die dimensions

Hello. The die sizes of the Intel Core 2 Extreme X6800 (dual-core) and AMD Athlon 64 FX-62 are listed, respectively, as 143 mm^2 and 230 mm^2 at several websites. Assuming the dies are rectangular, not square, what are the dimensions of the rectangles? Can someone take measurements (preferably with calipers) for me ASAP? Any help would be greatly appreciated. Thanks!

I'm an engineer designing a heat sink. I need to know the dimensions because they are an important factor in the thermal spreading resistance from die to IHS. AMD and Intel documents have drawings of the IHS's but not of the dies. AMD tech support says this die info is proprietary. I'm not sure where the 143 mm^2 and 230 mm^2 come from originally, but they're the most common values listed on the Web so I'll cautiously accept them. My understanding is that gamers often remove the IHS to directly mount a water block, so I thought someone might be able to help. Maybe someone out there has at least a ruler.

You can see pictures of the various cores here:

http://64.233.179.104/translate_c?h...?q=cooling+masters&num=100&hl=en&lr=&safe=off

Print it out, grab your ruler and measure.

Are you designing a heatsink or a waterblock?

I have an A64 DTR Clawhammer 3000+.
It's die size is approximately 18.5mm X 12.5mm (I'd guess that it's the 230mm² one).
Since that's the computer I'm using, It's an estimate I measured from the thermal paste track on the waterblock I just removed.

The 18.5mm dimension is perpendicular to a line connecting the two hold-down bolts.

Since it's a DTR (desktop replacement) chip, it never had a heatspreader.
Being that it's the larger of the two core dimensions you have, I'd say shoot for that and they'll all be covered.

Thanks, Graystar and Diggrr!

Why is the dual-core shown at Graystar's link much more rectangular than the QX6700 quad-core's two dual-cores shown at, for example, www.anandtech.com/cpuchipsets/showdoc.aspx?i=2866?

No prob. I would've pulled off my waterblock and given more accurate measurements, but the battery in my calipers is dead.

to the forums, M8

*edit* That's the Kentsfield core ? (dunno intel chips much)...the only constant is change.
Looks like making a taylored waterblock is gonna get much more difficult, and the heatspreader is going to be the only consistent sizing (for now).

weisse said:

Why is the dual-core shown at Graystar's link much more rectangular than the QX6700 quad-core's two dual-cores shown at, for example, www.anandtech.com/cpuchipsets/showdoc.aspx?i=2866?

Click to expand...

You'll have to ask Intel.

The die in the X6800 should be the same size as in the QX6700 because the quad-core just doubles the quantity of the dual-core die...

weisse said:

The die in the X6800 should be the same size as in the QX6700 because the quad-core just doubles the quantity of the dual-core die...

Click to expand...

LMAO!!!

www.neoseeker.com/Articles/Hardware/Reviews/core2quad_qx6700/
"The Core 2 Quad takes it one step further by placing two dual core dies into the same physical processor package - that is two of the Core 2 Duo processor dies are placed side by side and connected internally, with both chips sharing the pins on the BGA 775 package."

My understanding is, dimensionally speaking, the X6800 and QX6700 dies should be the same. Since I don't read German, I'm guessing the dual-core shown at Graystar's link isn't the Intel dual-core (X6800) that I asked about.

weisse said:

Since I don't read German, I'm guessing the dual-core shown at Graystar's link isn't the Intel dual-core (X6800) that I asked about.

Click to expand...

It's a Smithfield in the link, alias Pentium D....

All of the Core 2 Duos should have the same die size, and the quattro should be 2x's the size of that if I am believing correctly. I am fairly certain though that all Core 2 Duos will have the same die size.

~jtjuska

Sorry, I just thought this emphasis on die size and location was funny. The problem has never been getting the heat TO the heat transfer medium, be it air or water. Copper’s thermal conductivity is some 650 times that of water, and several times that over air. The problem has always been getting the heat IN the medium. That’s what radial designs such as the Storm waterblock and heatpipes are all about.

Thank you, jtjuska!

np!!
I would still try to get confirmation but it is my understanding that the E6700's, E6600's, E6400's and E6300's are basically X6800's that didn't make the grade in some way. Whether they weren't stable with a higher multiplier or in the 63 and 6400's case maybe some of the cache had problems so they disabled it. It makes sense to me and is pretty much what I have seen so far.

~jtjuska

jtjuska said:

np!!
I would still try to get confirmation but it is my understanding that the E6700's, E6600's, E6400's and E6300's are basically X6800's that didn't make the grade in some way. Whether they weren't stable with a higher multiplier or in the 63 and 6400's case maybe some of the cache had problems so they disabled it. It makes sense to me and is pretty much what I have seen so far.

~jtjuska

Click to expand...

From my understanding this is accurate with most lines of chips.