Morris Chang - China is 5 years behind. TSMC's real competitor is South Korea

[soAsian]

TSMC Founder: Intel's Fab Push Is Ironic, Taiwan Engineers More Devoted Than Americans

The Taiwan Semiconductor Manufacturing Company's (TSMC) founder Mr. Morris Chang has shared his views on the company's current environment.

wccftech.com

Morris Chang full speech:

(Chinese)

I watch his full speech. I gotta say, Chang is quite accurate in pointing out his competitors. Chang said China is at least 5 years behind TSMC and 1 to 2 years behind on IC design compare to US/Taiwan. The cost of running a foundry in US is much higher than Taiwan and the real competitor to TSMC/Taiwan is South Korea because both countries is very similar from engineers to management.

He take couple jabs at Intel. Intel decline his proposal on investing in TSMC and poop-poop on foundry business model. He said it's ironic that Intel is coming to the foundry model with IDM 2.0. Chang talk about the history of semis industry from Fairchild all the way to Intel domination in semis from late 1980s until recently. TSMC have surpass Intel for the first time in 2021.

INTC current market cap: 258.10B USD

TSMC current market cap: 561.66B USD

I respect Morris Chang for what he've done for the semis industry.

 

[prime007]

I would politely disagree with Dr. Chang that Taiwanese engineers are more devoted than American engineers. However, there's no question that Taiwanese salaries are significantly lower than American salaries. I actually can't think of any job where a Taiwanese citizen would be paid more than his American counterpart. This is one of main reasons why many Taiwanese work abroad.

I found this article snippet rather interesting...I'm not sure if its actually accurate (Source: https://technosports.co.in/2021/04/...expects-to-equip-u-s-plant-in-september-2022/)

Hantang chairman Mr. Chen Chaoshui further elaborated on the cost saying that building a plant in the U.S. is expensive when compared to costs in Tawain. According to the executive, “If the United States builds one factory, Taiwan can build three or four factories.” Previously, Taiwan Power Semiconductor Manufacturing Co.’s Huang Chongren commented in the aftermath of Intel Corporation’s $20 billion Arizona investment that six Taiwanese engineers can be paid by the salary of a single American engineer.

 

[soAsian]

I would politely disagree with Dr. Chang that Taiwanese engineers are more devoted than American engineers. However, there's no question that Taiwanese salaries are significantly lower than American salaries. I actually can't think of any job where a Taiwanese citizen would be paid more than his American counterpart. This is one of main reasons why many Taiwanese work abroad.

I found this article snippet rather interesting...I'm not sure if its actually accurate (Source: https://technosports.co.in/2021/04/...expects-to-equip-u-s-plant-in-september-2022/)

No doubt the cost of labors is higher in the U.S.

 

[IanD]

No doubt the cost of labors is higher in the U.S.

Indeed, but there aren't many people in a gigafab, wage costs are negligible compared to equipment costs (same everywhere), building costs (much cheaper in Taiwan), water/power costs (cheaper/subsidised for fabs in Taiwan), and financing costs (ditto). Many of these cost differences are down to government policy (fab support as a strategic industry), the same applies in Korea.

 

[soAsian]

One thing that is surprising is Chang think China is 5 years behind. that's pretty short, isn't it? Does that mean China poaching is working?

 

[john32979]

I would politely disagree with Dr. Chang that Taiwanese engineers are more devoted than American engineers. However, there's no question that Taiwanese salaries are significantly lower than American salaries. I actually can't think of any job where a Taiwanese citizen would be paid more than his American counterpart. This is one of main reasons why many Taiwanese work abroad.

I found this article snippet rather interesting...I'm not sure if its actually accurate (Source: https://technosports.co.in/2021/04/...expects-to-equip-u-s-plant-in-september-2022/)

There are still some vestiges of Japanese influence in Taiwan where jumping ship from good or prestigious companies every 2-3 years is frowned upon. Rightly/wrongly, there's a sentiment in Asia that the American labor market is more adversarial and mercenarily in nature.

 

[Daniel Nenni]

Indeed, but there aren't many people in a gigafab, wage costs are negligible compared to equipment costs (same everywhere), building costs (much cheaper in Taiwan), water/power costs (cheaper/subsidised for fabs in Taiwan), and financing costs (ditto). Many of these cost differences are down to government policy (fab support as a strategic industry), the same applies in Korea.

Very true, inside the fab you no longer see people but there are many more outside of the fab than ever before.

