Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/government-to-use-dpa-to-compel-semiconductor-companies-to-share-information.14686/
        )

    [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
)

Government to use DPA to compel semiconductor companies to share information

Andy1299

Member

White House to help solve the shortage through usage of Defense Production Act to force semiconductor companies to share information in order to solve the crises

What does the community think? I can see the logical merit in checking on customers whether they are double or triple ordering but this seems to be focused on suppliers.

My opinion (that my better judgement in me suggests that I should not share) is: how will the White House (most with political or liberal arts degrees that don’t understand the complexities or lead times) help one of the most complex industries with thousands of the smartest people in world…
 
Supply and demand, when out of wack the situation will self-correct. An ideal time for TSM to increase production, but this will be based on firm orders.
 
Supply and demand, when out of wack the situation will self-correct. An ideal time for TSM to increase production, but this will be based on firm orders.

All foundries will increase production and capacity. Remember it does not take long to build a fab. It can be done in 18 months in Taiwan and Korea. In the US and other places that the government is not tightly coupled maybe 2 years but still not very long. The problem is that foundries are building much more capacity than needed. China has been buying up semiconductor fab equipment (both new and used) for the last 4-5 years. Most of which is still not in HVM but hopefully soon.

If you look at all of the fabs built and in progress we definitely have a glut coming. Which means low semiconductor prices, great for consumers and the electronics industry. Not so good if your fab and manufacturing costs are higher than the others. Advantage TSMC and Samsung for their semiconductor manufacturing cost leadership. When push comes to shove cost, not location of manufacturing, will determine the industry winners, absolutely.

 
All foundries will increase production and capacity. Remember it does not take long to build a fab. It can be done in 18 months in Taiwan and Korea. In the US and other places that the government is not tightly coupled maybe 2 years but still not very long. The problem is that foundries are building much more capacity than needed. China has been buying up semiconductor fab equipment (both new and used) for the last 4-5 years. Most of which is still not in HVM but hopefully soon.

If you look at all of the fabs built and in progress we definitely have a glut coming. Which means low semiconductor prices, great for consumers and the electronics industry. Not so good if your fab and manufacturing costs are higher than the others. Advantage TSMC and Samsung for their semiconductor manufacturing cost leadership. When push comes to shove cost, not location of manufacturing, will determine the industry winners, absolutely.

Hi Dan,
Almost all players to increase their 28nm capacity. Do you think all of them can make it in 2023? Do suppliers like Applied Materials, Lam Research and others can deliver such huge demand at the same time?
thanks
 
Hi Dan,
Almost all players to increase their 28nm capacity. Do you think all of them can make it in 2023? Do suppliers like Applied Materials, Lam Research and others can deliver such huge demand at the same time?
thanks
MCU, and automotive ASIC is not even 28nm, more like 180nm, and 350nm, and 200mm wafers. Really, really old equipment, and process.

It actually makes almost as big of a challenge than the lack of capacity itself. Nobody now makes 200mm stuff anymore. You can only buy second hand tools for 200mm, and they've been all sold out to Chinese 3rd tier fabs who seen that 200mm crunch was coming at least 5-6 years ago.
 
In fact, margins on 200mm are now almost as high as on the leading edge: 30%-50%, and if you are a fab, and scored one of those "panic tapeouts" — those were going for up to 1000%
 
Back
Top