Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/dumb-question-but-why-nobody-tried-to-make-ics-on-already-pre-cut-dies.14393/
        )

    [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
)

Dumb question, but why nobody tried to make ICs on already pre-cut dies ?

Paul2

Well-known member
So, the idea that came to my head is why nobody tried to cut dies first, and then grow ICs after they are put into some kind of a carried?

I understand that alignment, warpage, and metrology now becomes a per-die problem, instead of per-wafer, but so far it doesn't seem to be impossible with very capable modern-day nanometric drives.
 
Paul, think about Pizza for a moment, why do they bake the entire pizza all at once, instead of 12 individual pieces? Answer: It’s faster, cheaper and offers greater quality control to fabricate an entire 200mm or 300mm wafer at one time, instead of 5,000 times.
 
Paul, think about Pizza for a moment, why do they bake the entire pizza all at once, instead of 12 individual pieces? Answer: It’s faster, cheaper and offers greater quality control to fabricate an entire 200mm or 300mm wafer at one time, instead of 5,000 times.

At the same time, if the biggest single piece of dough is only enough for 300mm pizza, while your oven can handle way more, the gain from pre-cutting your dough to use the most of oven capacity is very obvious too.

And you can replace, mix, test, and match pieces of dought in between the cooking stages too, to always use your oven capacity for the best, and tastiest pizza pieces only.
 
Presumably because most of the cost of silicon chips is often the depreciation cost of the expensive machinery used (until it's fully depreciated) so the throughput of raw silicon through these must be maximised. And processing as a single wafer achieves that. It must also be far easier to align die on a single wafer than 100 or more individual die.
 
Presumably because most of the cost of silicon chips is often the depreciation cost of the expensive machinery used (until it's fully depreciated) so the throughput of raw silicon through these must be maximised. And processing as a single wafer achieves that. It must also be far easier to align die on a single wafer than 100 or more individual die.

Yes, but think.

If you can put many times more dies into the machine at a single time, than as if it was on a single wafer.

Modern day wafer stages, i believe, have a lot of reserve for bigger mass.
 
Back
Top