i7 4790k overclocking help

[jjohnthedon1]

hi guys iv just done my first overclock and was wondering if my temps were ok.

i dont no my ambient temps so lets say there about 20 degrees as im inside and its summer 

i have overclocked to 4.7gz with vcore set to 1.24 

i have passed all of the stress test i threw at it

intel extrem tuning i hit 60 degrees 

prime 95 ( version 28 something) blend i hit 62 

prime small fft 90 !!! that seemed realy high to me but i know that prime is not the best with the intel chips
when playing bf4 64 server my top temp was 55 degrees

do these seem good temps for my overclock ?? ( im new to overclocking)

[typographie]

Those are very good temperatures of course, but you didn't say how long you spent stress testing. That voltage seems suspiciously low to me, but good for you if it's stable.

[jjohnthedon1]

only spent about 10 mins on each stress test as they did not peak anymore in terms of heat, im going to run a overnight test now 
my cpu seems to do well on voltage as at stock speeds it runs at 1.16v and i stressed that for 24 hours to make sure 

[SlowMixit]

10 minutes isnt enough to do even basic stability test, temperatures in your case isnt the problem, its if the cpu is stable in longer than 10min load, run 12hour test and if no errors in AIDA, Intel Extreme Tuning Utility, Prime95 or some similar stability test utility then you can say its stable.

 

Mine runs 12hour Prime95 stability test stable at 1.295v 4.8Ghz @ 64 celcius, custom watercooling loop, delidded and mounted waterblock with naked ivy kit.

 

Out of interest, whats your mainboard? 

[Imakuni]

only spent about 10 mins on each stress test as they did not peak anymore in terms of heat, im going to run a overnight test now 

my cpu seems to do well on voltage as at stock speeds it runs at 1.16v and i stressed that for 24 hours to make sure 

10 min is a little to small of a sample. Try leaving it overnight, 8h or so is a better estimate. Still, even with all the testing in the world, if you encounter problems, raise that voltage / lower that multiplier!

 

About prime 95, they use a special instruction set that extracts every single ounce possible out of the Haswell CPUs, that's why it heats to those absurd numbers. You'll never actually see them being used (except, duh, for Prime tests), so you shouldn't worry about it.

[jjohnthedon1]

ok thanks for the reply il leave a stress test running overnight ( not the small fft) and make sure its stable and il report back in the morning 
 

[JerkyMcDilerino]

ok thanks for the reply il leave a stress test running overnight ( not the small fft) and make sure its stable and il report back in the morning 

 

Make sure you use Prime95 v26.6 because latest version intend to cause CPU temperature spike up to 20C due to AVX 2 instruction. For Haswell and Haswell refresh, Intel recommend Aida64, Intel Extreme Tuning Utility, and Intel Burn Test(I could be wrong for this tool).

[jjohnthedon1]

10 minutes isnt enough to do even basic stability test, temperatures in your case isnt the problem, its if the cpu is stable in longer than 10min load, run 12hour test and if no errors in AIDA, Intel Extreme Tuning Utility, Prime95 or some similar stability test utility then you can say its stable.

 

Mine runs 12hour Prime95 stability test stable at 1.295v 4.8Ghz @ 64 celcius, custom watercooling loop, delidded and mounted waterblock with naked ivy kit.

 

Out of interest, whats your mainboard? 

my specs are 

i7 4790k

h110 (noctua fans)

gigabyte z97 gaming 5

msi 980ti 

650 superflow psu 

when u ran prime and got thoes temps was that small fft or blend ?

[jjohnthedon1]

Make sure you use Prime95 v26.6 because latest version intend to cause CPU temperature spike up to 20C due to AVX 2 instruction. For Haswell and Haswell refresh, Intel recommend Aida64, Intel Extreme Tuning Utility, and Intel Burn Test(I could be wrong for this tool).

im gona run intel extreme tuning utility overnight 

[arielfall]

Those are very good temperatures of course, but you didn't say how long you spent stress testing. That voltage seems suspiciously low to me, but good for you if it's stable.

I run 4.8 at 1.25v so that's not unbelievable at the least.

Don't run small fft with adaptive voltage OP. It spikes the voltage up like crazy hence the heat. 60s is good for gaming.

[jjohnthedon1]

I run 4.8 at 1.25v so that's not unbelievable at the least.

Don't run small fft with adaptive voltage OP. It spikes the voltage up like crazy hence the heat. 60s is good for gaming.

i no about the adaptive voltage thing im set at static atm and if it proves stable over night il set it with vcore offset

[typographie]

I run 4.8 at 1.25v so that's not unbelievable at the least.

 

I don't think it's unbelievable, just uncommon and lucky. OP didn't test for very long, so I wouldn't be surprised if there's some still some instability.

[Br3tt96]

I tested for 4 hours of Aida 64 and 4 hours of Intel xtu. I can get a stable 4.8 ghz @ 1.225v. I've just delidded so that helped my temps a bit

[jjohnthedon1]

so i ran intel extrema tuning stress whilst i was asleep and its still fine no crashes 

[0ld_Chicken]

so i ran intel extrema tuning stress whilst i was asleep and its still fine no crashes

you might be good then but i wouldn't be shocked if you find instability with another program our even just normal use. I've had overclocks pass 9 hours xtu and fail in cinebench 2x in a row. For me it's a constant validation process when I've got time. Whenever i think I'm good i usually add .025v or .05v since it shouldn't make much of a temp difference.

[jjohnthedon1]

you might be good then but i wouldn't be shocked if you find instability with another program our even just normal use. I've had overclocks pass 9 hours xtu and fail in cinebench 2x in a row. For me it's a constant validation process when I've got time. Whenever i think I'm good i usually add .025v or .05v since it shouldn't make much of a temp difference.

ty il add 0.025 to be safe then