To lap or not to lap?

[Jack Flash]

Is it worth trading the Intel warranty for slightly better temps from lapping?

 

[PeteRoy]

no

 

[firewolfsm]

No one really uses the warranty after they install processor and make sure it works. I lapped mine and am happy with it.

 

[haffey]

Originally posted by: firewolfsm
No one really uses the warranty after they install processor and make sure it works. I lapped mine and am happy with it.

Same. I hadn't tested temps before, but I lapped my Q9550 and TRUE, and temps are simply amazing. I don't know if they truly dropped by a couple degrees, but it was a good experience, and I wasn't going to use the warranty anyway. In fact, that thought didn't even enter my mind until this thread.

I would just make sure the damn thing works as it's supposed to for a little while. Then take it out and lap it.

P.S. I lapped to 10 micron. Do it.

 

[TidusZ]

Worth it.

 

[WaitingForNehalem]

no

 

[n7]

If you really love overclocking to that last little bit, sure.

Otherwise definitely not.

 

[Idontcare]

Originally posted by: Jack Flash
Is it worth trading the Intel warranty for slightly better temps from lapping?

Realize you asked one question but most folks responded to a different one.

Here's the question you should have asked:

Originally posted by: Jack Flash
Here's my situation: [insert your situation details here, "I want to OC more" or "I like my fans to be quiet so low cpu temps are nice" or "I like that peace of mind knowing my chip is warrantied", etc]

For a person in my situation, is it worth trading the Intel warranty for slightly better temps from lapping?

Instead here's what most people thought you wrote:

Originally posted by: Jack Flash
If you were in my shoes, and I valued exactly what you value, is it worth it to you to trade the Intel warranty for slightly better temps from lapping?

Reading this thread I gather some people do not value lowered temps (and whatever they might do with that enhanced thermal transfer budget) for themselves over the loss of the Intel warranty. Others don't value the warranty as much as they value the benefits of lapping.

It's not clear what value their opinions will have for you though as you never told us what you value regarding CPU temps or warranties.

 

[Zap]

Originally posted by: Idontcare
Reading this thread I gather some people do not value lowered temps (and whatever they might do with that enhanced thermal transfer budget) for themselves over the loss of the Intel warranty. Others don't value the warranty as much as they value the benefits of lapping.

I value my time.

IMO I can just spend the money on a better cooler (or go liquid cooled) which will result in real gains.

 

[DrMrLordX]

I dunno, if I put the money down on top-tier air cooling or water cooling or something more exotic than that, lapping would be the #1 thing on my to-do list for both IHS and HSF/blocks (unless they didn't need it).

 

[SickBeast]

I dunno about lapping, but I popped the heat spreader off my Opteron 165 and regretted it. It did nothing for my temps and I slightly damaged my processor.

 

[DrMrLordX]

IHS removal is mostly a no-go these days due to the solder they use now.

 

[SickBeast]

Well, it was never worth it. I don't know who came up with it but it's a really dumb idea in practice. It's one of those things that work in theory only.

 

[Zap]

I have a theory about this kind of thing. It is probably propagated by people having crappy temps and reading about how lapping can magically decrease temperatures. So, they take apart the system they built 6 months ago and remove the heatsink (and blow out the fan/fins with compressed air while they're at it) and then remove the gooped on thermal paste (hey everyone starts out a noob) and lap their CPU and heatsink. Then they carefully (now that they're no longer a noob) put a thin layer of thermal paste on and install the heatsink, and get magically lower temperatures.

That's my theory and I'm sticking with it. :evil:

 

[Bl0cks]

Does anybody actually read the warranty?

 

[faxon]

Originally posted by: Zap
I have a theory about this kind of thing. It is probably propagated by people having crappy temps and reading about how lapping can magically decrease temperatures. So, they take apart the system they built 6 months ago and remove the heatsink (and blow out the fan/fins with compressed air while they're at it) and then remove the gooped on thermal paste (hey everyone starts out a noob) and lap their CPU and heatsink. Then they carefully (now that they're no longer a noob) put a thin layer of thermal paste on and install the heatsink, and get magically lower temperatures.

That's my theory and I'm sticking with it. :evil:

im putting your theory to the test on friday or saturday Zap. check by my megahalems thread in cases and cooling. i have already done the thin layer and im gonna lap my megahalems for before/after temps, and possibly my cpu after that if im not satisfied yet. this is a brand new cooler and cpu, with no dust buildup and i dust it every other week anyway, as well as running the case open. depending on the cooler's stock base, sometimes it can make a HUGE difference in temperatures from before. i wish i still had before/after temps on my athlon 4000+ and vendetta2 both lapped. thing ran like 30c @ load and 18-20c idle in a 68f ambient room, from 28c idle and 40c load in the same room, and 40c idle and 60c load with stock cooler

 

[Idontcare]

Originally posted by: Zap
I have a theory about this kind of thing. It is probably propagated by people having crappy temps and reading about how lapping can magically decrease temperatures. So, they take apart the system they built 6 months ago and remove the heatsink (and blow out the fan/fins with compressed air while they're at it) and then remove the gooped on thermal paste (hey everyone starts out a noob) and lap their CPU and heatsink. Then they carefully (now that they're no longer a noob) put a thin layer of thermal paste on and install the heatsink, and get magically lower temperatures.

