Removing the processor heatspreader for lower temps

All about them.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
Devilsown
Posts: 23
Joined: Sat Mar 06, 2004 2:24 am

Removing the processor heatspreader for lower temps

Post by Devilsown » Wed Dec 28, 2005 2:51 pm

I was suprised that there's little written here about removing the heatspreader of a processor.

I've read on several OC forums that people are doing it and i've even seen a movie of it. It seems pretty simple, all you need is a sharp knife and a screwdriver.

So why not do it for extreme cooling?
On average it brings down your processor temps by 10 degrees. This sounds pretty good! I reckon this way fanless cooling should be quite ok (if you're not overclocking and maybe even unclocking).

Of course it's not without risks and i was wondering if any people here have experimented with it...

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Wed Dec 28, 2005 3:42 pm

I was suprised that there's little written here about removing the heatspreader of a processor.

I've read on several OC forums that people are doing it and i've even seen a movie of it. It seems pretty simple, all you need is a sharp knife and a screwdriver.
My CPU didn't come with a heatspreader, but my guess is that it would be pretty hard to exert enough force on the chip to separate it from the heatspreader without doing some serious damage to the pins and/or processor itself.

Elixer
Posts: 520
Joined: Wed Jun 02, 2004 10:31 am
Location: Las Cruces, NM
Contact:

Post by Elixer » Wed Dec 28, 2005 4:19 pm

I don't know where you get 10C, articles about it I've seen showed 2-3C drop in temps. Doesn't seem worth risking damage to the processor besides voiding the processor warrenty. Also very few heatsinks will fit properly with the heatspreader removed without modification. I think most of us see it as too much effort and risk for small benefits. I do know though that I've seen someone on here who's done it with a couple of athlon 64s. You might try out the forum search some more if you haven't already.

frostedflakes
Posts: 1608
Joined: Tue Jan 04, 2005 4:02 pm
Location: United States

Post by frostedflakes » Wed Dec 28, 2005 9:56 pm

Keep in mind the 10*C average drop mainly to overclockers who's processors are probably pumping out 2x the heat than at stock settings. At stock, where the IHS poses less resistance to heat transfer, you'll probably only see a few degrees difference.

I had an old 2800+ Newcastle I pulled the IHS off and I think it dropped temps ~2*C on stock cooling and not OC'd.

IMHO not really worth it.

Devilsown
Posts: 23
Joined: Sat Mar 06, 2004 2:24 am

Post by Devilsown » Thu Dec 29, 2005 1:33 am

frostedflakes wrote:Keep in mind the 10*C average drop mainly to overclockers who's processors are probably pumping out 2x the heat than at stock settings. At stock, where the IHS poses less resistance to heat transfer, you'll probably only see a few degrees difference.

I had an old 2800+ Newcastle I pulled the IHS off and I think it dropped temps ~2*C on stock cooling and not OC'd.

IMHO not really worth it.
Now that's some valuable information, because i was wondering about this. Then i guess it's only interesting if you're into extreme overclocking.

jaganath: on respected Dutch OC forum, www.tweakers.net. It is however very easy to do without damaging the processor. But seeing what frostedflakes wrote, i won't risk it.

Erssa
Posts: 1421
Joined: Sat Mar 12, 2005 9:26 pm
Location: Finland

Post by Erssa » Thu Dec 29, 2005 2:53 am

frostedflakes wrote:Keep in mind the 10*C average drop mainly to overclockers who's processors are probably pumping out 2x the heat than at stock settings. At stock, where the IHS poses less resistance to heat transfer, you'll probably only see a few degrees difference.

I had an old 2800+ Newcastle I pulled the IHS off and I think it dropped temps ~2*C on stock cooling and not OC'd.

IMHO not really worth it.
I read xtremesystems forums regularly and I have never seen anyone talking about temp drops of 10c or even anything close to it. Usual claims are between 2-5c. Anyway for them it might be worth it, since they are doing it for the benchmarks 2c drop can be worth some points, or it can make a difference in running through prime95 at certain clocks. I have to agree with frostedflakes, it's not worth it. 2c temp difference will hardly make a difference. When you have set your fans running at 5-7v and you are happy with the sound profile of your computer, you can wonder if it is really worth removing the heatsink just to drop 2c off your cpu, whe it doesn't affect the sound levels. I'd rather take chances with 2c higher temps, and have my warranty.

ronrem
Posts: 1066
Joined: Sun Jan 16, 2005 2:59 am
Location: Santa Cruz

Post by ronrem » Thu Dec 29, 2005 5:29 am

Athlon XPs had no heatspreader-and there was always worry about the right shim,and not chipping the proc. a "naked" cpu would be kind of scary to put on a big Ninja. So you'd tend to go with a smaller,less effective HS and the end result-probably you run a bit HOTTER

Aris
Posts: 2299
Joined: Mon Dec 15, 2003 10:29 am
Location: Bellevue, Nebraska
Contact:

Post by Aris » Thu Dec 29, 2005 6:34 am

1. it voids your warrenty
2. its not completely fool proof, ive seen some people try it and it ripped the top layer of their core clean off.
3. the gain is very little.


i remember the Tbirds with no heat spreader, and i remember everyone complaining about how fragile they were, and how easy it was to crack the core, and how amd should use a heat spreader like intel does so we didnt have this huge hassle.

the fact is the majority of people WANT a heat spreader, even if it costs them a couple of degree's. just so they dont have to worry about making a mistake that costs them hundreds of dollars.

Slaugh
Posts: 774
Joined: Thu Dec 02, 2004 2:27 am
Location: Quebec, Canada

Post by Slaugh » Sat Dec 31, 2005 6:20 am

According to this article from Overclocking Masters (translated), removing the heatspreader has little effect at stock voltages. It barely lowers your CPU temperature by 4°C at full load. Removing it will automatically void the warranty and there's a chance that you ruin the CPU in the process. I don't recommend it to anybody!

Judging by this graphic, the only big difference is at 1.75V for an Opteron. So you need to overclock your CPU to the max to really see a difference...

ronrem
Posts: 1066
Joined: Sun Jan 16, 2005 2:59 am
Location: Santa Cruz

Post by ronrem » Wed Jan 04, 2006 7:34 pm

Risk vs Advantage ratio...not good

HaloJones
Posts: 109
Joined: Mon Nov 10, 2003 9:09 am
Location: London, England

Post by HaloJones » Thu Jan 05, 2006 3:44 am

I removed the IHS from my Opteron 146. It is very easy to do with just a razor blade as you're only cutting through some sealant. I have heard of some cases where the TIM between the die and the IHS was very poor and thus big temp drops were achieved. In my case, the TIM seemed fine and no temp difference was achieved. I did get a slightly better overclock and since I use screw-down waterblocks rather than heatsinks, there was little possibility of core damage.

YMMV

Post Reply