Will lapping be worth it?

Dirkyd

Honorable
Sep 22, 2012
17
0
10,510
Hey everyone,

I just thought I'd ask if you all think it would be worth lapping my heatsink and cpu.

What I'm working with on my OC.....

Asus crosshair V Formula Z
FX-8350
RM1000
Nepton 280L

So far I have hit the cpu jackpot and have a stable 4.6 OC at only 1.38v but I'm being held back by my comfort level for heat. At 4.6 it idles at 31c and after 10 min of prime 95 small ffts it hits around 68c. One thing holding me back is I couldnt use the fans the nepton came with due to their size and am now using less effective fans (already ordered better ones). I have been debating if lapping will be worth it or not or seeing if maybe im being to cautious with my temps. My goal is to get around 5.0 stable with good temps so every little bit is being considered.

 

Dirkyd

Honorable
Sep 22, 2012
17
0
10,510


Sorry that was a typo, I meant 5.0 but am curious how far I could get this chip.
 

Dirkyd

Honorable
Sep 22, 2012
17
0
10,510


None really, cpu seems fine enough, but the heatsink on the Nepton is very rough and what i am debating lapping.
 

bmacsys

Honorable
BANNED



I have lapped dozens of cpu's. You must also lap the base of your heatsink. Done right it is very tedious and time consuming. It basically does very little if anything as far as reducing temps. It also destroys any resale value your cpu may have. Delidding works but the risk of destroying your cpu is pretty high. To get to 5 GHz you probably need a custom loop. Even the best AIO won't cut it.
 

USAFRet

Titan
Moderator


That is all up to your DIY skills. Mess it up, and scratch one CPU cooler interface.
 
I'm not entirely clear on the details of amd cpu's, but I believe the fx 8350 begins to thermal throttle at 70c. I know there are two temps, core temps and package temps so where I'm unclear is what temp the 68c the op is referring to (package or core). If I had to guess, the 70c limit is core temp. If DirkyD is already hitting 68c at 4.6ghz, it's going to need much better cooling to go further won't it? Otherwise it will just be pushing max temps and throttling?
 

bmacsys

Honorable
BANNED


After 10 minutes of Prime95 you haven't done much in the way of stress testing. The heat really kicks in after about half an hour. I would say two hours of Prime95 and you have established a good threshold of stability.
 

Dirkyd

Honorable
Sep 22, 2012
17
0
10,510


68C is cpu temp and seems to level around there, but under high load during gaming and normal use it maxes out at around 42C. I'm anticipating going cutom down the road, but atm 280L is what I have.
 

Dirkyd

Honorable
Sep 22, 2012
17
0
10,510


Yeah, Ive only done the 10 min as I was stepping it up, I havent reached my goal or a wall outside of my temp comfort level. So far it has run normal load np and hasnt skipped a beat. I will do more intensive testing once I get new fans installed.
 

bmacsys

Honorable
BANNED



You certainly have a nice rig. Good luck to you.
 
Oh ok, 42 under normal conditions is a big difference from the 68 under stress. I don't know how effective lapping is, would probably depend on the individual cooler. When lapping was more popular the cooler bases weren't machined all that great and had a rougher milled finish than a lot of them do now. Some coolers intentionally have a slightly crowned base designed to increase pressure in the center of the cpu over the cores to provide better contact.

I've noticed a lot of cooler manufacturers are paying more attention to machining the bases and they come with a high mirrored finish (doesn't always mean 'true' or 'flat'). The reason it may depend on the individual cooler, take for instance the hyper 212 evo. I've seen some with really rough pitted looking bases and the one I received was in really good shape. Probably just manufacturing tolerance error or one slipped by quality control.
 

Dirkyd

Honorable
Sep 22, 2012
17
0
10,510


Thats the only reason im looking at it atm, Ive check my cpu and it seems pretty flat and I dont think lapping will make much of a difference. However the Neptons block is poorly machined and very rough, I was seeing if anyone has experiencing in just doing the heatsink and if it made much of a difference.
 
The last one I lapped was an old zalman cnps flower type air cooler for a p4 2.8c. I might have lapped the cm v8 (not the newer gt) for my e8400 but can't remember. The base on my dark rock pro 3 looked pretty nice actually and seemed fairly flat checking with a metal straight edge so I didn't bother. You could always try it, I used to use varying grits of sandpaper ending with around 2000 I think and using a mirror for a base and some light water on the sandpaper.

One way to tell, use something flat like a mirror (just flat glass) or piece of granite tile to put the sandpaper on. Use something like 1800-2000 fine grit just to check it (no real risk of scuffing the surface like with a rougher grit). Use a marker like a sharpie or something and 'color' the base with it. Gently work it on the fine grit a few passes and then turn it over and see. The marker should reveal any places where it's not making full contact. If it's just a tiny bit around the edges I wouldn't bother, if you wear the marker off around the edges and there are spots left near the middle portion it may be worth lapping it. Depends if you feel like messing with it too.
 

Dirkyd

Honorable
Sep 22, 2012
17
0
10,510
Thanks for the input guys, Ive decided I'll just hold off on lapping until I can fully test with the new radiator fans. The increased airflow and static pressure should make a big difference and at least allow me to fully stress test at 4.6 and then take it from there.