What is the casing on a CPU made of?

[Arkamedies]

Hope I am asking the correct question here. I am trying to figure out what the outer section of the top of a cpu is made of.

I am specifically talking about the part that you would normally place your shim over when installing a coling device.


In this one I am actually seeing a problem with the possition of the contacts for what I want to do, but that too can be overcome.

Anyone know what material it is?

[kevin_bouchard]

i was told it was made of ceramic(dont know if this is true though)
I do know that intel pent 4 has a copper outer casing(i think amd should do the same):)

[Arkamedies]

Thanks, that would make sense... now... how to attach to liquid transfer block directly to it... any adhesive clues?

It has been done, and there are several still in use and working fine, but I have no clue what page they were on. Anyone know the links I speek of?

[Warlord2]

like direct die cooling?


people use JBweld for that

[Arkamedies]

And there is no ill effect long term that would cause leakage from such a rigid epoxy?

[kevin_bouchard]

you want to be careful about heating and cooling(contractions) it may cause cracks, so I would probably use something flexible. One of the major problems(aside of leaks) is the fact that you cant change your cooling setup as easily as before,and i cant see direct die cooling getting better results than a all copper heatsink watercooled (because of surface area). Ya i know that the copper is only in contact with the die (as the water would be) but copper coducts better heat, so that it can spread it out to the water(awoiding the bottleneck between die and water with the placement of copper)this is true unless you have a pump that pumps out alot of water.
This is only my theory and don't have any really data to defend it.
Does anyone think the same?

[Arkamedies]

Actually the heat dissipation of water is a great deal higher than copper, or any other metal as far as that goes. But you are right, there are several good reasons not to do direct cooling.

All the items you listed are big setbacks to direct cooling... perhaps I should just do the normal water block thing!

Thanks for all the help and comments guys and gals. Your a pile of gems in an otherwise dirtclod world! *grin*

[Ridenow]

Direct die water cooling was done on older processors that did not have the bridges the current AMD processors have. You would have to find some way to insulate them. As for the thermal expansion, the ceramic is not affected much and I do not think it would be a problem.

[kevin_bouchard]

Arkamedies water isnt even close to conducting heat better than copper. It is in fact better than air and that is why watercooling gets better cooling results. Waters heat conductivity is about 0.055, still are is .0003 and copper is 3.937(higher the better).
The only reason why you may think that water is more conductive than copper is because it mever contacts the die 100%(and that is where thermal grease comes,which is about 0.04).
Another thing is that copper has the tends to keep the heat,so that may also be one of the reasons.
After taking in these facts i still cant see direct die cooling being better(unless you have a older cpu like Ridenow said).
For my waterblock I am probably going to have homemade aluminum waterblock(doesnt tend to trap the heat)and a copper plate under it(to spread the heat to the entire alum block).
It is kind of the way alpha makes there heatsinks(and any other copper base heatsink)

And Ridenow I wasint talking about the ceramic die expanding and contracting but was talking about how the epoxy may do this(but now that i think about it, it shouldn't get that :mad: )


I hope this clears some stuff up and helps you on your watercooling journey :D