I have been in fabs all over the world and have spent a lot of time in Taiwan and I see a big cultural difference in the workplace. TSMC employees are considered Rock Stars in Taiwan. Much like Intel employees here in Silicon Valley many many years ago. And yes TSMC Taiwan salaries are lower than the US but the performance bonuses put them easily on par with the world.

The bottom line: TSMC has developed a formula of foundry success over the past 30+ years and it is now impossible to replicate. Unless Elon Musk get's into the foundry business I so no threat from anyone to the TSMC business model.

 

[Jert]

Taiwan/TSMC over the last 30 years did an amazing and fantastic job in keeping a heavy lid on the cost of engineers in Taiwan even as the company and technology complexities both have exploded. Today even mainland China engineers cost more than Taiwan sometimes by a significant amount (and mainland China suppose to be the land of the cheap...). I guess that really worked well for TSMC but I am not so sure if it worked equally well for the majority of TSMC salaried workers who may have seen somewhat stagnated wages. Dr. Chang is a legend, no doubt about it. But so are Noyce, Moore and Grove.

 

[hist78]

I would politely disagree with Dr. Chang that Taiwanese engineers are more devoted than American engineers. However, there's no question that Taiwanese salaries are significantly lower than American salaries. I actually can't think of any job where a Taiwanese citizen would be paid more than his American counterpart. This is one of main reasons why many Taiwanese work abroad.

I found this article snippet rather interesting...I'm not sure if its actually accurate (Source: https://technosports.co.in/2021/04/...expects-to-equip-u-s-plant-in-september-2022/)

I got chance to watch the full length of Dr. Chang's speech today. If we put his words in the context of the whole speech, I believe what Dr. Chang meant is that Taiwanese engineers, technicians, and production line workers are more willing to work in the "manufacturing" business, in comparison to current US situation.

Honestly, in US it's the reality that many jobs in the manufacturing sector are less glamour, less income, and less quick rewards than jobs at Google, Amazon, hedge funds, financial investment, or business consulting.

 

[Daniel Nenni]

I got chance to watch the full length of Dr. Chang's speech today. If we put his words in the context of the whole speech, I believe what Dr. Chang meant is that Taiwanese engineers, technicians, and production line workers are more willing to work in the "manufacturing" business, in comparison to current US situation.

Honestly, in US it's the reality that many jobs in the manufacturing sector are less glamour, less income, and less quick rewards than jobs at Google, Amazon, hedge funds, financial investment, or business consulting.

True, and this is compounded by the work at home trend that manufacturing does not readily support. So it is only going to get worse in the USA because we love working at home, absolutely.

 

[Barnsley]

I got chance to watch the full length of Dr. Chang's speech today. If we put his words in the context of the whole speech, I believe what Dr. Chang meant is that Taiwanese engineers, technicians, and production line workers are more willing to work in the "manufacturing" business, in comparison to current US situation.

Honestly, in US it's the reality that many jobs in the manufacturing sector are less glamour, less income, and less quick rewards than jobs at Google, Amazon, hedge funds, financial investment, or business consulting.

On a microscale this can be seen in Singapore , where there is a large semicon prescence for its size , however the majority of folk in the FAB cleanroom are foreigners , outside the FAB it may be 50/50 at best.
Long hours and constant meetings through the day put off local graduates from wanting to go into the industry.

 

[IanD]

I got chance to watch the full length of Dr. Chang's speech today. If we put his words in the context of the whole speech, I believe what Dr. Chang meant is that Taiwanese engineers, technicians, and production line workers are more willing to work in the "manufacturing" business, in comparison to current US situation.

Honestly, in US it's the reality that many jobs in the manufacturing sector are less glamour, less income, and less quick rewards than jobs at Google, Amazon, hedge funds, financial investment, or business consulting.

Same in the UK except without Google and Amazon... :-(

 

[rtrauben]

South Korea then is then the next Chinese takeover target after Hong Kong.

 

[Maxim]

Very true, inside the fab you no longer see people but there are many more outside of the fab than ever before.

I have been in fabs all over the world and have spent a lot of time in Taiwan and I see a big cultural difference in the workplace. TSMC employees are considered Rock Stars in Taiwan. Much like Intel employees here in Silicon Valley many many years ago. And yes TSMC Taiwan salaries are lower than the US but the performance bonuses put them easily on par with the world.

The bottom line: TSMC has developed a formula of foundry success over the past 30+ years and it is now impossible to replicate. Unless Elon Musk get's into the foundry business I so no threat from anyone to the TSMC business model.

Daniel, you did no finish your comparison (of cultural differences).
While engineers/employees are considered rock stars in Taiwan - how are they viewed in the US, in your opinion?
As expendables, as small replaceable cogs in a big machine?

