Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/samsung-and-tsmc-seeking-to-spend-their-way-to-worldwide-domination-of-advanced-ic-technology.13902/page-2
        )

    [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] => 2021270
            [XFI] => 1050270
        )

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

Samsung and TSMC Seeking to Spend Their Way to Worldwide Domination of Advanced IC Technology

IanD

Active member
Sapphire Rapids does have 4 CPU dies (no I/O die) butted together, but they're still pretty big, around 350mm2 each going by the delidded photo here:


With 56 cores total (14 per chiplet) this is about 1400mm2 total which is 25mm2 of 10nm per core. AMD Milan has something around 640mm2 of CPUs split over 8 x 8core chiplets (80mm of 7nm each, 10mm2 per core) plus a 420mm2 14nm I/O die (7mm2 per core), which is a lot cheaper per mm2 -- probably about half that of 7nm. So total silicon area for AMD is about 25% less than for Intel -- and for 64 cores, not 56, so 35% less per core, 40% of which is cheaper 14nm.

Add this onto the lower yield for Intel (CPU die are 350mm2 in lower-yielding Intel 10nm compared to 80mm2 in higher-yielding TSMC 7nm), and this suggest at least a 2x die cost advantage for AMD.

Intel can't easily deal with this by using more smaller CPU chiplets, their inter-die I/O is very short (low latency) on the edges that butt together, this only works for a 2x2 array -- they've basically sawn one enormous CPU chip into 4 big pieces. To use more smaller chiplets they would need a complete re-architect with longer higher-latency links between chips, in other works a proper inter-CPU network -- probably via an I/O die like AMD.

In other words, to compete with AMD at TSMC (or even in their own fab) they need an AMD-style architecture, not an Intel-style one...
 

prime007

Active member
This article from Nikkei Asia just adds (a Japanese perspective) to what's been said...TSMC is pulling ahead.

Of particular note was a comment by a Japanese executive regarding Intel:
Intel and Washington underestimated TSMC and Dutch peer ASML, allowing the two to pull ahead in terms of technical capabilities over the past several years, according to an executive at a Japanese chipmaking equipment manufacturer that has long done business with both.

"It's at a level where [intel=] can't catch up," the executive said.
[/intel]
[intel=][/intel]
 

prime007

Active member
Adding another case to the "Samsung can't yield" narrative...

Asus recently had a investor conference call. In it, they stated that the reason it hasn't sold as many graphics cards in the first three months of the year was because of a drop in supply from Nvidia. What this means is that the current shortage of Nvidia's 30-series GPUs isn't going to get any better anytime soon....but also an increase in graphics card pricing.

Asus' statement:
"Our guess is that the gap might have been caused by lower yields upstream [potentially indicating Nvidia's own production at Samsung]"

Source:
 
Top