip, I wasn't. Manufacturing is a culture, and no-one does it better than China.
Click to expand...
Not so sure of your thoughts. First off, people who understand semiconductors were only caught off guard until they realized the brute force way SMIC achieved 7nm. It was a way to demonstrate “progress” without being long-term viable for anything other than limited-run showpieces.

But the bigger question I have is whether the typical industrial model China uses for everything from housing, to ICE cars, to batteries, to EVs, to solar panels, to semiconductors can work for high-end chip manufacture. I thought the quote below in a discussion on the success of Chinese EVs based on a strategic focus starting in 1992 told an interesting narrative. The system works by spreading central and local government money far and wide, with development and factories essentially created by every local government with over 5M in population. That’s why China had 700 EV companies in 2017, 500 in 2019, about 300 in 2023, and only about 100 today of which only 3 are profitable. We’ll likely see 1-2 global brands come out of that.


IMG_0052.jpeg


Chinese semiconductor companies seem to use the same model. Kind of like the US VC model, but with every local government shotgunning to create their own local champion.

www.tomshardware.com

Major Chinese semiconductor company goes bankrupt — 23 others recently withdrew IPO applications

Taiwan's media warns about the resurgence of failed semiconductor projects in China.
www.tomshardware.com www.tomshardware.com
 
KevinK said:
Not so sure of your thoughts. First off, people who understand semiconductors were only caught off guard until they realized the brute force way SMIC achieved 7nm. It was a way to demonstrate “progress” without being long-term viable for anything other than limited-run showpieces.

But the bigger question I have is whether the typical industrial model China uses for everything from housing, to ICE cars, to batteries, to EVs, to solar panels, to semiconductors can work for high-end chip manufacture. I thought the quote below in a discussion on the success of Chinese EVs based on a strategic focus starting in 1992 told an interesting narrative. The system works by spreading central and local government money far and wide, with development and factories essentially created by every local government with over 5M in population. That’s why China had 700 EV companies in 2017, 500 in 2019, about 300 in 2023, and only about 100 today of which only 3 are profitable. We’ll likely see 1-2 global brands come out of that.
Click to expand...
Well, I think semi people seem to think there is something special about semi manufacturing. But auto people used to think there was something special about auto manufacturing, and before that steel makers used to think there was something special about steel manufacturing.

At the end of the day it's manufacturing. There is an increasing level of complexity involved but it's still manufacturing, and the same principals of quality, yield, flexibility, rate ect apply. And the complexity of it is something that over time you learn to manage through specialized systems, equipment, design tools, supply chain, industrial organization and so on. There is no technology that a country of 1.4 billion very smart and hard working people is not going to be able to figure out. Other companies and countries just have a long head start. There are some gaps that Daniel pointed out when it comes to collaboration and IP, but I also think China will improve in these areas.

People who think China will never be able to produce a completive semiconductor manufacturing company I think are being really arrogant.
 
You must log in or register to reply here.
Back
Top