Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/tsmc-arizona-struggles-to-overcome-vast-differences-between-taiwanese-and-us-work-culture.20774/page-3
        )

    [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
)

TSMC Arizona struggles to overcome vast differences between Taiwanese and US work culture

Let's see if Dresden "struggles" with the TSMC build:


Japan certainly did not "struggle" like AZ did.
Lagging node with a Partner. Everything done in Az was poorly thought out and denial of reality and blind hope the TSMC Taiwan way is the way to success, hope is a very bad strategy and the cost to the Taiwanese employees is sad and inexcusable, but the leadership in Fab21 and HQ care little it seems. Seen in almost every policy and communication or the lack of honesty and transparency.
 
Lagging node with a Partner. Everything done in Az was poorly thought out and denial of reality and blind hope the TSMC Taiwan way is the way to success, hope is a very bad strategy and the cost to the Taiwanese employees is sad and inexcusable, but the leadership in Fab21 and HQ care little it seems. Seen in almost every policy and communication or the lack of honesty and transparency.

Why do you think TSMC did it?

From what I know customers pressured TSMC into the AZ build, Morris Chang and others were against it, and it cost Mark Liu his job. Personally I think globalization is a smart play for TSMC. It makes the silicon Shield all that much stronger.
 
TSMC-Latest-Production-Capacity-Layout 2024.jpg
 
Lagging node with a Partner. Everything done in Az was poorly thought out and denial of reality and blind hope the TSMC Taiwan way is the way to success, hope is a very bad strategy and the cost to the Taiwanese employees is sad and inexcusable, but the leadership in Fab21 and HQ care little it seems. Seen in almost every policy and communication or the lack of honesty and transparency.
The political motivations far outweighed the economic ones. That's what happened. AZ was the wrong location from the beginning. It's kind of like marrying someone for political reasons; it will not be smooth down the road.
 
Why do you think TSMC did it?

From what I know customers pressured TSMC into the AZ build, Morris Chang and others were against it, and it cost Mark Liu his job. Personally I think globalization is a smart play for TSMC. It makes the silicon Shield all that much stronger.
As Jim said politics, Mark retired for 2 reasons 1) succumbing to pressure and 2) the huge mismanagement of the whole thing.

TSMC will need to undergo a culture transformation as large as Intel is attempting if they hope to move leading edge manufacturing out of Taiwan, they are nowhere near with the current leadership.

If Apple and others wanted to derisk they would have pushed / forced it years ago. Fab21 even on original schedule wouldn’t derisk any customer, all politics
 
As Jim said politics, Mark retired for 2 reasons 1) succumbing to pressure and 2) the huge mismanagement of the whole thing.
TSMC will need to undergo a culture transformation as large as Intel is attempting if they hope to move leading edge manufacturing out of Taiwan, they are nowhere near with the current leadership.
If Apple and others wanted to derisk they would have pushed / forced it years ago. Fab21 even on original schedule wouldn’t derisk any customer, all politics

I do not think TSMC will ever develop leading edge technology in the US or anywhere else. It will always be in Taiwan. The fabs in Japan and Dresden are mature technologies and by the time AZ is in HVM at N5/3, N2 will be ramping up. And when they switch to HNA-EUV it will require new fabs so that will be Taiwan only for years to come.
 
I do not think TSMC will ever develop leading edge technology in the US or anywhere else. It will always be in Taiwan. The fabs in Japan and Dresden are mature technologies and by the time AZ is in HVM at N5/3, N2 will be ramping up. And when they switch to HNA-EUV it will require new fabs so that will be Taiwan only for years to come.
That would match my experience. My PCB manufacturer did all their R&D in Taiwan and we never ran anything that hadn't been developed in Taiwan first.
 
Rather than focusing on 'laziness,' another aspect people have highlighted is 'obedience.' Below is a story from 25 years ago, shared by Shang-Yi Chiang (former TSMC Co-COO) in an oral interview.
Reference: https://archive.computerhistory.org/resources/access/text/2022/07/102792671-05-01-acc.pdf (page 34)
----
Chiang: I went through one experience. I was very shocked. I share with you, it's an interesting story. In
1999, TSMC at that time tried to start this wafer fab in Oregon called WaferTech. And the performance
was very poor. Was far behind TSMC fab. Had given them all kinds of headaches. So, finally they
decided one morning the VP in charge of manufacture, he got a list of 20 people, and he called one by
one to his office. Everyone go to his office for ten minutes. And early ones didn't know, because they
never talked to this VP--

Fairbairn: Right.

Chiang: -- you know, what he wants? And then later on, these guys come out, "Oh, why did he call you
in the office?" "TSMC want to send you to Wafer Tech, for assignment for two years. You go back to start
to apply for visa, get your passport. And be ready in three weeks."

Fairbairn: And go back to Taiwan?

Chiang: From Taiwan to U.S.

Fairbairn: Oh, Taiwan to the U.S. Oh, I-- move the people from there. Okay.

Chiang: Right. So, they moved 20 people, including the fab manager, and in every area, all the
engineering area and manufacturing area, Manufacture Supervisor, QRA, sent the entire team, 20
people.

Fairbairn: Mm hm, somebody from each group.

Chiang: Right. That's the way they did it. I look at that, wow. And three weeks later, 20 people on the
same airplane to U.S. And they live in the same place. They told me every morning, these people they all
carpool, they had ten cars, just drive to work <laughs> and go home. And they fixed that fab. And then
during this time, I was shocked, because in U.S. that's not the way they do it. If you're in HP, if we have
an assignment in France and some people look at people profile, "Oh, this may be a candidate." His boss
will probe him. "Oh, we have the job there. It's very interesting. And it's very important for the company."

Fairbairn: Right. Try to convince him.

Chiang: Try to convince him.