 

[tooLongInEDA]

Agree with the comments here that foundry does seem to find a better "cultural fit" in Taiwan and South Korea than the US (in particular) and Europe. Foundry seems (from the outside) like a very "collective" activity that operates on a large scale with large teams and few obvious stars. It might well be that the only reason that silicon fabs started in the US (and well within living memory centered in Silicon Valley) was that a) this was then new and exciting and b) no one else was doing it. Remember driving down 101 30 years ago and seeing the billboards advertising components ?

The US has - rightly or wrongly (I think rightly) - moved on as it always does to more exciting, leading edge stuff which suits the culture. The Valley moved from largely hardware to largely software/apps a long time ago.

I'm not sure I'm totally convinced about this sudden need to "re-shore" wafer fabs. The IDMs have been globally spread with global supply chains even when they were largely vertically integrated and all the vulnerabilities that entails (Kobe earthquake etc) for decades. They've actually kept going through the Covid interruption surprisingly well.

Whereas getting a new bicycle or new bike parts has become far more difficult !

Really not sure we should be beating ourselves up here when it is advanced semis and software that made mass homeworking during the pandemic possible. TSMC - and many others - deserve a medal, not criticism. I haven't noticed them obviously price-gouging while this has been going on.

Too many commentators with no knowledge these days. Semis is arguably more stable and less cyclical than in the past. I don't seem to hear the dreaded word "allocation" as much as I used to.

 

[Maxim]

Agree with the comments here that foundry does seem to find a better "cultural fit" in Taiwan and South Korea than the US (in particular) and Europe. Foundry seems (from the outside) like a very "collective" activity that operates on a large scale with large teams and few obvious stars. It might well be that the only reason that silicon fabs started in the US (and well within living memory centered in Silicon Valley) was that a) this was then new and exciting and b) no one else was doing it. Remember driving down 101 30 years ago and seeing the billboards advertising components ?

The US has - rightly or wrongly (I think rightly) - moved on as it always does to more exciting, leading edge stuff which suits the culture. The Valley moved from largely hardware to largely software/apps a long time ago.

I'm not sure I'm totally convinced about this sudden need to "re-shore" wafer fabs. The IDMs have been globally spread with global supply chains even when they were largely vertically integrated and all the vulnerabilities that entails (Kobe earthquake etc) for decades. They've actually kept going through the Covid interruption surprisingly well.

Whereas getting a new bicycle or new bike parts has become far more difficult !

Really not sure we should be beating ourselves up here when it is advanced semis and software that made mass homeworking during the pandemic possible. TSMC - and many others - deserve a medal, not criticism. I haven't noticed them obviously price-gouging while this has been going on.

Too many commentators with no knowledge these days. Semis is arguably more stable and less cyclical than in the past. I don't seem to hear the dreaded word "allocation" as much as I used to.

An interesting question is why Japan is not in this group (any more) of cultures that are favorable for semiconductor manufacturing.

 

[tooLongInEDA]

An interesting question is why Japan is not in this group (any more) of cultures that are favorable for semiconductor manufacturing.

I thought about that. You're right - it seems to fall into that group. The thing is - Japan and Europe have both fallen by the wayside in semis over the last 20 years - and Japan more so than Europe. Where is Japan heading in semis right now ? I guess it's arguable that Japan is doing reasonably in terms of semi manufacturing output ratiod to population. But they aren't obviously a strong player in foundry any more (those still awesomely strong in several other industries).

Perhaps this is going to be an industry like aircraft (Boing, Airbus) or aircraft engines (GE, Rolls-Royce) that can only support a very few huge global players ? So the number of successful countries might be capped by that.

 

[Staf_Verhaegen]

An interesting question is why Japan is not in this group (any more) of cultures that are favorable for semiconductor manufacturing.

I think main point is that it is too hierarchically structured and not enough initiative left at the lower level.

 

[Maxim]

I think main point is that it is too hierarchically structured and not enough initiative left at the lower level.

I think that's true, but the question is - why was that not a problem in 1960-80s, and became a problem in 1990-s and later?

Also - is there less hierarchy and seniority in Chinese / Korean / Taiwanese cultures vs Japanese?

 

[count]

Perhaps this is going to be an industry like aircraft (Boing, Airbus) or aircraft engines (GE, Rolls-Royce) that can only support a very few huge global players ? So the number of successful countries might be capped by that.

This is my view. There is room for at most three large foundries, and my guess is it'll be down to two once we get to 3nm.