Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/tsmc-acquires-innolux-fab-for-nt-17-14-billion-530m-usd.20820/
        )

    [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] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

TSMC acquires Innolux fab for NT$17.14 billion ($530M USD)

gc_epictetus

New member
https://www.digitimes.com/news/a20240815PD229/tsmc-innolux-5.5g-fab-acquisition.html

Where is TSMC going with this new factory? Expanding CoWoS or developing panel level fan out capabilities?

CoWoS-L may have fundamental limitations on # bridges. This & recent public news on NVIDIA B200 yield issues may just be process immaturity that will be resolved or could be a longer term issue. C.C. Wei indicated in earnings (July 18) that they are limited to retical limits of 5-6x chip size and panel-level fan out technology will give up to 10x, but that is at least three years out.

It is public that TSMC has started development on panel level fan-out, and it is aiming for large form factor AI products, but it will take at least 3 years minimum. CoWoS still should be the key packaging solution prior that and is still constrained.

The Innolux factory could give them a factory shell to build more CoWoS capacity, of they could be leveraging the panel display manufacturing equipment to build the panel level fan out technology.

If advanced packaging is evolving to be a higher yielding part of the business, should we expect others to jump on Fan Out Panel as they draft off TSMC?
 
Back
Top