Lapping

valerian · 9 Nov 2007 at 15:27

at the threshhold, got my g0 quad, got the heatsink, see sig, can anyone give me some compelling evidence to lap them both, not 2-3c, something worth spitting in the face of the warranty for?

Brian Stuart · 9 Nov 2007 at 15:44

Unless the temps are high under load and there is a vast difference between the cores then I would not risk my warranty. What cooler do you have?

Van Diemen · 9 Nov 2007 at 16:13

I got my Q6600(G0) up to 3.5, but the temps on load are just too high for my liking (70C). So I'm about to lap CPU & Heat sink, will let you know when Im done.

wiiija · 9 Nov 2007 at 16:47

Yep, I'm doing mine this weekend too.
I have a 6600 B3 and the core difference is:

0:45
1:36
2:45
3:37

I'm hoping all cores will be nice and level after the lapping. I'll post what difference it made after its done

frustin · 9 Nov 2007 at 17:18

mine are like that accept that my core 1 and 2 are 41 and 40 and my 0 and 3 are 45.

unfortunately, although lapping brought the temps down for sure and they are very flat surfaces now, the core temp differences stayed the same.

my temps are high because my heatsink (the ninja) fan can only be placed; sucking warm air from the back of the gfx card. i'm buying a new one soon.

valerian · 9 Nov 2007 at 18:00

Brian Stuart said:
Unless the temps are high under load and there is a vast difference between the cores then I would not risk my warranty. What cooler do you have?

ifx-14

core 1&2 60s
core 3&4 50s

Cob · 9 Nov 2007 at 18:03

Like 61 61 59 59?

valerian · 9 Nov 2007 at 18:09

Cob said:
Like 61 61 59 59?

no like 60 60 50 50 or 59 59 51 52

Cob · 9 Nov 2007 at 18:33

Are you sure all four cores are being stressed? That's the kind of result you would get if running Orthos instead of Prime95 v25.4 on a quad.

w3bbo · 9 Nov 2007 at 21:57

I have just finished re-building a rig for a friend and his B3 Q6600 has the same problem. Core 0 is 10c higher than core 3. After lapping the difference is down to 5c with a drop of around 10c overall(load).

mrthingyx · 9 Nov 2007 at 22:19

Whilst you're all fired up for voiding your warranties, why don't you just take the heatspreaders off the chips?

w3bbo · 9 Nov 2007 at 23:08

Because they are soldered on. The chances of killing a chip by lapping it are slim whereas removing the IHS with a blowtorch and prying the IHS away from the core is much more likely to end in tears.

cozm01 · 10 Nov 2007 at 00:14

Good luck.

mrthingyx · 10 Nov 2007 at 11:54

w3bbo said:
Because they are soldered on. The chances of killing a chip by lapping it are slim whereas removing the IHS with a blowtorch and prying the IHS away from the core is much more likely to end in tears.

Ah, cool - I thought it would be something like removing the IHS on AMD64s...

wiiija · 10 Nov 2007 at 14:20

Mines done and I'm glad I did it, there is still a core temp difference between them but its lower, the main bonus is my load temps, they have dropped dramatically!

Before, when starting prime the fan would immediately spin at full speed and my temps would rise instantly to 65 and then 70+ after 10 minutes

Now the temps went to 50 (gradually) when starting prime, the fan didnt even bother speeding up untill it hit 56 and then the temps settled at 53 under full load!

Its worth noting that I laso lapped my heatsink at the same time (Freezer 7 pro).

Cob · 10 Nov 2007 at 14:28

**** me that's quite a difference.

wiiija · 10 Nov 2007 at 14:41

Indeed, thats stock speed though, at 3ghz and 1.38 volts it settles at 65 under prime load, still lower than it was before at STOCK speed.

I can only presume both the cpu and heatsink were seriously uneven.

valerian · 11 Nov 2007 at 00:04

this ifx-14 was appallingly domed
started lapping it, arms aching

ThatJoe · 11 Nov 2007 at 00:26

Nice to see you got a good result from it bud.

Biffa · 11 Nov 2007 at 20:15

Just put another Q6600 together and one core is WAY out, 9-10degrees diff.

Might have to add lapping to the list of skills I've learnt this month.