CPU & Heatsink Lapping

[nicklmg]

Amazon: http://geni.us/33uO
NCIX: http://bit.ly/1asWg3w

 

Does CPU lapping matter? The answer may be a bit more interesting than a simple "no" this time around...

 

 

[tt2468]

1 minute ago, nicklmg said:

Amazon: http://geni.us/33uO
NCIX: http://bit.ly/1asWg3w

 

Does CPU lapping matter? The answer may be a bit more interesting than a simple "no" this time around...

 

 

Oh god the horror

EDIT: 777! This must be my lucky day!

[courtiousdoge]

5 minutes ago, tt2468 said:

Oh god the horror

EDIT: 777! This must be my lucky day!

ikr

[Stefan1024]

It's a great tip to flatten my selfmade heatsinks and I thank it will make a difference there. But for the already well machined CPU heatsinks it shouldn't make a big difference unless the manufacturer tuned the machine wrong.

[Deli]

I want to see if the result is the same if Luke laps the IHS of the CPU to go with the cooler.

[Progressor]

Wait.. you guys actually bothered to take 5min and research the topic about a video you're making instead of just half-assing yet another "workshop" video with plainly false conclusions.. what? Better be late than never I guess, right? 

Anyway.. regarding the topic, this is one I actually would not expect much of a result if any. Unless the contact surface of the heatsink is severely scratched or uneven you shouldn't see a difference, the thermal paste should be able to take care of filing any micro scratches or lightly uneven surfaces.  

[zMeul]

this was one horrible video

 

the title sais CPU and heatsink lapping, Luke only did the heatsink

these days, heatsinks are flat enough, but CPUs still aren't

 

there are claims that lapping both the CPU and heatsink to a flat mirror finish eliminates the need for a thermal compound; never tested it for myself tho

 

if you start a project, see it through the end

 

example of lapped CPU: http://forums.anandtech.com/showthread.php?t=2300800

[Opferlamm113]

At 3:31  you say cable management and airflow direction have no impact, this may be true for the CPU, but restricted or hotter airflow can have a negative impact on other parts like HDD e.g. Please be careful if you make statements like that.

[THELOU101]

@Progressor @zMeul

Yeah I agree, the only way this would make any quantifiable difference is if no thermal paste is applied, although I don't see why you would want to do that.

 

[EricX2]

Once you researched and found it didn't work with the heat pipe style heat sink why didn't you redo the work with a stock heat sink or any random cheap heat sink which would likely have less flat surface to begin with?

[ORCACommander]

I may be misremembering but concave vs convex cooler boil down again to an intel vs amd argument. AMD and intel use opposite stamping directions when fabricating the Integrated HeatsSink. IIRC Intel IHS bow in a convex fashion which means a Concave cooler would perform better.

For lapping to be done properly both the Heatsink and the processor need to be lapped in order for them to be both Leveled surfaces. One or the Other being lapped invalidates it as it still keeps a convex or concave surface in the equation.
 

I think Luke may have over done it with the grit progression but i suppose that isn't a bad thing

 

Luke never mentioned if he also lapped the processor he tested with the lapped heatsink

[907rider]

Idea: use Linus's home made water block from scrapyard wars, lap that and see if it makes a difference.. If lapping a CPU heatsink/waterblock would ever make a difference, it would be with that home made deallo.

[DXMember]

wait? you want us to sand-grid our coolers?

isn't that what thermal interface material is used for?

what kind of a sick April's fool joke is that?

[Admiral Snackbar]

Most of the stuff I have seen where lapping actually did something is on the CPU heatspeader and NOT the cooler. Any decent cooler already has a flat/polished enough base, but CPU heatspreaders are notoriously NOT flat. 

 

This is my i7-3770k. Notice how when I started sanding tons of material came off the edges not not the middle? I noted a 5 degree improvement from this.

 

[-Old-Tom-]

The result is kind of what I expected as with modern high quality thermal compounds and the high mounting pressures they will take care of any minor defects or convex curvedness.

 

@Admiral Snackbar did you record your ambient temps as if the room temperature decreased by 5 degrees then it basically means no change. although if you have a flat heatsink then taking the edges off your CPU spreader as you show may have made some difference.

 

It would be nice to see a video done on this CPU spreader lapping combined with the CPU being de-lidded and thermal paste replaced. maybe even a heatsink direct on die comparison would be good also.

[Admiral Snackbar]

@-Old-Tom- yes the ambient temp was the same.