Here's a good read with some great results. So who's going to try it?

http://inventgeek.com/Projects/DiamondG ... rview.aspx

If it's too good to be true it's probably not.

I can't see the differences being that extreme. I just can't.

Plus I don't have a system to try it on.

The science is there it would seem.

http://en.wikipedia.org/wiki/Thermal_conductivity

With synthetic being better than natural diamonds.

But what about when the paste dries out, you'll have diamond dust abrading your lungs.

Screw lung cancer I want my CPU a few degree's cooler!

very interesting, and very tempting....someone has to look into this, i wish i lived in halifax to have one of you give me a ahnd, could def be a good weekend project....

It would have been nice to see the results for various mixtures, from straight silicon grease to the final mixture ratio of diamond dust.

I think this might have more to do with the fresh silicon grease than diamond dust.

The thermal chart in a previous post provides for a cross sectional area in size based upon a solid substance.

Considering the diamond dust is suspended in the silicon grease, there cannot be 100% contact between all of the dust particles, thus a less than 100% contact between the CPU and heat sink.

The next interesting item to consider is for this diamond dust to apparently performing better than a 2000 grit lap of both the cpu and heat sink. Based upon previous experience with that CPU, getting even an idle temp at 28C in normal room temp would be outstanding and I'm sure, well beyond the physical limitations of the Artic Freezer Pro 7 to remove that sort of heat.

The final point that I'd like to make is for everyone to remember the material which the heat must travel through before it is dissipated.

From the cpu core through silver based solder to the copper based IHS with external silver coating, diamond silicon grease, then a pure copper HS of the Arctic 7 Pro cooler. Notice the copper is still the limiting factor of least thermal conductivity and it has not been removed from the system. The copper IHS can only allow so much heat to pass to whatever paste, grease, etc.

Now if you notice the Pentium D has a 130W Thermal Design Power and was overclocked, thus even more heat; and the Arctic 7 Pro has a max of 130W with a thermal resistance of 0.17°C/Watt. If one is generating more heat than which the cooler can dissipated, then we are seeing plain old magic in getting the temp even lower.

It's a shame there's no comments section on the page because I'm sure there would be lots of good questions like that for the author.

As a side question - do people still remove the IHS's from their chips anymore or have they found that it doesn't make as much of a difference compared to just lapping it? I remember that was the big thing to do when they were introduced on the Intel chips, but haven't heard much about it since.

Yeah, no point in really removing them to get maybe 1 or 2 degrees cooler than lapping. I've always had good lapping results and I've seen some bad things happen when removing IHS's.

IIRC Intel started firmly attaching the IHS to the die, resulting in some scrap silicon if you tried to remove it.

You might want to read this if you're interested in doing such things.

http://www.xtremesystems.org/forums/sho ... p?t=222040

BTW, direct water on die cooling is is less efficient than leaving the IHS in place.

I am actually working on a mod to solder a water shroud to my X6800, with the IHS acting as the bottom, then letting the water flow through, thereby eliminating the IHS/waterblock interface and the need for paste or lapping.

When I go i7 I'm getting one of these for overclocking:

http://www.ocztechnology.com/products/c ... ocz_cryo_z

U could put a hot quarter on your CPU and still keep amazingly low temps haha.

Only downside is you have to silicone your entire motherboard to prevent water damage.

You might be disappointed, the Cryo_Z's Max Stable Load is 120W, but the current I7's TDP is 130W. This might be the reason they do not advertise compatibility for the I7 and do not make a bracket for the i7?

Arriving in Sept are some new i7s and i5 with much lower TDPs but in a new socket LGA1156.

Yeah but they look so uber cool. By i7 I mean "My next major CPU upgrade" which in all likelihood would be an i7.

But still, that's a cool product, and water cooling is just a little too sketchy for me.

they are cool!

This is what I ran for while

http://www.pcstats.com/articleview.cfm?articleID=1793

I bought it for the FX-53 (939) then to a 4400 X2, then to the X6800, a E6700 and I think an E8400...and perhaps on the Q9450.

there are no doubts these phase change units get good performance, but as the die shrinks and the silicon becomes better, the increase over water/chilled water gets smaller.

Brainfart

http://www.innovationcooling.com/overview.htm

HWC had a sampling dealie going back in April

http://www.hardwarecanucks.com/forum/ov ... eaway.html

Results

http://www.hardwarecanucks.com/forum/ov ... sults.html