OC'd 4690k core temps

[Acerpwns]

so these are my CPU temps running AIDA 64. why is core 4 so low compared to the others

PIC:

 

?

[Zeke]

Your core temp difference is normal, you don't have a problem.  Don't forget, each core is basically an individual processor, and they usually operate individually from each other.  A running program will usually use only one or two cores, so all the cores are not running at the same time, so the ones that are in use will be warmer than the ones that aren't.  There are programs and gadgets available that show the amount of usage of each core, when using one you'll see some cores in use at various levels, and others not, including the hyper-threaded virtual cores.

 

It is possible that the thermal compound on the CPU cooler is not spread evenly across the top of the CPU, and if that happens to be just above one or more of the cores, you might see a temp difference.  That is very unlikely, given how small the cores are compared to the surface of the CPU, which is designed to spread out the heat from the CPU components.  A poorly mounted CPU cooler could cause this too, but you would see bad CPU temps in general.

 

Stolen from here

 

Also if those are your temps under load, don't even bother with that, that's some great cooling you got there.

[Acerpwns]

Your core temp difference is normal, you don't have a problem.  Don't forget, each core is basically an individual processor, and they usually operate individually from each other.  A running program will usually use only one or two cores, so all the cores are not running at the same time, so the ones that are in use will be warmer than the ones that aren't.  There are programs and gadgets available that show the amount of usage of each core, when using one you'll see some cores in use at various levels, and others not, including the hyper-threaded virtual cores.

 

It is possible that the thermal compound on the CPU cooler is not spread evenly across the top of the CPU, and if that happens to be just above one or more of the cores, you might see a temp difference.  That is very unlikely, given how small the cores are compared to the surface of the CPU, which is designed to spread out the heat from the CPU components.  A poorly mounted CPU cooler could cause this too, but you would see bad CPU temps in general.

 

Stolen from here

 

Also if those are your temps under load, don't even bother with that, that's some great cooling you got there.

I know cores perform on their own but their temperatures should still be "close" to each other. Having 53 degrees on core 3 and 46 on core 4 doesn't seem right imo. Maybe how the thermal paste spread?

[Acerpwns]

And yep 100% load. H100i is great <3

[Zeke]

I know cores perform on their own but their temperatures should still be "close" to each other. Having 53 degrees on core 3 and 46 on core 4 doesn't seem right imo. Maybe how the thermal paste spread?

8 degrees is not a big difference, and yeah, the thermal paste might have something to do with that, but I wouldn't worry if I were you.

[ProKoN]

your max temp on core 1 is 51c max temp on core 4 is 46c

 

this is perfectly normal as mentioned above.

 

even with a reapplication of thermal compound you are likely to have the same result

 

core 1 and 2 will typically run hotter than 3 and 4

 

the only way to even out the temps is either to delid or run naked. both methods are not advisable.

 

+\- 5c is not that bad. ive seen 10c ore more difference in haswell chips ive owned.

 

your temps and core temps difference is perfectly normal\fine

[Acerpwns]

8 degrees is not a big difference, and yeah, the thermal paste might have something to do with that, but I wouldn't worry if I were you.

But 8 degrees could mean the difference between thermal throttling and not oh and how long should I stress test before I can conclude its stable? I'm at 17 mins with AIDA 64. Should I use other software aswell like Prime95 and Memtest86?

[Acerpwns]

your max temp on core 1 is 51c max temp on core 4 is 46c

 

this is perfectly normal as mentioned above.

 

even with a reapplication of thermal compound you are likely to have the same result

 

core 1 and 2 will typically run hotter than 3 and 4

 

the only way to even out the temps is either to delid or run naked. both methods are not advisable.

 

+\- 5c is not that bad. ive seen 10c ore more difference in haswell chips ive owned.

 

your temps and core temps difference is perfectly normal\fine

Now it's 55 on core 1,2, and 3 and 48 on core 4

[Zeke]

Just stress tested my 2600K with Prime 95 for a little bit, got a difference of 12 degrees. #1 was the coolest, #4 the hottest.

 

But 8 degrees could mean the difference between thermal throttling and not   oh and how long should I stress test before I can conclude its stable? I'm at 17 mins with AIDA 64. Should I use other software aswell like Prime95 and Memtest86?

Prime95 for two hours. If it handles that okay - it will handle pretty much everything.

[ProKoN]

Now it's 55 on core 1,2, and 3 and 48 on core 4

 

pretty normal

 

rest assured everyone's chips runs like this as well

[Acerpwns]

Just stress tested my 2600K with Prime 95 for a little bit, got a difference of 12 degrees. #1 was the coolest, #4 the hottest.

 

Prime95 for two hours. If it handles that okay - it will handle pretty much everything.

  I was told I shouldn't use Prime95 with Haswell Refresh but I'm still not sure why

pretty normal

 

rest assured everyone's chips runs like this as well

I wasn't really worried, just curious and confused

[ProKoN]

prime95 is fine as long as your on manual voltage. dont run it on adaptive voltage or your vcore will spike and so will your temps.

 

intel xtu is my preferred program for stressing haswell chips

[Acerpwns]

prime95 is fine as long as your on manual voltage. dont run it on adaptive voltage or your vcore will spike and so will your temps.

 

intel xtu is my preferred program for stressing haswell chips

Hmm in my bios it doesn't give me an option to change voltage to manual or adaptive. Just the clock speeds

For how long should I run XTU?

