Removing the IHS...no brainer?

I remember reading a while ago, like a year ago about how removing the heat spreader didn't improve heat transfer that much, and in some instances, made it worse.

I can't remember where I put that link, but the test was conducted by that guy who made his own CPU die simulator, and the results were inconclusive.

I'm wondering if it's still controversial, or if removing the heatspreader from, say, an Opteron, is just a given. Any thoughts? Thanks guys!

Good reading here and here.

Well even if that guy made some very controlled tests it all depends on AMD doesn't it. He may have had a few chips with IHS that were put on very well.

Look tons of guys have popped theirs and seen some very convincing results, and lot's of guys have also trashed their chips or seen minimal OC gains after. Usually these last guys get better temps even if their OC doesn't do much.

I think with X2's for example, it is easy to say if the IHS is done badly when you have higher than usual temps and one core is doing much better than the other. I think the best answer is to see how a given chip does. If it's good then your probably fine, one thats doing poorly is another story.

Macaholic, the second link you gave me was the article I was referring to, thanks!

El<(')>Maxi, I see your point... I guess it just depends on the chip. I remember reading a post a while back (again, I don't have the link), about a guy who removed his Intel IHS to reveal thermal goo splattered all over the place. Clearly, that chip was put together very sloppily, and probalby hindered performance. I guess there's no way to know for sure how well a heatspreader is applied, and taking it off to have a peek pretty much renders it useless.

I guess there's really no way of knowing until I try it, but I guess I was just wondering if recently it became a fact that removing the heatspreader on recent A64s and Opterons is clearly a better choice than leaving it on or not.

didnt read the links, but I've never seen anyone with a worse OC or higher temps after removing the IHS, unless they trashed their CPU in the process. With the X2s, it makes a huge difference, for others, it depends on the chip.

Hmmm, I see... I guess I'll have to think about it... Maybe I'll test it with the IHS and if I'm experiencing heat problems I'll remove it or something.

Anyone know what bonds the IHS to the chip? Is it a gummy sort of adhesive, or is it a hardened resin? Is it possible to remove the remaining black residue that we always see after taking off the IHS without damaging the chip?

GV2NIX said:

Anyone know what bonds the IHS to the chip? Is it a gummy sort of adhesive, or is it a hardened resin? Is it possible to remove the remaining black residue that we always see after taking off the IHS without damaging the chip?

Click to expand...

Anyone?

its really risky removing the black gummy stuffyou might scratch too deep and get circuits and stuff.

It's RTV

It's not really hard to remove the RTV on the A64's...I don't have an X2 to try it on, but I'm sure it's similar.
You need to take your time, and trim a little at a time, let the tool do the work, not your muscles. Most people just use a single edged razor or an Exacto knife.
My A64 Winnie was so cupped in the center that it was being held up by the corners of the IHS (with a major air gap over the [email protected]" or .25mm). Not the best for heat disappation. That's why AMD puts that thick (floormat) TIM pad on there, it will brigde the gap, and suffice under "normal" use.
If you want optimum heat transfer for OC'ing then you may need to lap the IHS, or remove it if it's not fairly flat. You won't be able to tell how well it's making contact under the IHS though without running some tests, or removing it. Mine had almost no TIM on the core..(it was all on the IHS) lol. So, even with a well lapped HSF fan it wasn't working properly. With the IHS lapped it got somewhat better, but with the spreader removed and by making some spacers the proper height from non-conductive G-10 glass to shim the corners it improved much more.

Yea I want to remove my IHS on my x2 4400 but just don have the cash if I screw it up, My temps witha a iwaki md20 rlt, a g5, and a bong are reported as 35-40C idle, and I have seen as high as 69 load at only 1.5 volts. It makes me wonder if I have ****ty Ihs to Core contact or not. I think it is my mother board a lower the voltage really has no effect on temps, but mine will not go above 2.55-2.6 ghz so that suggests it could be ****ty contact (all on a epox 9npa+). I know it is not a problem with my cooling system as every part has been swapped and the G5 Never gets warm. Oh, my 7800gtx that is maxed out never goes over 33c with maze 3. I either have bad contact or a bad Mobo. Should i atempt to remove it?

CCUABIDExORxDIE said:

its really risky removing the black gummy stuffyou might scratch too deep and get circuits and stuff.

Click to expand...

Its not too risky. Just gotta take your time. I know it's easier with the P4 Northwood 12-caps. They have a VERY SMALL gap in the IHS's bonding adhesive. You stick the end of the blade in there (you don't have to cut anything to get it there), and then just gently slide it around the chip. The tool will cut through very easily.

You don't even need to remove the black gummy stuff. It is extremely low.

Yeah, I know I don't have to remove the black stuff, but I was just wondering how difficult is it is to do so since I wanted to make the chip look as clean as possible, haha. I know it doesn't matter whether I remove it or not.

Bhugarov, what kinds of tests do you think I should run to determine if there is good contact between the IHS TIM and the core? Just monitor temps and see whether they just keep rising even though I have adequate cooling?

Well, I just removed it and I am convinced that my motherboard just does not read temps accurate at all as it did not change anything there i am stiill sgowing load all the way up to 69 c lol on a bong witha storm g5 and a iwaki md20rlt, and voltage does not do a thing for temps. I am going to overclock now.

I am at 2650 and before the highest Could get was about 2.5-2.57. That is just my first try, will be back later.

Is the IHS held in place by a soft adhesive, or is it a hardened resin?

GV2NIX said:

Is the IHS held in place by a soft adhesive, or is it a hardened resin?

Click to expand...

It is just RTV silcone.

Is there any household product that dissolves RTV silicone rubber?