6600k delid results

Vishera · Apr 21, 2016

EarthDog said:
Its not so much that as much as it is other variables that are unaccounted for...

For example, how do we know his cooler wasn't mounted properly or the TIM application wasn't where it needed to be for the initial results? This is why review sites use testing methods like multiple mounts and TIM applications to remove those sort of variables. Ambient temperature could also be different (1:1 ratio - but still likely 1-2C difference if they have central air conditioning).

Its possible indeed, but not remotely common to see such a drop. In this case, I do believe the OP because he said the paste that was there was dry and cracking... I do not believe that is normal and was the issue. For those with 'normal' paste between the die and IHS, temp drops are much less.

What about those who ditch the IHS entirely and opt to put the heatsink directly on the die? Surely having a water block directly on the die, instead of said block cooling a piece of metal which THEN cools the die, would yield some decent results.

Kenrou · Apr 21, 2016

trents said:
Wait, I thought Intel had gone back to using a fluxless solder metal-based TIM with Skylake. Aquaman78, what you describe in your delidding doesn't seem to support that.

The sealant feels like black rubber to be honest ?

EarthDog · Apr 21, 2016

Vishera said:
What about those who ditch the IHS entirely and opt to put the heatsink directly on the die? Surely having a water block directly on the die, instead of said block cooling a piece of metal which THEN cools the die, would yield some decent results.

Indeed it should yield better results. But the risk of doing something like that, to me, FAR outweighs any benefits from doing so.

trents · Apr 21, 2016

Vishera said:
What about those who ditch the IHS entirely and opt to put the heatsink directly on the die? Surely having a water block directly on the die, instead of said block cooling a piece of metal which THEN cools the die, would yield some decent results.

Not nearly as much as you might think compared to leaving the lid on but using something like Liquid Ultra for TIM under the lid. I've seen the results of someone who systematically tested this theory and going to a bare die changed temps almost none at all.

Woomack · Apr 22, 2016

There can be one big issue when you put cooler directly on die. It needs perfect pressure and placement. Skylake die isn't exactly in the center. There are shims to help in that or you can make it yourself but it still won't guarantee that pressure will be good enough. I can tell you only that I was trying to make it with my i3 and board was almost always showing 00 or 19 even though I was tightening water block screws really hard ( not using tools but didn't want to bend PCB ). I made it boot once or twice but I'd rather have 2-3*C more and CPU which boots everytime.

6600k delid results

Vishera

Member

Kenrou

Member

EarthDog

Gulper Nozzle Co-Owner

trents

Senior Member

Woomack

Benching Team Leader

Similar threads