TSMC could drive Arizona's economy — let's not miss our moment

[Daniel Nenni]

Opinion: TSMC is about a lot more for Arizona and America than we realize. Can we work in good faith to solve our differences amicably?​

Editorial board
Arizona Republic

The machines of the modern world are exquisitely small. They are labyrinths of multilayered transistors and electric circuits that in miniature rival the detail and design of sprawling metropolises.

To see these new devices in all their complexity — in their billions of components — you must put them under microscope.
Whisper thin and virtually weightless, these semiconductors and microchips are marvels that control and operate fighter jets and computers, cellphones and automobiles — virtually all of our modern electronics.

They are drivers of the present and the future and could become one of the most significant drivers of the Arizona economy if we don’t miss our moment.

TSMC could drive Arizona's economy — let's not miss our moment

Opinion: TSMC is about a lot more for Arizona and America than we realize. Can we work in good faith to solve our differences amicably?

www.azcentral.com

 

[Daniel Nenni]

Personally I think the "moment" has been missed. These rags dragged TSMC through the mud with all of the union nonsense I doubt TSMC and others will favor AZ for future projects.

Besides, what if China attacks AZ? We need to regionalize and build in New Mexico and Texas to reduce the risk of semiconductor supply chain disruption.

 

[blueone]

Personally I think the "moment" has been missed. These rags dragged TSMC through the mud with all of the union nonsense I doubt TSMC and others will favor AZ for future projects.

Besides, what if China attacks AZ? We need to regionalize and build in New Mexico and Texas to reduce the risk of semiconductor supply chain disruption.

New Mexico would make a lot of sense for TSMC, since they followed Intel to Phoenix (and in Morris Chang's mind, Oregon), they could also follow Intel to Rio Rancho (near Albuquerque). Intel just invested an additional $3.5B there, so it's probably high time for TSMC to get on the New Mexico bus in case China attacks the Valley of the Sun.

 

[nghanayem]

New Mexico would make a lot of sense for TSMC, since they followed Intel to Phoenix (and in Morris Chang's mind, Oregon), they could also follow Intel to Rio Rancho (near Albuquerque). Intel just invested an additional $3.5B there, so it's probably high time for TSMC to get on the New Mexico bus in case China attacks the Valley of the Sun.

I think to really be secure they as Dan said need to follow intel to the canned Ft. Worth Texas Fab. Building a fab in Sherman next to good ol' TI would also be an essential step.

 

[benb]

TSMC in AZ is a fab shell at this point, with equipment installation underway. If you do a good job installing tools, the fab yield will reflect this.

It’s pretty much 100% on people to do a good job. That’s it. Not labor politics, geopolitics, just the intelligence and capabilities of a person, usually a man, to not *#$# it up. Would it be better if more women were installers? Probably. Would it be better if more of the installers were from Taiwan or China? Or Germany? Or Japan? Or any country that takes industry more seriously than the US does?

I’ll let you all answer that last question.

 

[Daniel Nenni]

TSMC in AZ is a fab shell at this point, with equipment installation underway. If you do a good job installing tools, the fab yield will reflect this.

It’s pretty much 100% on people to do a good job. That’s it. Not labor politics, geopolitics, just the intelligence and capabilities of a person, usually a man, to not *#$# it up. Would it be better if more women were installers? Probably. Would it be better if more of the installers were from Taiwan or China? Or Germany? Or Japan? Or any country that takes industry more seriously than the US does?

I’ll let you all answer that last question.

I'm doing research on the TSMC fab in Camas Washington which has always been viewed as a stepchild (my observation). It is a very successful fab and has done quite well over the years (I know people that worked there). Yet Morris Chang has thrown shade their way on many occasions. Why do you think that is? I think that may answer the question, or maybe not.

 

[nghanayem]

I'm doing research on the TSMC fab in Camas Washington which has always been viewed as a stepchild (my observation). It is a very successful fab and has done quite well over the years (I know people that worked there). Yet Morris Chang has thrown shade their way on many occasions. Why do you think that is? I think that may answer the question, or maybe not.

My guess is he is just an old man stuck in his ways, and when he was making his bones in the industry he worked at TI in the time of rot in the US semi industry (and I suppose US manufacturing in general) when the Japanese were doing just about everything better (except for I guess maybe logic design). If my understanding of the timeline is correct he went back to TW before you saw the US firms adopt LEAN and six sigma manufacturing principles (I think Gaussian quality curves with lower std dev were also a big deal since at the time US auto firms were doing skewed quality distributions that tried to maximize the number of units at the ideal quality at the cost of greater line variability). The 90s was also when you saw folks like intel/Micron catch up on process tech and had IBM path the "copper brick road" for the industry.

 

[blueone]

I'm doing research on the TSMC fab in Camas Washington which has always been viewed as a stepchild (my observation). It is a very successful fab and has done quite well over the years (I know people that worked there). Yet Morris Chang has thrown shade their way on many occasions. Why do you think that is? I think that may answer the question, or maybe not.

I've been to Camas several times and I've seen the WaferTech facility. (Actually, the facility is even outside Camas, near Mill Plain.) It has always been a mystery to me why TSMC chose that location. It looks like WaferTech was placed to fail.

I'm looking forward to what you find out.

 

[Daniel Nenni]

My guess is he is just an old man stuck in his ways, and when he was making his bones in the industry he worked at TI in the time of rot in the US semi industry (and I suppose US manufacturing in general) when the Japanese were doing just about everything better (except for I guess maybe logic design). If my understanding of the timeline is correct he went back to TW before you saw the US firms adopt LEAN and six sigma manufacturing principles (I think Gaussian quality curves with lower std dev were also a big deal since at the time US auto firms were doing skewed quality distributions that tried to maximize the number of units at the ideal quality at the cost of greater line variability). The 90s was also when you saw folks like intel/Micron catch up on process tech and had IBM path the "copper brick road" for the industry.

A friend who was at TI during the Morris Chang era told me that TI tried to be the jack of all trades but was the master of none. This is when they went into the consumer electronics market and lost focus in semiconductors. Now TI seems to be the master of analog? TI is one of the all time great semiconductor companies, absolutely.

 

[delong.height]

Personally I think the "moment" has been missed. These rags dragged TSMC through the mud with all of the union nonsense I doubt TSMC and others will favor AZ for future projects.

I think TSMC came to US with an attitude that views American workers as being expensive and not diligent enough to Taiwanese employees. That's true. But if you come to US, you got to have the spirit that it shall be treated as a long-term project, you should want to expand in the US, instead of relying on Taiwanese employees that are being brought to US. If you don't want to expand AZ project, maybe it's better to expand elsewhere? how about Singapore? How about Thailand, they are a bit far from China for sure. So why come to US?

Besides, what if China attacks AZ? We need to regionalize and build in New Mexico and Texas to reduce the risk of semiconductor supply chain disruption.

I would rather believe China to unite Taiwan by 2030. Not a guess, but a high probable thing to happen

 

[Jozo035]

I think TSMC came to US with an attitude that views American workers as being expensive and not diligent enough to Taiwanese employees. That's true. But if you come to US, you got to have the spirit that it shall be treated as a long-term project, you should want to expand in the US, instead of relying on Taiwanese employees that are being brought to US. If you don't want to expand AZ project, maybe it's better to expand elsewhere? how about Singapore? How about Thailand, they are a bit far from China for sure. So why come to US?

I would rather believe China to unite Taiwan by 2030. Not a guess, but a high probable thing to happen

Well, You can't just open fab and expect that there is thousands of idle engineers waiting to take any position... TSMC needs to provide adequate training/education or pay higher salaries to attract professionals from other sites. They will face same difficulties in Singapore or Thailand or europe. Even in China, They have/had training programs and at the same time very good salaries specially in trier n districts.

As for "unification": Try considering one apolitical reason. How will standard of living improve after joining mainland?

 

[delong.height]

Well, You can't just open fab and expect that there is thousands of idle engineers waiting to take any position... TSMC needs to provide adequate training/education or pay higher salaries to attract professionals from other sites. They will face same difficulties in Singapore or Thailand or europe. Even in China, They have/had training programs and at the same time very good salaries specially in trier n districts.

Yes, that is not expected for any foreign companies.

As for "unification": Try considering one apolitical reason. How will standard of living improve after joining mainland?

Come on, we are not Taiwanese, why shall we care about how their standard of living are going to improve/decline after they joined the mainland. But one thing for sure, they can go to mainland at ease, and so does people from mainland. Electricity may not be an issue as well.

And China knows that if they unite Taiwan by 2025 or any time before 2030, the west will lose control of 90% of leading edge and majority of overall semiconductor manufacturing capabilities immediately, which is an economic nuclear bomb they'll be able to possess. Having a war with US, it may not be a problem if that is taken into account.

 

[hist78]

I'm doing research on the TSMC fab in Camas Washington which has always been viewed as a stepchild (my observation). It is a very successful fab and has done quite well over the years (I know people that worked there). Yet Morris Chang has thrown shade their way on many occasions. Why do you think that is? I think that may answer the question, or maybe not.

TSMC fab in Camas Washington eventually became profitable but it took a lot of time, efforts, and many top engineers dispatched from Taiwan to achieve it (or to salvage it, from TSMC's point of view).

I won't call it a nightmare experience for TSMC but one thing TSMC had learned from it is that it must tightly manage the whole construction and operation process from day one. Whenever it's necessary, TSMC doesn't hesitate to bring in global resources from its own staffs and/or from TSMC's suppliers/partners to help out. At the end of day, any cost overrun, bad quality, or production delay will hurt TSMC, Apple, Qualcomm, AMD, Nvidia or even Intel greatly.

Nowadays people might understand that to ask a UAW worker who has experience in GM or Chrysler to teach Toyota how to build a quality car probably won't work very well.

But now there are a group of HVAC and sheetmetal union workers who believe they know how to build a fab. They believe they are entitled to it. Instead of collaborate and partnership, they chose to badmouth or blackmail TSMC.

