Another 6700K delidded

FTC · Nov 17, 2016

Hi,

This is another of those 'already seen' threads... but just to share my experience..

Yesterday I delidded my 6700K.. I used the rockit cool delid tool, with CLU between core and IHS, and thermal grizzly between IHS and cooler (NH-C14S). It was easy but scary and took a bit of time since I double-triple checked every step and my ITX case is not very friendly to work on space wise.

In terms of gains, it was before running at 4.3Ghz (4 cores), 4.5Ghz (1-2 cores), and temps normally did not exceed 80C in Prime95. In OCCT after 20 minutes max reached was 83C. After delid I found a decrease of 12C, OCCT maxes now 71C and P95 top is 69C, so I have already pushed it further to 4.6Ghz (all cores) and still obtain 3C-4C better temps than before ... although I had to increase voltage from 1.23v to 1.3v.

In summary, I think delidding is absolutely worth, but doing it requires care and patience... and do not it if not prepared to assume the risk.

Witchdoctor · Nov 17, 2016

Nice work man, glad it worked out for you :thup:

Silver Surfer · Nov 17, 2016

No Offense Please. This has got to be a little on the disappointing side for you?

A 300mhz increase is just not worth it to me, I mean if you had gone from 4.3ghz to 5ghz that would be worth Break dancing over, but not just 300mhz.

Anyway I'm happy for you that you did not screw up the chip, Congratulations.

FTC · Nov 17, 2016

Silver Surfer said:
No Offense Please. This has got to be a little on the disappointing side for you?

A 300mhz increase is just not worth it to me, I mean if you had gone from 4.3ghz to 5ghz that would be worth Break dancing over, but not just 300mhz.

Anyway I'm happy for you that you did not screw up the chip, Congratulations.

Hi, no problem. First of all I did this to learn and have some fun given the fact that I was not 100% happy with temps under load, but I am perfectly aware that I do NOT really need the added performance (even if it is very well welcome). All in all I did not expect much more than this, And the more important part is that I am not finished yet. I will try to push this proc to its limits under controlled temps. Just pushing 4.6Ghz w/1.31v has been done as a 'fast start', because I had tested it previously w/fewer cores and knew it would run w/1.28v aprox, so to be on the safe side I pushed 1.31v and did a few stability tests w/all cores. So my expectation is to be able to run 4.7 or 4.8 Ghz in the end, fully tuned and with temps under 80C. Will post in this thread again my progress..

Jeff G · Nov 17, 2016

I never have temps reach 80C on my 4.7 overclock, and that's pushing 1.4v.
I was contemplating a delid (more like finding a local, experienced member to do it for me!), but in the end I really wouldn't be happy if all it did was get me to 4.8 when I'm already running 4.7 daily.
Do you plan to push this further now to see what your max overclock is? I'd be really curious to see how far you can go now that you have a much cooler running CPU.

EarthDog · Nov 17, 2016

Nice one OP! Thanks for sharing!

Jeff, you are, for all intents and purposes, voltage limited anyway. I wouldn't bother with yours.

FTC · Nov 17, 2016

Jeff G said:
I never have temps reach 80C on my 4.7 overclock, and that's pushing 1.4v.
I was contemplating a delid (more like finding a local, experienced member to do it for me!), but in the end I really wouldn't be happy if all it did was get me to 4.8 when I'm already running 4.7 daily.
Do you plan to push this further now to see what your max overclock is? I'd be really curious to see how far you can go now that you have a much cooler running CPU.

jeff, thanks for your msg. Of course.. you are on water cooling and I'm on air and w/an ITX case.. so I really needed to do something if wanted to push this further and not being so much 'temps limited'. We'll see the limits of my proc in the next couple of days..

- - - Updated - - -

EarthDog said:
Nice one OP! Thanks for sharing!

Jeff, you are, for all intents and purposes, voltage limited anyway. I wouldn't bother with yours.

Thanks Earthdog, I try to share all my related experiences w/OC in this fora, however this tends to be only every couple of years.. when I change the rig. (I was still using my dothan based rig OC to 2.6Ghz from 2008 until 2 months ago...).

Silver Surfer · Nov 17, 2016

FTC said:
Hi, no problem. First of all I did this to learn and have some fun given the fact that I was not 100% happy with temps under load, but I am perfectly aware that I do NOT really need the added performance (even if it is very well welcome). All in all I did not expect much more than this, And the more important part is that I am not finished yet. I will try to push this proc to its limits under controlled temps. Just pushing 4.6Ghz w/1.31v has been done as a 'fast start', because I had tested it previously w/fewer cores and knew it would run w/1.28v aprox, so to be on the safe side I pushed 1.31v and did a few stability tests w/all cores. So my expectation is to be able to run 4.7 or 4.8 Ghz in the end, fully tuned and with temps under 80C. Will post in this thread again my progress..

I'm considering the 6700K myself so I am interested to see just how far you do get with it.

Good Luck to You!

Jeff G · Nov 17, 2016

FTC said:
jeff, thanks for your msg. Of course.. you are on water cooling and I'm on air and w/an ITX case.. so I really needed to do something if wanted to push this further and not being so much 'temps limited'. We'll see the limits of my proc in the next couple of days..

