Delidded 6700k

[SubLimation7]

This isn't going to be much of a post, I'm still working on the PC.  But I had to report my early results from delidding.  At 1.5v 5ghz it will not break 58c on hottest core, and it immediately drops back to around 55c under 100% load.  Right now its at 1.38v and doesn't break 48c on hottest core.  I will upload some pictures and stuff later, but to all you other delidders out there take your time and do it right with the best stuff, and don't put ANY adhesive under IHS.  And youll get 30+c drops like this consistently, I've done it on 3570k and now a 6700k.  I will be doing another 6700k soon so I will have another thread about that when I'm done.

 

Asus z170-a

I7 6700k

Noctua d15

 

UPDATE: Added some pictures of temps while installing some games and then a stress test + installing games in background.  I get the time of stress test isn't long, but believe it or not temps weren't increasing with time.  I'll upload more pics later.

 

UPDATE: Got a core 0 error around 11 minutes, so I boosted LLC to level 4 which gave me a small voltage bump under load, and temps are still beautiful AF if you don't mind my saying so... 

See new pics for detailed info from 1 hour stress test.

[RadiatingLight]

Question: Why doesn't intel do this to all CPUs that leave the factory?

[SubLimation7]

Just now, RadiatingLight said:

Question: Why doesn't intel do this to all CPUs that leave the factory?

Well like almost anything that comes off an assembly line, there's room for improvement.

 

With that being said, I believe intel isn't allowed to use the heavy metals they use in liquid metals due to toxicity.  But I haven't really researched the topic very much in all honesty.  You'd think they would at least reduce the "gap" created by the adhesive they use.

[done12many2]

1 hour ago, SubLimation7 said:

This isn't going to be much a post, I'm still working on the PC.  But I had to report my early results from delidding.  At 1.5v 5ghz it will not break 58c on hottest core, and it immediately drops back to around 55c under 100% load.  Right now its at 1.38v and doesn't break 48c on hottest core.  I will upload some pictures and stuff later, but to all you other delidders out there take your time and do it right with the best stuff, and don't put ANY adhesive under IHS.  And youll get 30+c drops like this consistently, I've done it on 3570k and now a 6700k.  I will be doing another 6700k soon so I will have another thread about that when I'm done.

 

Asus z170-a

I7 6700k

Noctua d15

 

UPDATE: Added some pictures of temps while installing some games and then a stress test + installing games in background.  I get the time of stress test isn't long, but believe it or not temps weren't increasing with time.  I'll upload more pics later.

 

Post some 30 to 60 minute OCCT shots.  1 minute into the test and OCCT isn't really doing much.  

 

Congrats on the delid!

[SubLimation7]

Heres a shot of 10 minutes, max temps are pretty BS, it would show those temps once in blue moon for 1/4 second then go back to around ~48c per core.  I'll upload more later with the charts and everything so you guys can see the actual temps as they happen instead of just a current/min/max chart, bout to up LLC to level 4 and let it run for an hour while I build this other system...

 

Edit: forgot to say thanks for the props! 

[Cyborgsmith]

21 hours ago, SubLimation7 said:

This isn't going to be much a post, I'm still working on the PC.  But I had to report my early results from delidding.  At 1.5v 5ghz it will not break 58c on hottest core, and it immediately drops back to around 55c under 100% load.  Right now its at 1.38v and doesn't break 48c on hottest core.  I will upload some pictures and stuff later, but to all you other delidders out there take your time and do it right with the best stuff, and don't put ANY adhesive under IHS.  And youll get 30+c drops like this consistently, I've done it on 3570k and now a 6700k.  I will be doing another 6700k soon so I will have another thread about that when I'm done.

 

Asus z170-a

I7 6700k

Noctua d15

 

UPDATE: Added some pictures of temps while installing some games and then a stress test + installing games in background.  I get the time of stress test isn't long, but believe it or not temps weren't increasing with time.  I'll upload more pics later.

I've been considering doing this with my 6700k, very inconsistent temps 45-75 on water across cores.
Did you use one of the deliding tools?

 

[SubLimation7]

26 minutes ago, Cyborgsmith said:

I've been considering doing this with my 6700k, very inconsistent temps 45-75 on water across cores.
Did you use one of the deliding tools?

 

Yes I did

 

PM me if your interested in me delidding for you.

 

Also, I'm curious, what is your frequency/voltage with those temps?

