Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/exclusive-look-inside-sk-hynixs-720-billion-ai-memory-expansion.25701/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/EmailDomainReplace] => 1000010
            [SemiWiki/HtmlMailer] => 1000200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2031270
            [XFI] => 1060170
        )

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

Exclusive Look Inside SK Hynix's $720 Billion AI Memory Expansion

soAsian

Well-known member
South Korean chip giant SK Hynix, which makes the majority of the world's high-bandwidth memory chips for artificial intelligence, is undergoing the world’s largest memory fab buildout, betting $720 billion to triple capacity by 2034. This secretive memory king is also the hottest new entrant to the U.S. market, following a blockbuster Nasdaq debut in July.

SK Hynix, whose market cap has jumped more than fivefold in the past year and now tops $1 trillion, is making a big bet that extreme demand from AI is here to stay. Soaring memory prices exacerbated by capacity constraints are changing the fabric of the classic boom-and-bust industry, with AI chip giants like Nvidia willing to take on more risk as they scramble to secure supply.

This is a never before seen look inside the highly secured company from the largest factory being built in South Korea to an inside look at a cleanroom where memory is made, to SK Hynix’s first ever U.S. facility coming in Indiana.


massive incentives from US and Indiana but cutting edge still stay in South Korea and to breakout memory cycle and to be more profitable. contract is now 5 years instead of 1 year.

 
Tripling wafer output by 2034.. that's about a compound rate of 15% (1.15X) over 8 years.

That seems .. not insane - there are lots of other possible / future in-demand uses for memory besides AI LLMs.

The Korean foundires also have the advantage (over Micron) of less restrictions in selling memory to the Chinese.
 
Tripling wafer output by 2034.. that's about a compound rate of 15% (1.15X) over 8 years.

That seems .. not insane - there are lots of other possible / future in-demand uses for memory besides AI LLMs.

The Korean foundires also have the advantage (over Micron) of less restrictions in selling memory to the Chinese.

Providing it has less than 25% US IP contained within its manufacture.
 
Back
Top