Has anyone removed Conroe's heatspreader?

RichardJSampson

Distinguished
Aug 23, 2001
135
0
18,680
Or is this just an incredibly asinine thing to do? Seems to me like the spreader could be a weak link between the CPU and the sink. Does the fact that the C2D have a heatspreader make high-end thermal compounds like AS5 more negligible in the cooling setup?

Yeah, I'm pretty much asking stupid questions at this point. The rig you see in my signature is not yet built as I'm still waiting for all the components to arrive. Thus, I wanted to get my stupid questions out of the way first before I destroy my brand new system.

Thanks friends!
 
Yep, there is tasty cheese underneath. Just don´t eat it all at once.




Fun aside, some people have removed the spreader to improve cooling, yes. You have to remember that the heat spreader is not only there for cooling but also for protection. I remember the era of amds copper spacers. They didn´t use a heat spreader and you could (or had) to put a copper piece of metal around your cpu die so the pressure of the cooler didn´t break anything. Still, every other moron botched it and fragged the CPU. Ah, good times... :)
 
People have removed Conroe's heatspreader; try Googling "Removing a Core 2 Duo IHS" or something like that. It's a rather extreme measure to improve cooling. One article I saw put the temperature drop achieved from removing the Conroe IHS at about 3C max. On the flipside, that same article reported that there was a 75% chance of destroying the processor by doing so.

😳 The naked Conroe:
intel_ihs_removal_4.jpg
 
I thought with the recent Intel CPUs they *soldered* the die to the IHS, meaning heat transfer was almost perfect and you were almost sure to destroy the die removing the IHS?

On AMDs and older s478 Intels it can be worth it, but on a Core 2...?
 
Intel has soldered the dies and heatspreaders together since the introduction of Prescott. I believe the method of removing the IHS requires one to heat the solder enough for it to start melting, though it doesn't take long for the solder to resolidify. In other words, apply heat, then pry off the IHS as fast but as cleanly as possible. It takes some skill and coordination, but it seems like it's really more luck than anything.
 
Other than the thermal glue which you can remove with a razor, there's also the soldered part in which you need some heat to removed it. As I've read in the review once, they managed to removed the IHS but at the end the cpu was fried, due to the heat applied to removed the solder. You'd be better off getting a $150 water cooler than do this measure to improve cooling.
 
Other than the thermal glue which you can remove with a razor, there's also the soldered part in which you need some heat to removed it. As I've read in the review once, they managed to removed the IHS but at the end the cpu was fried, due to the heat applied to removed the solder. You'd be better off getting a $150 water cooler than do this measure to improve cooling.

Looks like I'll be keeping the heatspreader on! While we're on the subject of cooling, here's the solution I've decided to implement. I'll have an Amtec 900 case. (http://www.newegg.com/Product/Product.asp?Item=N82E16811129021) This thing has 4 fans, including a 200mm chimney fan so that should help. I will also be using a Thermaltake Large Typhoon (http://www.newegg.com/Product/Product.asp?Item=N82E16835106061) -- I sure as hell hope it fits in the case!

One thing I've noticed about the HSF I bought is that the surface is less not tremendously smooth. Should it be lapped first? Do you guys think this cooling solution should get me at least 3.5GHz (it's an e6700) at full load and pass Prime95?

Thanks for that fascinating post, Mex!