[Cyborgsmith]

25 minutes ago, SubLimation7 said:

Yes I did

 

PM me if your interested in me delidding for you.

 

Also, I'm curious, what is your frequency/voltage with those temps?

Voltage topping out at 1.39 @4.2ghz

 

[Cyborgsmith]

They only use adhesive on the outer edge of the heat spreader, its thermal compound between the die and heat spreader. Biased on recent LTT content Linus plans to reassemble the 6700k he delidded with thermal adhesive, wouldn't mind seeing how well that works as the thermal adhesive is supposed to have excellent thermal transfer. 

 

Gonna tag @nicklmg maybe we can see a followup video 

[SubLimation7]

11 hours ago, Cyborgsmith said:

They only use adhesive on the outer edge of the heat spreader, its thermal compound between the die and heat spreader. Biased on recent LTT content Linus plans to reassemble the 6700k he delidded with thermal adhesive, wouldn't mind seeing how well that works as the thermal adhesive is supposed to have excellent thermal transfer. 

 

Gonna tag @nicklmg maybe we can see a followup video 

They do put adhesive under the IHS around edge which makes a small gap.  I don't really understand what your trying to say?  But yeah they definitely need to remake the delid video... that wasn't representative at all of an even half decent delid.  And I guess I'd be curious to see thermal adhesive, but I can't imagine it performs close to conductonaut, which is twice as thermally conductive as coolaboratory, so I don't see the point in why you would use thermal adhesive...

 

You peaked my interest enough to do a bit of "research". Now this is only one brand of thermal adhesive, but it has lower thermal conductivity than even kryonaut, which is a non-conductive TIM (electrically speaking). And you better pray you don't have to remove IHS after thermal adhesive application. Like I stress in my first post, if your going to delid... do it right.   

[SubLimation7]

Just did another 6700k to replace my girlfriends 3570k build which was delidded (I think I'm going to start saying relidded) as well.  Identical results (unfortunately), her chip was from a later shipment from intel so I was hoping it worked out better, but it needs the exact same voltage for the exact same frequency and is running at the exact same temps (+/- 1-2c).

 

I'm starting to wonder what the rest of the world is doing wrong when relidding, its incredibly easy... and you can put non conductive tim on the parts your worried about getting liquid metal on to protect them so there's almost no risk assuming you don't slice into the board or something stupid.  But I'm still trying to wrap my head around why my temps are so much better than everyone else's, I guess other people just tend to rush?  Or maybe the mounting pressure I use? Idk...

[SubLimation7]

21 hours ago, Cyborgsmith said:

Voltage topping out at 1.39 @4.2ghz

 

 That seems like quite high voltage for 4.2ghz, you shouldn't need more than 1.3 even with a not so great chip.  Maybe check your LLC under digi+ power control assuming you have an ASUS board, and theres another setting under AI tweaker that your supposed to disable when overclocking as well, something to do with external VRM, but if you look around it will say they recommend to disable the setting when overclocking.

 

If you go down to 1.3v good chance you'll shave around 10c off.

[Cyborgsmith]

2 hours ago, SubLimation7 said:

 That seems like quite high voltage for 4.2ghz, you shouldn't need more than 1.3 even with a not so great chip.  Maybe check your LLC under digi+ power control assuming you have an ASUS board, and theres another setting under AI tweaker that your supposed to disable when overclocking as well, something to do with external VRM, but if you look around it will say they recommend to disable the setting when overclocking.

 

If you go down to 1.3v good chance you'll shave around 10c off.

Ya thats just letting the chip do its thing, no overclocking. Just it boosting to 4.2 has the voltage topping out at 1.39 

 

TBH I hadn't ever looked into the liquid metal, that stuff looks amazing. 

[SubLimation7]

14 hours ago, Cyborgsmith said:

Ya thats just letting the chip do its thing, no overclocking. Just it boosting to 4.2 has the voltage topping out at 1.39 

 

TBH I hadn't ever looked into the liquid metal, that stuff looks amazing. 

Well that certainly makes sense, the most recent 6700k build I did booted at 1.39 volts from it auto OCING, for some reason the TPU switch was on and it only pushed it to 4.2ghz and that was with a delid processor, so I would highly recommend doing it manually.  Like I said if you do 1.3v I imagine you'll still be stable.  I have 2 6700ks both at 4.7ghz @ 1.39v.