Does IHS (integrated HeatSpread) on Tualatin/P4 - Hinder or help Overclocking?

[Peter007]

Law of "ThermoFlow" in Physics dictates that:

1. Greater Surface area dissipate more heat.

from this point-of-view, a Integrated Heat Spreader (IHS) like the P4/ Tualatin is a good idea since it spread the tiny core surface to a larger area for better cooling.

However, the same thermodynamic also dictates that.

2. Direct Contact with the Heat source alway dissipate more heat than indirect contact.

from this point-of-view, (IHS) is a bad idea.

____________________________________________
Now all that ThermoDynamic stuff I quoted are just Theory.

What is the REAL WORLD Opinion and Experiences?

Does IHS hinder or help overclocking and can you share your experience why it is better or worst???

Also, I read people ripping of the IHS on the Tualatin, but noone has done it on the P4..........Why not?

 

[MrNatural]

Well, I have removed the IHS off a couple of my P4s and find they cool alot better without it than they do with it. About 2-3c cooler. Guess that puts the IHS in the indirect heat dispersment area huh?

Add on. The removal of the IHS isn't that much of an improvement to allow for higher OCing but it will help cool the unit down a little more.

 

[jazztrumpet216]

Yes it does, but only partially for the reason you stated... the thermal interface compound Intel uses is absolute garbage. If they used ASIII it wouldn't be so bad, but they use some cheap generic stuff. I removed my IHS from my PIII (it is one of the newer Coppermine-T processors with the IHS) and it lopped about 3C off the load temps. Maybe someday I'll get adventurous and get some Arctic Alumina adhesive and try remounting it to see what it does, but for now I like it right where it is.

 

[MrNatural]

Yes it does, but only partially for the reason you stated... the thermal interface compound Intel uses is absolute garbage. If they used ASIII it wouldn't be so bad, but they use some cheap generic stuff. I removed my IHS from my PIII (it is one of the newer Coppermine-T processors with the IHS) and it lopped about 3C off the load temps. Maybe someday I'll get adventurous and get some Arctic Alumina adhesive and try remounting it to see what it does, but for now I like it right where it is.

I already tried that with my P4 and it really didn't help. The IHS simply put more distance between the core and the HSF which means it took longer for the heat to be disapated thus the temps went back up just like if I had not even removed the IHS to begain with. It wasn't as bad but it was pretty close to the same.

 

[deez]

I also got a 3C temp drop with my cele with IHS removed although I was not able to OC any higher

The P4 uses a different type of mounting system so depending on your heatsink as MrNatural said it could actually go up a bit

 

[ShiFtY2001]

you might get a couple of degrees less, but its hardly worth voiding your warranty for imho. but then again, you might be more 1337 than me!

rig: cele 1.0A @ 1.4, Abit VH6-T, 512MB RAM, Geforce 2 Ti, s3trio 64 (lol)

 

[deez]

As soon as you overclock you void your warranty on your CPU and yes i know it cant be detected but if you burn out your CPU by overclocking then its your responsibility to buy a new one

RMA's are for factory defects

 

[TASOS]

I 've also seen a 3C difference @full load when i removed it
on my 1.0A Tually....also using AS II

 

[Rio71]

I 've also seen a 3C difference @full load when i removed it
on my 1.0A Tually....also using AS II

i agree 3°C - 4°C good for default o'ced Speed
my Shim is made of Plastic and Tape like ol' man ... THX

 

[Peter007]

Thanks Everyone for thier input !

It seem that 100% recommendation is the IHS removal.

So much for the heat dissipation over much greater surface area theory.

 

[soil]

IIRC there are air space between the IHS and the die. I suspect the IHS is a device to reduce the effects of supercooling such as TEC and/or water-cooling.

Theory: Intel knows that P4/Tualatin can be o/c much higher if Vcore are increased and there are good heat dissipation and people will not buy those CPU of higher clocks. So to fight o/c Intel added a copper plate over the die but allows a small air space between it. The air space reduce the heat transfer effectiveness at higher temperature.

 

[deez]

I wasn't able to overclock any higher with the IHS removed from my cele but temps did drop about 3C. Actually I haven't heard from anyone that removing the IHS allows for a better OC just lower temps.

 

[soil]

I wasn't able to overclock any higher with the IHS removed from my cele but temps did drop about 3C. Actually I haven't heard from anyone that removing the IHS allows for a better OC just lower temps.

but did you increase the Vcore when you o/c?

 

[deez]

My vcore was already at 1.775 and increasing it to 1.825 did not improve stability at higher FSB's even with the IHS off. My system will run Windows as high as 150 FSB but not stable. Highest stable OC is 145 FSB

 

[MrNatural]

IIRC there are air space between the IHS and the die. I suspect the IHS is a device to reduce the effects of supercooling such as TEC and/or water-cooling.

Theory: Intel knows that P4/Tualatin can be o/c much higher if Vcore are increased and there are good heat dissipation and people will not buy those CPU of higher clocks. So to fight o/c Intel added a copper plate over the die but allows a small air space between it. The air space reduce the heat transfer effectiveness at higher temperature.

Where the hell did you come up with that one. There is no air space between the die and the IHS. Also Intel doesn't use copper on it's IHS. Some of the things ppl toss out. Geez. the core contacts the IHS by way of a blob of compound. The IHS is attached to the surface by way of a heat activated sealant and a vacume. That's what that small hole is for in the IHS. It so that a vacume can be produce between the IHS and the CPU body so the sealant get even contact all the way around and to also allow the IHS to have a solind contact with the core. If there was the slightest area of space in between the core and the IHS then there would not be enough heat transfer from the core to preven thermaling.

So again I will repeat myself. There is no gap of air between the core and IHS and there is no copper in the IHS.

 

[soil]

So again I will repeat myself. There is no gap of air between the core and IHS and there is no copper in the IHS.

You mean the IHS is touching the die, are you sure?

Are you also sure that the IHS has no copper? What is it made of then?

 

[soil]

My vcore was already at 1.775 and increasing it to 1.825 did not improve stability at higher FSB's even with the IHS off. My system will run Windows as high as 150 FSB but not stable. Highest stable OC is 145 FSB

so if you increase the Vcore beyond 1.825V, remove the IHS, use TEC and/or watercool, are you sure it can't run any faster?

 

[MrNatural]

You mean the IHS is touching the die, are you sure?

Are you also sure that the IHS has no copper? What is it made of then?

I have removed the IHS on a number of P4s and I can tell you that the die touches the IHS with a blop of compund between the 2.

The IHS is an aluminum/steel aloy. No copper used. Just pick one up and look at it. Geez. Do you know the difference between copper and aluminum?

 

[soil]

I have removed the IHS on a number of P4s and I can tell you that the die touches the IHS with a blop of compund between the 2.

The IHS is an aluminum/steel aloy. No copper used. Just pick one up and look at it. Geez. Do you know the difference between copper and aluminum?

After you have removed the IHS how do you know that IHS is touching the die?

For the IHS I saw messages saying it's made of nickel plated copper. Can you scratch the IHS a bit?

 

[Yodums]

Is plastic a good material for a shim? I would use some other kind of squishy thing like AMD Pads, because if the plastic is a little high, then it can be bent easily like the AMD pads.

Yodums