What those union leaders did is exactly going against the reputation and economic opportunities they truly need.

Can we blame TSMC for coming to US and encountering the very same reasons that have forced countless American's manufacturing moved to foreign countries?

 

[blueone]

TSMC fab in Camas Washington eventually became profitable but it took a lot of time, efforts, and many top engineers dispatched from Taiwan to achieve it (or to salvage it, from TSMC's point of view).

I won't call it a nightmare experience for TSMC but one thing TSMC had learned from it is that it must tightly manage the whole construction and operation process from day one. Whenever it's necessary, TSMC doesn't hesitate to bring in global resources from its own staffs and/or from TSMC's suppliers/partners to help out. At the end of day, any cost overrun, bad quality, or production delay will hurt TSMC, Apple, Qualcomm, AMD, Nvidia or even Intel greatly.

How do you know these details? Also, WaferTech does not produce the chips that most interest Apple, Qualcomm, AMD, Intel, and Nvidia. 35nm to 16nm processes probably have different design customers.

Nowadays people might understand that to ask a UAW worker who has experience in GM or Chrysler to teach Toyota how to build a quality car probably won't work very well.

Assembly line workers don't teach other automakers how to build cars. Also, due to numerous rental car experiences (I was President's Circle level at Hertz for five years) I've driven many new US and Japanese cars, including Toyotas. There isn't a significant perceptible quality difference. If anything, I preferred a Chevy Malibu over any Toyota vehicle I ever got. (All of our personal vehicles are from German manufacturers, and have been for about 20 years, so I'm not defending a personal purchase decision.)

But now there are a group of HVAC and sheetmetal union workers who believe they know how to build a fab. They believe they are entitled to it. Instead of collaborate and partnership, they chose to badmouth or blackmail TSMC.

What those union leaders did is exactly going against the reputation and economic opportunities they truly need.

Union leaders only care about the union's power. I was once a member of a national-level US union, and it was a disgusting experience. Nothing they have said or done in AZ to TSMC surprises me at all.

Can we blame TSMC for coming to US and encountering the very same reasons that have forced countless American's manufacturing moved to foreign countries?

American manufacturing moved overseas to lower production costs, not because of similar problems to what TSMC has had in AZ. Also, TSMC's executive rhetoric has not helped their cause at all. Union leaders cannot let precedents to be set where they are perceived as reasonable and practical. They obviously like to be perceived as tough and powerful.

 

[hist78]

How do you know these details? Also, WaferTech does not produce the chips that most interest Apple, Qualcomm, AMD, Intel, and Nvidia. 35nm to 16nm processes probably have different design customers.

Assembly line workers don't teach other automakers how to build cars. Also, due to numerous rental car experiences (I was President's Circle level at Hertz for five years) I've driven many new US and Japanese cars, including Toyotas. There isn't a significant perceptible quality difference. If anything, I preferred a Chevy Malibu over any Toyota vehicle I ever got. (All of our personal vehicles are from German manufacturers, and have been for about 20 years, so I'm not defending a personal purchase decision.)

Union leaders only care about the union's power. I was once a member of a national-level US union, and it was a disgusting experience. Nothing they have said or done in AZ to TSMC surprises me at all.

American manufacturing moved overseas to lower production costs, not because of similar problems to what TSMC has had in AZ. Also, TSMC's executive rhetoric has not helped their cause at all. Union leaders cannot let precedents to be set where they are perceived as reasonable and practical. They obviously like to be perceived as tough and powerful.

"How do you know these details? Also, WaferTech does not produce the chips that most interest Apple, Qualcomm, AMD, Intel, and Nvidia. 35nm to 16nm processes probably have different design customers."

1. Because I talked to and/or listened to those were involved in saving TSMC fab in Camas Washington (WafeTech).

2. It's true that WafeTech is not making the leading edge products. TSMC didn't choose Camas to expand it into more advanced process manufacturing. Among may reasons, one obvious one is that do they really want to get burned twice when there are many other better locations in the world?

 

[blueone]

1. Because I talked to and/or listened to those were involved in saving TSMC fab in Camas Washington (WafeTech).

Pardon my skepticism and curiosity but...

Why do Intel and Micron fabs never need "saving" in the US? I admit to being skeptical of Intel choosing central Ohio for a gigafab complex, but I'm not in the mood to bet against them.

Did your sources tell you why TSMC chose an exceptionally inconvenient and seemingly inappropriate location for the WaferTech facility?

 

[hist78]

Pardon my skepticism and curiosity but...

Why do Intel and Micron fabs never need "saving" in the US? I admit to being skeptical of Intel choosing central Ohio for a gigafab complex, but I'm not in the mood to bet against them.

Did your sources tell you why TSMC chose an exceptionally inconvenient and seemingly inappropriate location for the WaferTech facility?

Please let Pat Gelsinger know the good news that Intel doesn't need him and his rescue plan. Now he can cash in his stock options and happily retire.