That's my theory and I'm sticking with it. :evil:

We are dealing with masses of people, so naturally there will be a distribution of reasons and results.

I say that because I want to agree with you and preserve your opinion while adding the firsthand result of my own lapping experience which would contradict your statement (and I don't want to do that).

In 2007 I bought the hardware to buy five identical Q6600 rigs, same mobo, same ram, same hd, etc. With 5 rigs I valued "quiet" while at the same time I valued OC speed. I also valued not heating my home up 10C from my 6 quads (I had a QX6700 already).

Lower temps makes for lower Vcc stable rigs at any given clockspeed. That's straightforward device physics. You drop your operating temps 10C and you can decrease your Vcore while remaining stable at the same GHz. At the same time a drop in temps means you can run your case fans and/or HSF speeds all the lower and keep those decibels low.

So I went about systematically determining just how much lapping and which thermal paste, etc, I needed. I did no lapping but tested AS5, ceramique, TX-1, and TX-2. I lapped just the IHS on some, on others I just lapped the HSF. Still on others I lapped both.

On my better lap jobs, where I went to 1500 grit and really worked hard to get a flat surface, I found I did not need TIM at all and my temps were the same. I still used TIM though, simply because I had it. AS5 because it is the easiest to remove IMO with the artic clean solutions. TX-2 is like removing putty, PITA.

At any rate all said and done because I lapped I was able to run my Q6600's (all 5 of them) at their stock VID settings while overclocked to 3.3GHz (9x367) with the temps never breaking 60C on small FFT with the Tuniq fan manually turned down to 800rpm. Before I lapped I could hit 3.3GHz but it took more Vcore and I had to crank up the fans just to keep the temps in the high 70's with small FFT.

For me and what I cared about (power consumption and noise), lapping made a difference in the metrics I cared about.

But it is a very subjective process, no two people will lap their chips in exactly the same way, so to be sure there will be folks who got frustrated and did a rush job and ended up with a crappy experience from their less than stellar lap job. Or as you say, their benchmark for what a non-lapped system is like was also handicapped by the user being at a point on the learning which still had a considerable first-derivative.

But making the argument that lapping an IHS or HSF makes no difference is equal to arguing that all TIM's really have the same thermal conductivity and it is just user error that accounts for the measured differences.

 

[TidusZ]

Originally posted by: Zap
I have a theory about this kind of thing. It is probably propagated by people having crappy temps and reading about how lapping can magically decrease temperatures. So, they take apart the system they built 6 months ago and remove the heatsink (and blow out the fan/fins with compressed air while they're at it) and then remove the gooped on thermal paste (hey everyone starts out a noob) and lap their CPU and heatsink. Then they carefully (now that they're no longer a noob) put a thin layer of thermal paste on and install the heatsink, and get magically lower temperatures.

That's my theory and I'm sticking with it. :evil:

I go through compressed air like nobodys business, I clean out my heatsink minimally once a month if not twice, and I saw a pretty good decrease in temps on my e4300 after lapping it compared to having everything clean but no lap. I lapped my q9550 before I even stuck it in the motherboard so I have nothing to compare it to, but I've been more than happy with its overclocking and cooling performance. It does take a fair bit of time to do, but if you have top end air cooling like I have, water cooling costs a shitload more, and I'd rather spend a few hours lapping and watching tv than spending cash (and then working for several hours setting up water).

 

[Idontcare]

Originally posted by: TidusZ

Originally posted by: Zap
I have a theory about this kind of thing. It is probably propagated by people having crappy temps and reading about how lapping can magically decrease temperatures. So, they take apart the system they built 6 months ago and remove the heatsink (and blow out the fan/fins with compressed air while they're at it) and then remove the gooped on thermal paste (hey everyone starts out a noob) and lap their CPU and heatsink. Then they carefully (now that they're no longer a noob) put a thin layer of thermal paste on and install the heatsink, and get magically lower temperatures.

That's my theory and I'm sticking with it. :evil:

I go through compressed air like nobodys business, I clean out my heatsink minimally once a month if not twice, and I saw a pretty good decrease in temps on my e4300 after lapping it compared to having everything clean but no lap. I lapped my q9550 before I even stuck it in the motherboard so I have nothing to compare it to, but I've been more than happy with its overclocking and cooling performance. It does take a fair bit of time to do, but if you have top end air cooling like I have, water cooling costs a shitload more, and I'd rather spend a few hours lapping and watching tv than spending cash (and then working for several hours setting up water).

Even when spending a shitload of cash on the cooling side of it, it actually becomes all the more a question why you'd leave the IHS/cooling system interface as the rate-limiting portion of the thermal transfer topology.

I give you an example from experience...my $1.5k QX6700 and my $1k VapoLS. Without lapping I got 3.73GHz small-FFT stable. With lapping I got 4GHz stable.

For the ~1-2hrs I spent lapping the IHS and the evaporator head on the vapoLS I got an additional 7% performance out of the >$3k rig for the following year that I ran it like that.

Not that anyone has to value their time/money as I do, but for me in my situation with that hardware it was silly of me to run it as long as I did before I took the time to lap it and clock it even higher. It makes a difference, it is the weakest link in the thermal transfer topology after all.