Crunching comparison 2600K and 3770K

[D_A]

I've done it once on an AMD Opteron but that only required a fine blade and a steady hand, there was no soldering involved. I removed it and replaced the TIM (used ICDiamond) then put it all back together. It gained me a few degrees but nothing I HAD to have. I was crapping myself the whole time that I might kill the chip.

[Sparky]

I took the lid off my old Opteron 165 because its temps were out of control. I never put the lid back on, I just did an old-school heatsink on the die mount with some arctic silver 5. Worked just fine once I modified my heatsink's mount to work with the topless CPU.

[D_A]

Mine's a 185 and I could never get the mount sorted without the IHS. Doesn't matter now, it's been reliable ever since.

[itznfb]

I've tried popping the top on my 3770k and I can't get anything under the lid. Surgical blades, exacto knives, razor blades.... I even tried my SOG Visionary and a small hammer
For the life of me I couldn't get any blade between the lid and the chip.

[retro77]

CPU surgery? count me out

[Sparky]

CPU surgery? count me out

Where's your sense of adventure!

[OldChap]

Could this be a good excuse for sending nice cheap Bought in the USA processors here to UK in the months to come...

Sir, I regret to inform you that during the course of testing the metal lid on this cpu has become detached.... Therefore: Item value $10

[itznfb]

Victory is mine!

300_0387.jpg

[itznfb]

So far my max temps are 11C-13C cooler. Win.

[D_A]

Are you running it bare or did you replace the budget TIM under the lid with something decent?

[itznfb]

Are you running it bare or did you replace the budget TIM under the lid with something decent?

AS5 on both sides.

[D_A]

I would have gone for ICDiamond, but AC5 is still better than the standard stuff, obviously. Nice reduction in temps

[itznfb]

I would have gone for ICDiamond, but AC5 is still better than the standard stuff, obviously. Nice reduction in temps

I'll use something better next time I remove the chip. It was all I had lying around.

[D_A]

On-hand is always a positive attribute.

[Conquistador SW]

What did you use to get the lid off in the end?

I want to try the same, but also have a hard time finding something that fits in between the IHS and the PCB (or whatever it is called) I have a surgical scalpel, but even that seems to be too thick.

[masterg]

i dont have the money to buy another 3770k if broken, but dang. with that stuff i could push my chip stable higher

[haschioz]

use the liquid metal "paste". it gave some ocers 100-200mhz (not sure which) of additional headroom with lower temps. a 10°C difference didnt result in any higher oc.

[itznfb]

What did you use to get the lid off in the end?

I want to try the same, but also have a hard time finding something that fits in between the IHS and the PCB (or whatever it is called) I have a surgical scalpel, but even that seems to be too thick.

Went to Lowes and picked up a pack of TaskForce scraper blades and a Kobalt snap blade retractable knife. I used the scraper blade to work the corners with the CPU standing on it's side blade pushing down rocking back and forth. Then connected the corners with the knife. Took about 10 minutes to do the first one. I've done 3 more since for other people which took about 5 minutes. Takes a lot longer to get that crappy glue off though.

300_0389.JPG

i dont have the money to buy another 3770k if broken, but dang. with that stuff i could push my chip stable higher

Send it to me. I'll pop that sucker off.

use the liquid metal "paste". it gave some ocers 100-200mhz (not sure which) of additional headroom with lower temps. a 10°C difference didnt result in any higher oc.

From the results people have posted they are getting higher more stable clocks due to the temp decrease alone. Some TIM works better than others but if you reduce idle and load temps by 10c you will get better clocks.

[itznfb]

Prior to removing the IHS I wasn't really able to go above 44x. Currently I'm running 47x @ 1.325 vcore. I'm not sure it's stable though. Seems to be crunching away but when I opened Chrome and clicked on my XS bookmark I got an Aww snap message that it failed to load the page. Clicked again and it's been working fine. I got this message on WCG client also:
5/24/2012 8:21:38 AM Suspending computation - CPU usage is too high
5/24/2012 8:21:48 AM Resuming computation

[OldChap]

I'll do this with the next one I buy. I want to try a comparison of intel paste with ICD, mx4 and liquid pro. followed by Liquid pro with ICD, mx4 and Liquid pro. Anyone used that Indigo extreme?

[itznfb]

amlett did a similar comparison already: http://www.xtremesystems.org/forums/...MX4-PK1-LQP-IX

I definitely wasn't stable with 47x @ 1.325 vcore and I don't see the point in going further since I wouldn't use a vcore any higher than that for 24/7 anyway. So I stopped it back down to 46x and I'm running @ 1.275 vcore right now. 200mhz improvement by replacing stock TIM with AS5. Not too shabby.

[OldChap]

IB is ALL about the cooling I reckon. Using the MOAF and rad with warm ambient today I have air in 23.5° and water out is 24.5° giving me an average core temp of 63°

It may be chip dependent but if you can keep temps low I think you may need less volts

[itznfb]

IB is ALL about the cooling I reckon. Using the MOAF and rad with warm ambient today I have air in 23.5° and water out is 24.5° giving me an average core temp of 63°

It may be chip dependent but if you can keep temps low I think you may need less volts

I agree. I'm currently stuck with Corsair H100 as my best method of cooling.... it's a great compact system but obviously it's not going to handle much of an OC on IB.

[OldChap]

First post edited with Aida64 Mem and Cache benchmarks

[WFO]

OC, not sure how much difference it makes if any, the latest version of Real Temp is 3.70 and you are using the GT version designed for Gulftown.

http://www.techpowerup.com/downloads/2089/mirrors.php