Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/where-does-the-semiwiki-community-think-the-chips-act-money-should-go-and-why.16466/
        )

    [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
)

Where does the SemiWiki community think the Chips Act money should go and why?

Arthur Hanson

Well-known member
As an investor, I will have a vested interest in the answers. Also, where should it definitely not go? And why on both, Thanks
 
CHIPS Act money (corporate welfare) should go to Micron to better compete in memory. Only 2 percent of global memory supply is manufactured in the US, and all of that is produced by Micron. Without memory there is no logic. Micron is battling Korean giants and needs our help.

GF and Skywater are lost causes. Intel, Samsung, and TSMC do not need it. The chip shortage is over and for those of you who think building fabs in the US will solve our geopolitical semiconductor problems you are wrong. It takes the whole world to build chips. Will US consumers pay more for US based chips? Not likely.
 
  • I would imagine that US government will be considering several factors:
  • * Type of FAB/tech (Daniel suggested memory FABs, personally I have no idea what government/country priorities are).
  • * The funds for FABs should be matching private/company investments (i.e. the company investing more should get more). R&D funds shouldn't necessarily follow this rule.
  • * How advanced is the tech/FAB the funds will be invested into.
  • * Company domicile
  • * Companies doing R&D in US might get higher priority (say, if the company has a FAB in US but R&D is done abroad, such FAB does less for advancement of US semicoductor industry compared to the case where both R&D and FABs are in US)
 
CHIPS Act money (corporate welfare) should go to Micron to better compete in memory. Only 2 percent of global memory supply is manufactured in the US, and all of that is produced by Micron. Without memory there is no logic. Micron is battling Korean giants and needs our help.

GF and Skywater are lost causes. Intel, Samsung, and TSMC do not need it. The chip shortage is over and for those of you who think building fabs in the US will solve our geopolitical semiconductor problems you are wrong. It takes the whole world to build chips. Will US consumers pay more for US based chips? Not likely.

I read the act text, and there is zero clue to whom money will actually go.

Here is the showstopper:
▪ $19 billion in FY22, including the $2 billion legacy chip production funding
▪ $5 billion each year, FY23 through FY26

We understand that $19B is around the figure US promised TSMC, but there is no concrete regulations I found stating that.

Devil is in details.
 
CHIPS Act money (corporate welfare) should go to Micron to better compete in memory. Only 2 percent of global memory supply is manufactured in the US, and all of that is produced by Micron. Without memory there is no logic. Micron is battling Korean giants and needs our help.

GF and Skywater are lost causes. Intel, Samsung, and TSMC do not need it. The chip shortage is over and for those of you who think building fabs in the US will solve our geopolitical semiconductor problems you are wrong. It takes the whole world to build chips. Will US consumers pay more for US based chips? Not likely.
Well done article by Nikkei on the complexity of the global semiconductor industry supply chain. It absolutely takes the whole world to build chips.

 
With all of this discussion about the CHIPS ACT, there is absolutely no focus on Assembly, Packaging and Test. If the US is serious about building an ecosystem in the US, then we need these elements back in the US as well. There are probably some towns in the Midwest where the cost of labor is low enough that it might make some sense to do this.
 
With all of this discussion about the CHIPS ACT, there is absolutely no focus on Assembly, Packaging and Test. If the US is serious about building an ecosystem in the US, then we need these elements back in the US as well. There are probably some towns in the Midwest where the cost of labor is low enough that it might make some sense to do this.
Exactly. Intel getting those money should build Packaging and Test in US.
 
With all of this discussion about the CHIPS ACT, there is absolutely no focus on Assembly, Packaging and Test. If the US is serious about building an ecosystem in the US, then we need these elements back in the US as well. There are probably some towns in the Midwest where the cost of labor is low enough that it might make some sense to do this.

I doubt the chance to survive for a large scale packaging and testing operations in the US for consumer semiconductor products. In the semiconductor food chain, the packaging and testing segment typically is a labor intensive and low margin business. Using the industry leader ASE Group as an example. Its net profit margin was single digit percentage for most years except recently quarters ( increased to 10-11%) .
 
With all of this discussion about the CHIPS ACT, there is absolutely no focus on Assembly, Packaging and Test. If the US is serious about building an ecosystem in the US, then we need these elements back in the US as well. There are probably some towns in the Midwest where the cost of labor is low enough that it might make some sense to do this.

MidWest is in fact cheaper than South China for net-cost of labour, but still nowhere near Vietnam, or rural Malaysia where their existing packaging plant been running for decades.
 
Packaging is labor intensive, cost is for sure a lot higher in Midwest.
But that what subsidies are supposed to do.
To support higher labor cost instead of support lower production yield.

Packaging also create more jobs.
 
Packaging is labor intensive, cost is for sure a lot higher in Midwest.
But that what subsidies are supposed to do.
To support higher labor cost instead of support lower production yield.