[ProKoN]

on gigabyte boards its tricky to set adaptive voltage

 

where you would normally enter your vcore and vring voltage push "N"

 

select normal

 

this will unlock the voltage offset option

 

you will have to use the voltage offset to run adaptive voltage

 

only do this once you have figured out what voltage is required for your desired cpu frequency

 

i suggest a minimum of 8 hours stress testing with xtu. for my personal overclocks i require xtu to run for 18-24 hours before im satisfied with my overclock

[Acerpwns]

on gigabyte boards its tricky to set adaptive voltage

 

where you would normally enter your vcore and vring voltage push "N"

 

select normal

 

this will unlock the voltage offset option

 

you will have to use the voltage offset to run adaptive voltage

 

only do this once you have figured out what voltage is required for your desired cpu frequency

 

i suggest a minimum of 8 hours stress testing with xtu. for my personal overclocks i require xtu to run for 18-24 hours before im satisfied with my overclock

so i have it at 4.4 @ 1.2v that seemed stable in the 30 minutes i tested it. should i set it to adaptive now or keep it manual or whatever the preset is?

[OCing Noob]

so these are my CPU temps running AIDA 64. why is core 4 so low compared to the others

PIC:

 

CPU Temps.png?

 

SAME WITH ME . Core 4 is lowest and core 2 is highest always . i think weaker core jump up fast in temps . when its on load or may be TMI is less at that core rather then others .

[OCing Noob]

And yep 100% load. H100i is great <3

 

I'm getting the same temps what H55 . means performance is same ??????

[ProKoN]

so i have it at 4.4 @ 1.2v that seemed stable in the 30 minutes i tested it. should i set it to adaptive now or keep it manual or whatever the preset is?

i suggested 8-24 hours lol

 

if you wanna run with your 1\2 hour stress test go for it. play some games and run some benchmarks!...mind you 4.4ghz at 1.2v seems about right! 

 

i would only set adaptive once you are certain your oc is stable. run the computer for a 2-3 days under normal operation

 

when you set the voltage offset you will have to know what your motherboard is selecting as your default vcore. 

 

 

 

default vcore + offset voltage = max voltage 

 

1.000V + .200V = 1.22V

 

this requires a bit of trial and error!

[Acerpwns]

I'm getting the same temps what H55 . means performance is same ??????

i wouldnt count on it... im not sure why tho

[OCing Noob]

so i have it at 4.4 @ 1.2v that seemed stable in the 30 minutes i tested it. should i set it to adaptive now or keep it manual or whatever the preset is?

 

 

on gigabyte boards its tricky to set adaptive voltage

 

where you would normally enter your vcore and vring voltage push "N"

 

select normal

 

this will unlock the voltage offset option

 

you will have to use the voltage offset to run adaptive voltage

 

only do this once you have figured out what voltage is required for your desired cpu frequency

 

i suggest a minimum of 8 hours stress testing with xtu. for my personal overclocks i require xtu to run for 18-24 hours before im satisfied with my overclock

 

i really don't believe in hours idea at all . its like following the next sheep . i tested rig for 7 hours 3 hours then 30 mints . results where same and what i learned for AIDA64 if your rig is not stable it will go down in first 10 mints if not and your rig is stable for 30 mints its stable no need to kick your power bill and cup TMI for 8 hours. its my opinion i'm not disagreeing from your experience . i have a bad habit to believe in something with personal experience and few examples " what happened to others in the same situation"  

[OCing Noob]

i wouldnt count on it... im not sure why tho

 

And its on 60% fan speed . Push Pull . its hell of quite too. i can't even hear my rig even sitting close to ti

[ProKoN]

i really don't believe in hours idea at all . its like following the next sheep . i tested rig for 7 hours 3 hours then 30 mints . results where same and what i learned for AIDA64 if your rig is not stable it will go down in first 10 mints if not and your rig is stable for 30 mints its stable no need to kick your power bill and cup TMI for 8 hours. its my opinion i'm not disagreeing from your experience . i have a bad habit to believe in something with personal experience and few examples " what happened to others in the same situation"  

 

fair enough. its your hardware and your experience, feel free to do what you want we all have different opinions and thats all good with me. i try to share what i know and what has worked best for me.

[Acerpwns]

i suggested 8-24 hours lol

 

if you wanna run with your 1\2 hour stress test go for it. play some games and run some benchmarks!...mind you 4.4ghz at 1.2v seems about right! 

 

i would only set adaptive once you are certain your oc is stable. run the computer for a 2-3 days under normal operation

 

when you set the voltage offset you will have to know what your motherboard is selecting as your default vcore. 

 

 

 

default vcore + offset voltage = max voltage 

 

1.000V + .200V = 1.22V

 

this requires a bit of trial and error!

 

 

i know but its late here so i dont have 8/10 hours to stress test. thanks for the help!

[ProKoN]

i know but its late here so i dont have 8/10 hours to stress test. thanks for the help!

 

 

thats when i stress test 

 

 

while i sleep

 

i dream of waking up to a computer still stressing

 

often i wake up to a bsod.

 

its all part of the process!

[Acerpwns]

thats when i stress test 

 

 

while i sleep

 

i dream of waking up to a computer still stressing

 

often i wake up to a bsod.

 

its all part of the process!

that seems like a waste of time tho, when it blue screens you should reset it immediatly so you can continue stess testing and fix it accordingly

 

but hey what do i know, im still a noob