Fairbairn: Right.

Chiang: "And just think about it," two weeks later, "Oh, have you thought about that? Well, good. Go talk
to your family and see how it works." And a month later, "Oh, it looks good," you go there to interview.
And come back, oh, everything seems to go on, and you talk about your package. Oh, later on, go with
your wife to house-hunting. They wait until summer, your kids out of school, you go there six months later.
At TSMC, it's three weeks.

Fairbairn: It's 20 people, boom.

Chiang: Twenty people.

Fairbairn: How long did they stay there in Oregon?

Chiang: Two years.
This really resonated with me. The expectation of compliance was very high in my experience.

Most of my co-workers that were unhappy either complained about the level of dedication to the company that was expected at the expense of their personal life, or disliked the dictatorial management style. Personally, I expected the demands on my time in an HVM environment and am willing to let the boss be the boss and make the final call.

That said, I also feel it is my responsibility to voice my opinion and I believe I can deliver better results if I understand the why behind the plan, but ultimately it is the bosses call. My boss didn't seem to have any issues with that, because he knew at the end of the day I would do what he asked me to do.

Most of the complaints I saw around in the articles around the AZ plant seemed to center on these two issues.
 
This really resonated with me. The expectation of compliance was very high in my experience.

Most of my co-workers that were unhappy either complained about the level of dedication to the company that was expected at the expense of their personal life, or disliked the dictatorial management style. Personally, I expected the demands on my time in an HVM environment and am willing to let the boss be the boss and make the final call.

That said, I also feel it is my responsibility to voice my opinion and I believe I can deliver better results if I understand the why behind the plan, but ultimately it is the bosses call. My boss didn't seem to have any issues with that, because he knew at the end of the day I would do what he asked me to do.

Most of the complaints I saw around in the articles around the AZ plant seemed to center on these two issues.
Imagine this: one of TSMC AZ's purposes could be to poach talent from INTC AZ and ultimately cripple it. Now, with INTC half dead already, there is little drive to pour resources for minimal return.
 
Imagine this: one of TSMC AZ's purposes could be to poach talent from INTC AZ and ultimately cripple it. Now, with INTC half dead already, there is little drive to pour resources for minimal return.
Given the local hire attrition rate doubt that. But maybe TSMC will soon have a flood of new candidates with CPM and ISP
 
I do not think TSMC will ever develop leading edge technology in the US or anywhere else. It will always be in Taiwan. The fabs in Japan and Dresden are mature technologies and by the time AZ is in HVM at N5/3, N2 will be ramping up. And when they switch to HNA-EUV it will require new fabs so that will be Taiwan only for years to come.
Agreed, development for TSMC will always be in Hsinchu and Intel in Hillsboro. The other nuance is leading edge will stay in Taiwan unlike Intel so did/does move outside of the US. No amount of political or customer pressure will bring a giga to US, Fab21 is a token press / political site. Makes for good pictures and press but the real money and volume will always stay in Taiwan till the rouge province is retake
 
Imagine this: one of TSMC AZ's purposes could be to poach talent from INTC AZ and ultimately cripple it. Now, with INTC half dead already, there is little drive to pour resources for minimal return.

The funny thing is, when TSMC's AZ plan was announced, many were pessimistic, believing that TSMC AZ would lose a lot of talent to Intel AZ due to Intel offering better salaries and a superior working environment.
 
The funny thing is, when TSMC's AZ plan was announced, many were pessimistic, believing that TSMC AZ would lose a lot of talent to Intel AZ due to Intel offering better salaries and a superior working environment.
This could well be true, but the purpose might include learning some tricks from their counterparts, which will cause deeper fear within TSMC.
TSMC has grown so fast by absorbing BKMs from companies like TI and Intel. The reverse could also happen.
 
Last edited:
If Apple and others wanted to de-risk they would have pushed / forced it years ago. Fab21 even on original schedule wouldn’t de-risk any customer, all politics
This is an inconvenient truth, Geopolitical risks aside, if the objective is fundamentally to increase the percentage of chips manufactured in the US, then TSMC's projects are steps in that direction. Small steps to be sure, but we've got to start somewhere.
 
Similar videos propaganda can be found about any other company.

There was an older video profile of several local hires, and several of the best and and promising candidates too, they no longer work there.

This video aligns with what I know personally. TSMC AZ has come a long way. TSMC is an expert at yield learning and customer learning. They are applying the same principles to employee learning. Think what you want but this is happening. TSMC AZ is doing well.
 
This video aligns with what I know personally. TSMC AZ has come a long way. TSMC is an expert at yield learning and customer learning. They are applying the same principles to employee learning. Think what you want but this is happening. TSMC AZ is doing well.
Well I hear another story from insider…
 
This is an inconvenient truth, Geopolitical risks aside, if the objective is fundamentally to increase the percentage of chips manufactured in the US, then TSMC's projects are steps in that direction. Small steps to be sure, but we've got to start somewhere.
Geopolitical risk is much lower than the Western world imagines. The Chinese have much lower territorial ambitions than the Western world. They ceded much of their land to Russia, Japan, Germany, etc., throughout history. Taiwan was ceded to Japan as a gift and returned after WWII. Therefore, they do not care as much a
Geopolitical risk is much lower than the Western world imagines. The Chinese have much lower territorial ambitions than the Western world. They ceded much of their land to Russia, Japan, Germany, etc., throughout history. Taiwan was ceded to Japan as a gift and returned after WWII. Therefore, they do not care as much about the island as you might think.
To increase chip manufacturing in the US, the first thing you need is a workforce to support it. Otherwise, you will need to import engineers and technicians from elsewhere, which might not work out well. TSMC is a result of Taiwanese engineers who were not satisfied with US fabs.
 
Back
Top