Packaging also create more jobs.

Average skilled manual labourer job of that level in Kuangtung will easily hit CNY 15000-25000 + benefits and taxes

That's a minimum of $2800 per month, with average upper hitting $4700 in total cost of hiring per month.

In US, you can pay pretty much only salary for the job of this level with few taxes, and close to none social ensurance/pension/medical/maternity benefits.
 
Average skilled manual labourer job of that level in Kuangtung will easily hit CNY 15000-25000 + benefits and taxes

That's a minimum of $2800 per month, with average upper hitting $4700 in total cost of hiring per month.

In US, you can pay pretty much only salary for the job of this level with few taxes, and close to none social ensurance/pension/medical/maternity benefits.

Labor cost in Midwest is not cheap. A store manager in a suburban Chicago Burger King told me last month that he can't find workers by paying $15 an hour.

He told me he's so happy that he "only" needs to work six days a week in July, compare to June.
 
Average skilled manual labourer job of that level in Kuangtung will easily hit CNY 15000-25000 + benefits and taxes

That's a minimum of $2800 per month, with average upper hitting $4700 in total cost of hiring per month.

In US, you can pay pretty much only salary for the job of this level with few taxes, and close to none social ensurance/pension/medical/maternity benefits.
ASE in Taiwan median salary only $20,000 per year. In China it will be a slightly lower than that.
The higher salary you mentioned are for engineers. Not labors in the production lines.
 
Labor cost in Midwest is not cheap. A store manager in a suburban Chicago Burger King told me last month that he can't find workers by paying $15 an hour.

He told me he's so happy that he "only" needs to work six days a week in July, compare to June.

You can hire undocumented immigrants for cheap,it's actually very common in harvesting season
 
ASE in Taiwan median salary only $20,000 per year. In China it will be a slightly lower than that.
The higher salary you mentioned are for engineers. Not labors in the production lines.

Actual assembly line workers easily bag CNY 12000-15000 per months, which means CNY 15000-22000 in total employment costs. That for somebody with 5-7 years experience, 2-3 years of technical school, and maybe few extra courses, and certificates.

For an OSAT worker with university education, $2800 net salary will be the minimum. For line lead, or a shift head, they can net $4k-$5k. That's higher than a lot mid-career university educated engineers get outside of Silicon Valley in the USA, and is completely outside of realm of possibility for American manual labourers, and a lot of non-tech white collars.

Yes, South China is way more expensive than Taiwan, that's why there were so many Taiwanese working there.

Taiwanese enterprises fully deserve all the just criticism for screaming "Taiwan expensive!" while ignoring the glaring fact of their Chinese factories costing them more than operating in TW.
 
Last edited:
  • Like
Reactions: VCT
You can hire undocumented immigrants for cheap,it's actually very common in harvesting season

Businesses in Midwest (or all over the US) have been doing that already. Otherwise the labor hourly rate can go even higher.
 
Actual assembly line workers easily bag CNY 12000-15000 per months, which means CNY 15000-22000 in total employment costs. That for somebody with 5-7 years experience, 2-3 years of technical school, and maybe few extra courses, and certificates.

For an OSAT worker with university education, $2800 net salary will be the minimum. For line lead, or a shift head, they can net $4k-$5k. That's higher than a lot mid-career university educated engineers get outside of Silicon Valley in the USA, and is completely outside of realm of possibility for American manual labourers, and a lot of non-tech white collars.

Yes, South China is way more expensive than Taiwan, that's why there were so many Taiwanese working there.

Taiwanese enterprises fully deserve all the just criticism for screaming "Taiwan expensive!" while ignoring the glaring fact of their Chinese factories costing them more than operating in TW.
Mistery solved: Intel is going to build packaging and assembly facilities in Italy:

Intel Close to Sealing $5 Billion Chip Factories Deal in Italy
Sources say there will be both semiconductor packaging and assembly plants.

 
@VCT seeing you giving a like, I think I have some solidarity here.

Tsai's next task will be to prevail over powerful OEM electronics lobby, who are doing Beijing's bidding by running their businesses in the mainland.

People in the West all exort their virtue, while lots of these folks are nasty, and are a big part of the problem.

Few misconceptions I want to correct our Western readers:

1. Taiwanese business ran to the mainland from unions, not high labour prices

2. If they could've made a dime by overtiming workers 5 hours a day, and get away with it, as they do in the mainland, they would've done so in Taiwan as well. They just cannot bribe the govt in Taiwan as they do in the mainland (5h overtime is well illegal under PRC's laws)

3. Taiwanese govt did not let Taiwanese business do business in the mainland as any kind of a good will gesture, but because of relentless lobbying for years to let them do so

4. TW business moguls in the mainland are no victims of worsening USA-PRC relationships by any account, they been all over greasing hands, and rubbing elbows with communist officials for the last 30 years
 
Last edited:
Back
Top