Overclock.net › Forums › Intel › Intel CPUs › Skylake Overclocking Guide [With Statistics]
New Posts  All Forums:Forum Nav:

Skylake Overclocking Guide [With Statistics] - Page 569

post #5681 of 11309
Lentil sized drop always was the most consistant for me. I used to use a credit card to spread it over but i stopped doing that once i did testing for myself and saw a single dot worked best.

I was going to prime the ihs and waterblock by putting a dot and spreading it around with a glove then wipe it all off with a paper towel (leaving behind a residue film) it always worked well but im not sure if it made any improvments
Taiga's 2nd wind
(14 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 6600K Gigabyte GA-Z170XP-SLI Gigabyte GTX 970 G1 Gaming G.Skill Ripjaws V series 
Hard DriveHard DriveCoolingOS
Western Digital 1TB HDD Mushkin Enhanced Eco2 SSD Silverstone TD02-Lite Windows 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Acer G257HU Roccat Isku Corsair CX600M Antec P182 
MouseMouse Pad
Roccat Kone+ SteelSeries 
  hide details  
Reply
Taiga's 2nd wind
(14 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 6600K Gigabyte GA-Z170XP-SLI Gigabyte GTX 970 G1 Gaming G.Skill Ripjaws V series 
Hard DriveHard DriveCoolingOS
Western Digital 1TB HDD Mushkin Enhanced Eco2 SSD Silverstone TD02-Lite Windows 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Acer G257HU Roccat Isku Corsair CX600M Antec P182 
MouseMouse Pad
Roccat Kone+ SteelSeries 
  hide details  
Reply
post #5682 of 11309
Quote:
Originally Posted by error-id10t View Post

You don't mention you setting LLC here, did you? LLC5 should raise it ~0.02v while if you used LLC4 it should be similar to what you set, eg; 1.35v (hwinfo vcore reading, not VID).

I have LLC at 4 in both scenarios.
post #5683 of 11309
question on temps. with a 4.6ghz@1.32v overclock, is 75c underload using OCCT 4.4.1 Small data set stress test high? or is that safe?

Welp, it passed the 1hr test and only touched 75c a few times. Temps stabilized at 74c. Load core voltage is 1.308

http://valid.x86.fr/htpnfi

I overclocked my memory too. stock is 2400mhz, 1.2v 15-15-15-35. Overclocked it to 2666mhz 1.34v 15-15-15-35. Time to see if that's stable


Edited by Nenkitsune - 2/13/16 at 7:32pm
Taiga's 2nd wind
(14 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 6600K Gigabyte GA-Z170XP-SLI Gigabyte GTX 970 G1 Gaming G.Skill Ripjaws V series 
Hard DriveHard DriveCoolingOS
Western Digital 1TB HDD Mushkin Enhanced Eco2 SSD Silverstone TD02-Lite Windows 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Acer G257HU Roccat Isku Corsair CX600M Antec P182 
MouseMouse Pad
Roccat Kone+ SteelSeries 
  hide details  
Reply
Taiga's 2nd wind
(14 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5 6600K Gigabyte GA-Z170XP-SLI Gigabyte GTX 970 G1 Gaming G.Skill Ripjaws V series 
Hard DriveHard DriveCoolingOS
Western Digital 1TB HDD Mushkin Enhanced Eco2 SSD Silverstone TD02-Lite Windows 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Acer G257HU Roccat Isku Corsair CX600M Antec P182 
MouseMouse Pad
Roccat Kone+ SteelSeries 
  hide details  
Reply
post #5684 of 11309
Quote:
Originally Posted by ChilledNineball View Post

Hey guys, first post, but this behavior was weird enough I thought it might merit one.

I'm running a 6700k on an Asus Maximus VIII Hero (UEFI 1402) at 100x45, with my vcore at 1.36, LLC 4. I'm using a Noctua NH-D14 for a cooler.

If I set manual voltage mode at 1.36v and run Prime95 (small FFT), my VID hangs around 1.36 and my temperatures are where I'd expect them to be.

If I set adaptive voltage at 1.35v+0.01 and run Prime95 (small FFT), my VID hangs around 1.4 and my temperatures are about 8 degrees higher than when I run in manual voltage mode.

I've repeated this process a number of times thinking I must have missed something, but it seems to be feeding my CPU way more voltage than it's supposed to with the given settings.

Has anyone else noticed this behavior? I'd really like to use adaptive mode, but the temperatures are just not acceptable at load.

