i5-6600K @ 4.5GHz, how safe is it to push further?

[RocLobster]

I've read that ultimately, OCing your CPU is safe until you start playing with voltage, and as a result, higher temperatures. If, for example, I kept increasing the voltage so that I could increase my clock speeds/multipliers, as long as my temperatures stay in a safe range, is there any harm to this? Is there a point where you can have too much voltage, but still be at a safe temperature? Or is temperature what really matters, and increasing voltage is fine until it starts getting too hot?

For a while, it was impossible for me to get my CPU hotter than 55°C. Then I started using an additional stress test (Prime95), and it got my CPU working, with the hottest core being 69°C. I know this is still safe, but while it's being stressed, the Vcore goes as high as 1.404v, with an average just under 1.4v. Temperature wise, I know I am okay, but from what I've read about voltages, 1.4 is pretty high...too high, in some peoples' opinions.


Just so there is no confusion, I wanted to make it clear that I don't really know what I'm doing. The only reason I've been successful thus far is because the BIOS settings are basically all set to auto. All I have to do is indicate the multiplier I want to apply, and it sets all the different voltages accordingly. It would appear to work just fine that way. The voltage is adaptive, by the way.

I guess the other main question I have is, is it worth it? Would trying to push this CPU to 4.6 or 4.7 be worth it? I've tried before and have always failed, although I now know what I believe the problem was (voltage would not auto scale high enough, so I was setting it manually..way lower than necessary). I guess the answer to this question depends on the answer to the first question. This rig is my baby, and I don't want to hurt it or cause any harm. I just spent 2 grand on it and I can't imagine damaging components.

Relevant system specs:
Intel i5-6600k 3.5GHz
Gigabyte Z170MX Gaming-5
MSI GeForce GTX 980 4G
16GB DDR4 2133 Dual channel RAM


Thanks everyone for your support

 

[niven]

This rig is my baby, and I don't want to hurt it or cause any harm.

So don't abuse it, you big pubic louse.
Joke apart, you already have a turbo mode at 3.9 GHz on this CPU so performance gain will be little and cost stability, noise and power bill.
Advice: Don't experiment on your children unless your name is Frankenstein.

 

[EarthDog]

1.4V is about the limit I would have on that CPU. Temps are fine at 69C stress testing.

As far as worth it for another 100-200Mhz.. not to most. But again, most would say to stop where you are already voltage wise.

 

[RocLobster]

Thank you both for the advice. I appreciate it; I'll leave it where it's at.

1.4V is about the limit I would have on that CPU. Temps are fine at 69C stress testing.

As far as worth it for another 100-200Mhz.. not to most. But again, most would say to stop where you are already voltage wise.

I got your PM, but my account does not appear to have the ability to respond or create PMs. If my username is really too vulgar for this site, change it to whatever. RockLobster works fine.

 

[notarat]

I've read that ultimately, OCing your CPU is safe until you start playing with voltage, and as a result, higher temperatures. If, for example, I kept increasing the voltage so that I could increase my clock speeds/multipliers, as long as my temperatures stay in a safe range, is there any harm to this? Is there a point where you can have too much voltage, but still be at a safe temperature? Or is temperature what really matters, and increasing voltage is fine until it starts getting too hot?

For a while, it was impossible for me to get my CPU hotter than 55°C. Then I started using an additional stress test (Prime95), and it got my CPU working, with the hottest core being 69°C. I know this is still safe, but while it's being stressed, the Vcore goes as high as 1.404v, with an average just under 1.4v. Temperature wise, I know I am okay, but from what I've read about voltages, 1.4 is pretty high...too high, in some peoples' opinions.


Just so there is no confusion, I wanted to make it clear that I don't really know what I'm doing. The only reason I've been successful thus far is because the BIOS settings are basically all set to auto. All I have to do is indicate the multiplier I want to apply, and it sets all the different voltages accordingly. It would appear to work just fine that way. The voltage is adaptive, by the way.

I guess the other main question I have is, is it worth it? Would trying to push this CPU to 4.6 or 4.7 be worth it? I've tried before and have always failed, although I now know what I believe the problem was (voltage would not auto scale high enough, so I was setting it manually..way lower than necessary). I guess the answer to this question depends on the answer to the first question. This rig is my baby, and I don't want to hurt it or cause any harm. I just spent 2 grand on it and I can't imagine damaging components.

Relevant system specs:
Intel i5-6600k 3.5GHz
Gigabyte Z170MX Gaming-5
MSI GeForce GTX 980 4G
16GB DDR4 2133 Dual channel RAM


Thanks everyone for your support

One thing you want to ensure is that your cooling solution is up to snuff to properly cool the CPU when you overvolt. If you decide to go for more than 4.6 lap the CPU until the heat spreader is flatter than Scott Weiland's EKG Readout to ensure good thermal conductivity to the heatsink/waterblock

 

[RocLobster]

One thing I didn't see addressed was my question about temperature related to voltage:

Is there a point where you can have too much voltage, but still be at a safe max temperature? Or is temperature what really matters, and increasing voltage is fine until it starts getting too hot?

I don't want to go over 4.5GHz, but I still want to know this. If temperature is what really matters, and voltage is only important in relation to the extra heat it produces, then I'm okay. But if this isn't the case, I think I may want to lower my voltage down from ~1.4 just to be safe and to not shorten my CPU's lifespan.

 

[EarthDog]

Is there a point where you can have too much voltage, but still be at a safe max temperature?

Perhaps... it typically doesn't work that way though.

Or is temperature what really matters, and increasing voltage is fine until it starts getting too hot?

Voltage (and temps) kill CPUs. You can kill a CPU with too much voltage even at LN2 temperatures.

 

[LutaWicasa]

I tend to always use the lowest voltage that also maintains stability

 

[RocLobster]

Perhaps... it typically doesn't work that way though.

Voltage (and temps) kill CPUs. You can kill a CPU with too much voltage even at LN2 temperatures.

I tend to always use the lowest voltage that also maintains stability

This makes a lot of sense to me, thanks for clarifying again. Under load, I average 1.38v, with a max of 1.4v, and the system is rock solid stable. Even so, I do not know if I should or want to lower my voltage (and multiplier as a result).

 

[Tír na nÓg]

You can try lowering the vCore with the same multi until you know your lowest stable voltage at a given frequency.

You can eithet have lowervCore at the same freq or a higher freq at the same vCore...

 

[wingman99]

For 4.5GHz you should be in the 1.30v-1.35vcore peak for a average overclock. Test with a stress test to save some time.