Makes sense, I was just wondering if the 12C drop you saw would be the same for me and allow me to get closer to the 5ghz range. But I'm guessing by Earthdog's response, that I will not be able to push voltage much higher. Either way, glad you were able to see some good results and it looks like you could push a little further still!

EarthDog · Nov 17, 2016

Most wouldn't go more than 1.4-1.45V. And you are already there really. But, I am not a huge proponent of delidding in the first place. THe Mhz gained for the average user just doesn't make the risk worth it.

dejo · Nov 17, 2016

once you start to oc the burden of voltage and mhz combine to brings temps up pretty quickly. going from just a decent heatsink to a great heatsink wont make a ton of difference in an overclock most of the time. I have delidded several chips and never really gained anything from it.
have a friend that has a 3770k that was crazy hot at low temps. stock clocks with a noctua hh-d15 would go to the 80s. he delidded and that chip wont greak 60c at 1.51v at 5.aghz even using intel burn test to bring the heat.

FTC · Nov 17, 2016

EarthDog said:
Most wouldn't go more than 1.4-1.45V. And you are already there really. But, I am not a huge proponent of delidding in the first place. THe Mhz gained for the average user just doesn't make the risk worth it.

Yes, I agree. The point is the risk ... BUT if you take it and survive the attempt, note that even if you 'only' win a couple of Mhz, and even if you have to buy a delid tool to make it safe, the cost is more or less in line with the cost difference of two processors with the same couple of Mhz difference in speed,...

trents · Nov 17, 2016

Silver Surfer said:
No Offense Please. This has got to be a little on the disappointing side for you?

A 300mhz increase is just not worth it to me, I mean if you had gone from 4.3ghz to 5ghz that would be worth Break dancing over, but not just 300mhz.

Anyway I'm happy for you that you did not screw up the chip, Congratulations.

That's an unreasonable measuring stick IMO. Each to his own, though.

Woomack · Nov 18, 2016

Actually all users see temps drop after delidding but not all see OC improvement so 300MHz in quite big difference. All depends from too many factors and the CPU itself. My i3 had about 10*C less but wasn't overclocking better on air/water ... it could make 100MHz more on dice. The same was with 3770K, 12*C less but no OC improvement but in this case not even on sub 0 temps.
You will see OC improvement when your cooling is holding you down and when is higher temp difference between CPU cores ( like 8*C+ ). In some cases cores have almost the same max temps and there is not much to improve.
So again 200-300MHz more is really good result on ambient temps.

FTC · Nov 19, 2016

Well, first update, I've since my last post completed stability tests for 4.6Ghz/1.3v AND for 4.7Ghz/1.35v vcore. Temps at this speed (4.7) are now similar to the temps I obtained before delidding at 4.3Ghz (for all cores running), that is 83C max core temp w/OCCT, and 79C w/Prime95. I may try in the following days to test 4.8Ghz but only for testing purposes, not 24x7 as at this point I would not feel confortable with higher temps than current ones in a daily basis.

Re. the stability tests, I basically use OCCT (both default and linpack tests), Prime95 and Real Bench, each one for about 1-1.5 hours and then IBT (very high settings) and a set of other minor tests such as XTU, Aida64 stability, some graphics benchmarks, ... for shorter periods, plus some gaming and general 'office work'. If all of these run without problems I consider a given pair freq/voltage as stable.

EarthDog · Nov 19, 2016

Temps are fine stress testing... they are g9ing to be well less doing everything else. If you want 4.8, the only thing holding you back is you.

FTC · Nov 19, 2016

EarthDog said:
Temps are fine stress testing... they are g9ing to be well less doing everything else. If you want 4.8, the only thing holding you back is you.

Well, I just tested 4.8Ghz with vcore=1.41v, and it ran ... for a while. Prime95 hanged 4 minutes into the test, and temps were already 86C for the hotter core, so it may need some more voltage to run at this speed, and temps will for sure approach 90C in prime95 .. probably 93-94 in OCCT if it is stable at all.. too much for my liking.

Soulcatcher668 · Nov 19, 2016

Time for that custom water loop. :chair:

FTC · Nov 19, 2016

Soulcatcher668 said:
Time for that custom water loop.

Well.. thanks for the comment but I'll leave it here (at 4.7Ghz). The cost would be prohibitive for 100Mhz gains and.. remember I have an ITX case.. In any case the performance of this rig is brilliant already...

deathman20 · Dec 5, 2016

Yeah I can run 4.8Ghz myself with 1.425V but for the heat and potentially pushing the chip not worth it at all. Instead I run a very comfortable 4.5Ghz with 1.25V and around 50C temps right now.

Another 6700K delidded

Member

Member

Member

Member

Member

Gulper Nozzle Co-Owner

Member

Member

Member

Gulper Nozzle Co-Owner

Senior Moment Senior Member

Member

Senior Member

Benching Team Leader

Member

Gulper Nozzle Co-Owner

Member

Member

Member

High Speed Premium Senior

Similar threads