Also, I think I must have lost the silicon lottery. I can't seem to hit 4.6 stable even going as high as 1.42v, but I'm at 3.5 stable at 1.35v-1.36v. Depressing.
How about something like 1.338 +0.020 offset? Also, any difference if you monitor vcore instead of vid using hwinfo64 (in the OP).
miso-o
(10 items)
 
  
CPUMotherboardGraphicsRAM
Intel 6700k @4830 Asus Hero VIII EVGA 980Ti Classy G.Skill 8GBx2 @3780 
Hard DriveHard DriveCoolingOS
Samsung 850Evo 250GB Crucial BX100 500GB Noctua D15S Windows 10 Home 
PowerCase
Seasonic X750 Corsair 730T 
  hide details  
Reply
miso-o
(10 items)
 
  
CPUMotherboardGraphicsRAM
Intel 6700k @4830 Asus Hero VIII EVGA 980Ti Classy G.Skill 8GBx2 @3780 
Hard DriveHard DriveCoolingOS
Samsung 850Evo 250GB Crucial BX100 500GB Noctua D15S Windows 10 Home 
PowerCase
Seasonic X750 Corsair 730T 
  hide details  
Reply
post #5685 of 11309
Quote:
Originally Posted by misoonigiri View Post

How about something like 1.338 +0.020 offset? Also, any difference if you monitor vcore instead of vid using hwinfo64 (in the OP).

So I just finished doing this, and do have some interesting information to report (and thanks for the suggestion, not sure why I didn't think of doing that). Here's the data I collected:

1.36v, manual
vid 1.36v under load
vcore 1.344v under load
vcore 1.36-1.376 idle

1.35v+0.01v offset, adaptive
vid 1.392v under load
vcore 1.392v under load
vcore 0.768v idle

1.338v+0.020 offset, adaptive
vid 1.387v under load
vcore 1.376v under load
vcore 0.768v idle

1.33v+0.020 offset, adaptive
vid 1.382v under load
vcore 1.376v load
vcore 0.784v idle

1.30v+0.020 offset, adaptive
vid 1.3527v under load
vcore 1.344v load
vcore 0.784v idle

So, strangely, when running in manual mode, the VID is 1.36v but it actually applies 1.344v under load, while keeping 1.36v and up to 1.376v at idle. Switching to adaptive, I was able to take things all the way down to 1.3v+0.02, which yielded the same load voltage as 1.36v does in manual mode. So far I'm stable running Prime95 27.9 for 10 minutes, small FFTs, and my vcore is staying at 1.344v.

I wonder if this could have somehow also been affecting my maximum OC. I really doubt it, and it's probably wishful thinking on my part, but I'm tempted to start pushing for 4.6 again. I'm unsatisfied with 4.5.

Thank you for your help.
post #5686 of 11309
Quote:
Originally Posted by ChilledNineball View Post

Warning: Spoiler! (Click to show)
So I just finished doing this, and do have some interesting information to report (and thanks for the suggestion, not sure why I didn't think of doing that). Here's the data I collected:

1.36v, manual
vid 1.36v under load
vcore 1.344v under load
vcore 1.36-1.376 idle

1.35v+0.01v offset, adaptive
vid 1.392v under load
vcore 1.392v under load
vcore 0.768v idle

1.338v+0.020 offset, adaptive
vid 1.387v under load
vcore 1.376v under load
vcore 0.768v idle

1.33v+0.020 offset, adaptive
vid 1.382v under load
vcore 1.376v load
vcore 0.784v idle

1.30v+0.020 offset, adaptive
vid 1.3527v under load
vcore 1.344v load
vcore 0.784v idle

So, strangely, when running in manual mode, the VID is 1.36v but it actually applies 1.344v under load, while keeping 1.36v and up to 1.376v at idle. Switching to adaptive, I was able to take things all the way down to 1.3v+0.02, which yielded the same load voltage as 1.36v does in manual mode. So far I'm stable running Prime95 27.9 for 10 minutes, small FFTs, and my vcore is staying at 1.344v.

I wonder if this could have somehow also been affecting my maximum OC. I really doubt it, and it's probably wishful thinking on my part, but I'm tempted to start pushing for 4.6 again. I'm unsatisfied with 4.5.

Thank you for your help.
I see, from your readings & results you should forget abt vid, just stick to vcore readings so its less confusing.
miso-o
(10 items)
 
  
CPUMotherboardGraphicsRAM
Intel 6700k @4830 Asus Hero VIII EVGA 980Ti Classy G.Skill 8GBx2 @3780 
Hard DriveHard DriveCoolingOS
Samsung 850Evo 250GB Crucial BX100 500GB Noctua D15S Windows 10 Home 
PowerCase
Seasonic X750 Corsair 730T 
  hide details  
Reply
miso-o
(10 items)
 
  
CPUMotherboardGraphicsRAM
Intel 6700k @4830 Asus Hero VIII EVGA 980Ti Classy G.Skill 8GBx2 @3780 
Hard DriveHard DriveCoolingOS
Samsung 850Evo 250GB Crucial BX100 500GB Noctua D15S Windows 10 Home 
PowerCase
Seasonic X750 Corsair 730T 
  hide details  
Reply
post #5687 of 11309
Quote:
Originally Posted by ChilledNineball View Post

So I just finished doing this, and do have some interesting information to report (and thanks for the suggestion, not sure why I didn't think of doing that). Here's the data I collected:

1.36v, manual
vid 1.36v under load
vcore 1.344v under load
vcore 1.36-1.376 idle

1.35v+0.01v offset, adaptive
vid 1.392v under load
vcore 1.392v under load
vcore 0.768v idle

