Well, after lapping my HS last night, I've had this nagging little voice in my head telling me to do the same from the CPU. I did the job totally with 800 grit sandpaper. Initially, I told myself I'd just buff what's there right now just to see if it's level. After about 30 laps in one direction and 30 in the other direction I discovered I had quite a concave IHS. So I just kept at it. Two 9x11 pieces of 800 grit later paper later I was left with a darn flat layer of copper looking back at me.
I was very happy to see that it lowered my temps doing the samex264.exe encoding test I have been using to benchmark them. Below are the numbers some of which you may have read in the thread I started about the heat sink last night. By the way, these numbers are my Q6600 @ 9x333 and as the caption in the image says, they are averaged over the entire encode:
The one thing that still has me a little miffed is the difference between the two dies (Cores 1/0 and Cores 2/3). Either there is something internally that's off perhaps related to the HIS (which I’m not going to remove so don’t suggest it) or the sensors aren't placed symmetrically and I’m being mislead. Oh well.
Here's the process I used: just a piece of 8x10 glass and tapped a strip of the 800 grit to it. A little splash of water and about 45 laps in one direction, clean sand paper and the CPU, then rotated 90 degrees and 45 laps in the other. Repeat. I then replaced the strip of 800 grit and repeated until I got this result (which by the way looks pretty scratched up due to the camera flash, and yes I know there's a little bit of the nickel left and that it still appears to be concave a little bit, but I didn't want to press my luck and will leave it as-is):
BTW, I've been running 9x366 (3.3 GHz) for the past hour stable on stock voltage. Temps are little higher (53, 53, 48, 47) but those aren't averaged like the first set was; it still has about 40 min to go before I have the full data set.
I've been thinking of lapping my CPU recently, so thanks for the info on how. My possessor would run at about 80c to 83c at the tjunction, at stock speeds on the stock cooler, under load, so I had to get a better HSF, but I plan to overclock my configuration in a few years, and I would like to get the best cooling out of my HSF if I do.
Interesting that temps get that high without any voltage increases. Potential disadvantage of quad core? I suppose that is a pretty high OC though, considering it will only go as high as it's weakest core and at stock voltage I'm impressed.
Yeah, I was surprised it booted and ran for about 5 minutes before I decided the temps were too high and dropped it back to 9x333. There is a lot of headroom in this chip. Kudos to Intel - great design.