Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/tsmc-to-build-fab-in-arizona-and-they-are-hiring.13209/
        )

    [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 to Build Fab in Arizona and They are Hiring!

Daniel Nenni

Admin
Staff member
From TSMC:

We’re delighted to catch up with you that TSMC had announced its intention to build and operate an advanced semiconductor fab in Arizona. This U.S. advanced foundry fab not only enables us to better support our customers and partners, we also wish to attract global talents to work with us to change the world.At TSMC, we are working consistently to provide the most advanced technologies to enrich human life. Join us to initiate and witness the new semiconductor era with remarkable people around the world. https://lnkd.in/gX-aEre

TSMC Career Opporunitites.jpg
 
Last edited:
I’m wondering if TSMC and Intel are in on this together. Intel becomes a big TSMC customer if TSMC builds fabs in AZ?
 
The problem in the us is the lack of development of engineers and stem in general that are the tip of the spear. US isn't the only country that has this problem.

If tsmc has the same problem in arizona that Intel does than the problem is bigger than Intel.
 
The problem in the us is the lack of development of engineers and stem in general that are the tip of the spear. US isn't the only country that has this problem.

If tsmc has the same problem in arizona that Intel does than the problem is bigger than Intel.
I believed recently WFM gives lots of hints about manufacturing inflection-work offshore. We might refer to the operation model of tsmc's China Fab16(16/12nm FinFET) as a reference point. Look forward to seeing the change soon.
 
I’m wondering if TSMC and Intel are in on this together. Intel becomes a big TSMC customer if TSMC builds fabs in AZ?
I do believe tsmc and Samsung are in the final short list of foundry vendor but intel needs to decide their foundry long term strategy first. It is tricky in competing with AMD and leveraging between Samsung and tsmc.
 
I highly doubt Intel will chose Samsung. AMD really is a threat to Intel and it would be a much wiser decision for Intel to partner with TSMC to better compete with AMD.

I remember when Altera and Xilinx were battling head-to-head. Altera was partnered with TSMC and Xilinx with UMC. Xilinx actually had a dedicated floor in the UMC Hsinchu HQ. At 40nm UMC fell behind TSMC so Xilinx jumped to TSMC and actually beat Altera to 28nm. The rest is history but Xilinx beat Altera to first silicon from that day forward. Intel would have to do the same and the only way to do that is to partner with TSMC, my opinion.

I do believe tsmc and Samsung are in the final short list of foundry vendor but intel needs to decide their foundry long term strategy first. It is tricky in competing with AMD and leveraging between Samsung and tsmc.
 
I highly doubt Intel will chose Samsung. AMD really is a threat to Intel and it would be a much wiser decision for Intel to partner with TSMC to better compete with AMD.

I remember when Altera and Xilinx were battling head-to-head. Altera was partnered with TSMC and Xilinx with UMC. Xilinx actually had a dedicated floor in the UMC Hsinchu HQ. At 40nm UMC fell behind TSMC so Xilinx jumped to TSMC and actually beat Altera to 28nm. The rest is history but Xilinx beat Altera to first silicon from that day forward. Intel would have to do the same and the only way to do that is to partner with TSMC, my opinion.
My opinion, if intel wants to squeeze AMD in market share, definitely he needs order capacity in tsmc in this "zero-sum and total capacity fixed " game. To meet more demand shortage, he can expand internal capacity and leverage foundries like Samsung or Globalfoundries(14nm only now). It takes time for preparation.
Regarding your example of Xilinx, I remembered it was long term-exclusive foundry contract with UMC, similar to AMD and Globalfoundries contract. When contract expired(?) Xilinx jumped out of UMC's wagon and into tsmc's. Heard Altera was very angry and then selected intel as 14nm foundry. Then we all know the following story.
 
The timeline fits. Intel looking at outsourcing 5nm in 2022. TSMC building a 5nm fab in Arizona where Intel has a large presence, based on lead times and the types of positions they are hiring for now you'd expect the TSMC Arizona fab to be online in 2022.
 
The timeline fits. Intel looking at outsourcing 5nm in 2022. TSMC building a 5nm fab in Arizona where Intel has a large presence, based on lead times and the types of positions they are hiring for now you'd expect the TSMC Arizona fab to be online in 2022.

In my estimation Intel will start with TSMC 3nm which is better aligned with Intel's 7nm process with chips available in 2023. AMD already has TSMC 5nm chips so I see no need for Intel to start behind AMD.
 
It won't be just about node size anymore. It will be if semiconductors are modded with a layer of diamond etc.
 
In my estimation Intel will start with TSMC 3nm which is better aligned with Intel's 7nm process with chips available in 2023. AMD already has TSMC 5nm chips so I see no need for Intel to start behind AMD.

AMD 5nm chips will not be 'buyable' before late '21 / early '22, will they?

Would it be possible to elaborate on how TSMC 3nm is / isn't equal to Intel 7nm considering design rules?
 
The TSMC 3nm PDKs are available now so why would Intel start at 5nm in 2021? AMD has 5nm production chips now but yes HVM starts in 2021. The point being if Intel is just starting designs at 5nm they will be a process behind AMD. And as I mentioned before porting existing Intel designs to TSMC is a fools errand. TSMC PDKs must be used to get chips competitive with AMD.

TSMC 5/3nm is still FinFET as is Intel 7nm. Based on Scott's numbers (which are estimated) I could see Intel designing to both 7nm to TSMC 3nm without too much problem. Certainly better than Samsung 3nm (due to device type):

1604432042510.png


AMD 5nm chips will not be 'buyable' before late '21 / early '22, will they?

Would it be possible to elaborate on how TSMC 3nm is / isn't equal to Intel 7nm considering design rules?
 
Do You think this fab will be build if Biden wins? What about US support towards TSMC (Taiwan) as Biden's priorities are completely different than Trump's?
 
Do You think this fab will be build if Biden wins? What about US support towards TSMC (Taiwan) as Biden's priorities are completely different than Trump's?

I'm not sure there are significant difference between Republicans and Democrats regarding the policy towards Taiwan, especially on TSMC.

Using the proposed "The American Foundries Act Of 2020" as an example, the sponsors are heavy weight senators from both parties. I believe Federal's subsidy for the TSMC Arizona project will be coming from this pool of money.

 
Do we need to politicize everything including this forum?

I fear that:

1- the CHIPS act is too small and too late
2- It does not address fundamental problems with manufacturing competitiveness in USA
3- The overwhelming # of fabs to be built in the next 3 years will be in Taiwan and China (around 19)
 
Do we need to politicize everything including this forum?

I fear that:

1- the CHIPS act is too small and too late
2- It does not address fundamental problems with manufacturing competitiveness in USA
3- The overwhelming # of fabs to be built in the next 3 years will be in Taiwan and China (around 19)

1. The initial production capacity is small but there's nothing to stop TSMC adding more capacity in the future.
2. With a smaller initial capacity, TSMC is managing the expectation and can have more means and ways to make it profitable with a high utilization.
3. It's true that it does not address the fundamental problems with manufacturing competitiveness in USA. But it can take at least 2 ~ 3 years to pass and implement a bigger and more ambitious "Made in USA" plan. I think by then it will be too late for the US semiconductor industry.
 
Back
Top