1.338v+0.020 offset, adaptive
vid 1.387v under load
vcore 1.376v under load
vcore 0.768v idle

1.33v+0.020 offset, adaptive
vid 1.382v under load
vcore 1.376v load
vcore 0.784v idle

1.30v+0.020 offset, adaptive
vid 1.3527v under load
vcore 1.344v load
vcore 0.784v idle

So, strangely, when running in manual mode, the VID is 1.36v but it actually applies 1.344v under load, while keeping 1.36v and up to 1.376v at idle. Switching to adaptive, I was able to take things all the way down to 1.3v+0.02, which yielded the same load voltage as 1.36v does in manual mode. So far I'm stable running Prime95 27.9 for 10 minutes, small FFTs, and my vcore is staying at 1.344v.

I wonder if this could have somehow also been affecting my maximum OC. I really doubt it, and it's probably wishful thinking on my part, but I'm tempted to start pushing for 4.6 again. I'm unsatisfied with 4.5.

Thank you for your help.
Also keep in mind software readings are only to a .016v resolution, so you could be over/under what software reads.

By my DMM, generally using Adaptive and under a high load such as P95 12K fft's on my M8 Hero, LLC 3 & LLC 4 gives a fair amount of droop, and LLC 4 actually overshoots the setting on a more moderate load such as x264. If you need/want a really accurate reading I would use a DMM.
post #5688 of 11309
Is it good to have vcore 1.33-1.34V all the time? When my clocks come to 800MHZ when idling my Vcore stays 1.33V but VID goes to 0.77V. Temps are very good but just asking.. rolleyes.gif
Skylake
(9 items)
 
  
CPUMotherboardGraphicsRAM
i5 6600K 4,5Ghz GA-Z170-HD3P Asus GTX 1060 6GB Dual OC 16GB Crucial BallistiX Sport DDR4 2400Mhz 
Hard DriveOSPowerCase
250GB Samsung 850 SSD Windows 7 64 Corsair RM650x Fractal Define R3 
Audio
Scarlett 2i2 
  hide details  
Reply
Skylake
(9 items)
 
  
CPUMotherboardGraphicsRAM
i5 6600K 4,5Ghz GA-Z170-HD3P Asus GTX 1060 6GB Dual OC 16GB Crucial BallistiX Sport DDR4 2400Mhz 
Hard DriveOSPowerCase
250GB Samsung 850 SSD Windows 7 64 Corsair RM650x Fractal Define R3 
Audio
Scarlett 2i2 
  hide details  
Reply
post #5689 of 11309
Quote:
Originally Posted by mandrix View Post

Also keep in mind software readings are only to a .016v resolution, so you could be over/under what software reads.

By my DMM, generally using Adaptive and under a high load such as P95 12K fft's on my M8 Hero, LLC 3 & LLC 4 gives a fair amount of droop, and LLC 4 actually overshoots the setting on a more moderate load such as x264. If you need/want a really accurate reading I would use a DMM.

Thanks for the information. Definitely something to keep in mind if I encounter any stability issues, as well.

Right now I'm just trying to think of some way to push past this wall I've encountered at 4.5. I just can't seem to get something stable at 4.6 or higher. Just giving it more voltage isn't doing it, and I appear to have some thermal headroom. BSODs at 1.36 manual, BSODs at 1.42 manual. Yet 4.5 is perfectly stable at 1.35v-1.36v manual. I'm really not used to hitting a wall that isn't thermal.
post #5690 of 11309
Quote:
Originally Posted by ChilledNineball View Post

Thanks for the information. Definitely something to keep in mind if I encounter any stability issues, as well.

Right now I'm just trying to think of some way to push past this wall I've encountered at 4.5. I just can't seem to get something stable at 4.6 or higher. Just giving it more voltage isn't doing it, and I appear to have some thermal headroom. BSODs at 1.36 manual, BSODs at 1.42 manual. Yet 4.5 is perfectly stable at 1.35v-1.36v manual. I'm really not used to hitting a wall that isn't thermal.
How abt trying LLC5 instead of LLC4. Since you were at 1.344 vcore at load for 4.5ghz, suggest you retry from 1.344 at uefi with LLC5.
miso-o
(10 items)
 
  
CPUMotherboardGraphicsRAM
Intel 6700k @4830 Asus Hero VIII EVGA 980Ti Classy G.Skill 8GBx2 @3780 
Hard DriveHard DriveCoolingOS
Samsung 850Evo 250GB Crucial BX100 500GB Noctua D15S Windows 10 Home 
PowerCase
Seasonic X750 Corsair 730T 
  hide details  
Reply
miso-o
(10 items)
 
  
CPUMotherboardGraphicsRAM
Intel 6700k @4830 Asus Hero VIII EVGA 980Ti Classy G.Skill 8GBx2 @3780 
Hard DriveHard DriveCoolingOS
Samsung 850Evo 250GB Crucial BX100 500GB Noctua D15S Windows 10 Home 
PowerCase
Seasonic X750 Corsair 730T 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › Skylake Overclocking Guide [With Statistics]