Overclocking a 4790K and feel close to 4.8ghz overclock. This is on an Asus Z97 Pro. I am at 1.34V and using Swiftech 240X watercooler. I think I have a bad seat on my paste. I have one core running almost 10C cooler than the others. I used GELID Extreme paste and it's a pain to get a nice coating of that. I was thinking of trying liquid metal Pro? Would that be ok with the 240X block?
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
4790K overclocking..
- Thread starter Copyright
- Start date
5way tuning puts it at 1.375V and 4.9ghz.. not stable and possibly because of heat..hitting around 87C.
Doing a Blend @ 4.8ghz 1.327V temps are around 78C peak and appears to be stable.. doing a small fft causes much more heat and instability.
Doing a Blend @ 4.8ghz 1.327V temps are around 78C peak and appears to be stable.. doing a small fft causes much more heat and instability.
Last edited:
Well gave up on 4.8ghz.. found 4.7 easily reached.. Running Prime at 1.26V so far and appears stable.
EDIT: Failed at 1.26V So far stable at 1.28V... I still have one core 10C hotter than the other 3.. makes me think I need to do reseat the waterblock.. maybe try liquid pro..
1.5 hours into Prime 95 4.7ghz @ 1.28V... Max temp 83C.. lowest core is 73C.. 10C difference still between hottest and lowest core. Seems to be 100% stable so I will take it.
EDIT: Failed at 1.26V So far stable at 1.28V... I still have one core 10C hotter than the other 3.. makes me think I need to do reseat the waterblock.. maybe try liquid pro..
1.5 hours into Prime 95 4.7ghz @ 1.28V... Max temp 83C.. lowest core is 73C.. 10C difference still between hottest and lowest core. Seems to be 100% stable so I will take it.
Last edited:
UnclePauly09
Gawd
- Joined
- Feb 28, 2009
- Messages
- 523
prime is not a good stability test for maxwell. It's way too harsh, use intel extreme tuning or x264 encoding it's more reasonable and realistic of a load. No regular program is gonna push your cores to 90c like prime will.
Thx, I will give that a try. It passed 2.5 hours of Prime doing small fft. At the moment I am using Manual voltage at 1.28V and its running stable. But using manual is not letting it throttle the voltage down when it throttles clock speed down. I believe I have to use offset or adaptive in order for that to work? Or is that not recommended? It's been so long since I did my own I can't remember.
Also, isn't this a Haswell chip? I thought Mawell was Nvidia's GPU?
Manual will have it run at a constant 1.28V. I usually do that to test stability then switch it over to adaptive when I've reached an oc that I'm comfortable with. Then I stress test again using adaptive to make sure everything is running smooth.
so adaptive just use the offset or OC voltage as well? What method would you use to get a stable 1.28V if 1.28V manual works?
From what I read 1.3V seems to be the most recommended top voltage but plenty are going above that. If you have good enough cooling you got some room left to OC that CPU. I finalized at 1.28V @ 4.7ghz. Anything above that took a lot more voltage to get stability.
Domingo
Fully [H]
- Joined
- Jul 30, 2004
- Messages
- 24,141
So, I'm actually having a bit of trouble OC'ing my 4790. Mine is on an MSI Z97 Krait Edition. I have an H110 cooler.
It idles in the low 20's but Prime 95's Small FFT's makes it skyrocket to the 70's and low 80's almost instantly. My cooler is seated correctly and I'm using a thin layer of Noctua thermal paste, too.
I'm just doing the basic change of the CPU ratio and CPU voltage. Am I missing something? I don't really know a lot about other settings to play with, but my voltage to get to 4.6ghz is only 1.20...so it's not like I'm feeding it a ton of juice. What other settings should I mess with?
I'm not even really looking for performance boosts so much as just trying it out...but I can't even get much to work with.
It idles in the low 20's but Prime 95's Small FFT's makes it skyrocket to the 70's and low 80's almost instantly. My cooler is seated correctly and I'm using a thin layer of Noctua thermal paste, too.
I'm just doing the basic change of the CPU ratio and CPU voltage. Am I missing something? I don't really know a lot about other settings to play with, but my voltage to get to 4.6ghz is only 1.20...so it's not like I'm feeding it a ton of juice. What other settings should I mess with?
I'm not even really looking for performance boosts so much as just trying it out...but I can't even get much to work with.
Yes core 4 is 10c cooler. Also the 4970K is supposed to not run as hot as the 4770K..I thought they fixed the thermal compound issue? My 4770K is delidded and the temp drop was huge.
I think more people should report here and tell us if their #4 core is also running cooler than the rest.
what I have just discovered during the CPU stress test is that #4 core uses less voltage than the rest.
that might have something to do with it, take a look:
free pic
All active cores in a Core i CPU use the exact same multiplier and the exact same voltage. The VID voltage is the voltage that a CPU core is requesting. It is not equal to the voltage that the CPU cores are actually receiving. You can have inactive cores in the C3 / C6 / C7 C States. The active ones are always locked together.
The other thing to keep in mind is that the temperature sensors that Intel uses are only accurate to +/- 5°C. That means one temperature sensor might be reading 5°C lower than the actual core temperature and another sensor might be reading 5°C higher than the actual core temperature. With these sensors, a difference of 10°C from hottest to coldest is completely meaningless. Intel engineered these sensors for thermal throttle and thermal shutdown control and that is it. They are not 100% accurate temperature monitoring devices.
The other thing to keep in mind is that the temperature sensors that Intel uses are only accurate to +/- 5°C. That means one temperature sensor might be reading 5°C lower than the actual core temperature and another sensor might be reading 5°C higher than the actual core temperature. With these sensors, a difference of 10°C from hottest to coldest is completely meaningless. Intel engineered these sensors for thermal throttle and thermal shutdown control and that is it. They are not 100% accurate temperature monitoring devices.
After delid I got normal prime temps/etc including AVX2. Before I would could peg my CPU at 100C on certain tests with a mild overclock. Intel's just retarded for still using crap paste and having IHS to die gaps because a CPU should be able to handle any load out there.
4.8 for me is doable at 1.31 though, wish I could hit 4.9... OCCT was the test that required the most voltage by far for me.
Right now 1.3V peaking 81C.... doing a 30 minute Intel XTU stress test and if it passes ill move up to 2 hours.. normally if I pass 2 hours I feel comfortable using the OC. Right now just gaming and other stuff it has been 100% stable with no issues. Very happy with the results.. my fan curves keep this PC dead silent at idle and it only makes minimal noise when gaming. During my stress test I turn all fans on 100% of course.
EDIT: Passed 30 min no problem.. stayed in the mid 70s most of the time.
EDIT: Passed 30 min no problem.. stayed in the mid 70s most of the time.
Last edited: