Questions about facts about overclocking..

[utisz]

I've recently started playing around with my cpu and motherboard, here are the few things's I can't find answers online quite easily:

1. As far as I know, we should not pass 1.4vcore for a "safe" overclock (x79). Is this because it causes more temperature, or is this because voltage harms the cpu directly? In other words, can I simply push it to -lets say- 1.7vcore if my temps are stable at 70ish (c) max?

2. Is there an easy way to find the "sweetspot" for the machine I'm using? Sometimes it feels pretty stable but I'm very sure I'm not using the full potential of my motherboard therefore I'm posibly wasting that potential in these manual tweaks... Is there a more automated version of doing this?

3. What are the requirements for a "stable" system? When can we be sure that the system is stable?

You don't have to answer all at once. Even answering one of those helps me. I'm sure these questions are fairly simple, but it's hard to get answers to these by reading other people's technical issues - it turns to treasure hunting within all this big data. Wanted to ask it directly to here to get some direct answers.

 

[Ubrales]

1. Higher voltages lead to higher heat, and heat will lessen the life of all electronics. Might even lead to failure. Personally, I would not go over the manufacturer's limits and recommendations.

2. Two different CPUs of the exact specs will behave differently as far as overclocking is concerned. This is because the die was cut from different areas of the large (12" dia.) silicon wafer. This is why some CPUs will overclock better and have lower temps.

3. A 'Stable' overclock is when the system runs fine at the desired OC with no BSODs.

 

[footbrake4life]

I had a lot of questions when I first started and when I read THIS[strike][/strike] it really helped me understand more about overclocking. I wouldn't recommend just upping the voltage of your CPU right away, you need to up the multipliers first and once your build becomes unstable then you boost your voltage. I would not recommend boosting your voltage more than .025, if your clock is still not stable after you boost the voltage than you can do 2 things. You can reduce the multiplier by .5 until you reach a stable clock speed. Or you can boost your voltage by another .025 I personally just decreased my multiplier until I found a stable clock. How you can check your systems stability is by running a program like prime95 for 10-30 minutes. What it does is it maxes out your processor to see if your processor will be able to handle it if that were to ever happen. Also if your temps get too hot I would shut down and decrease the multiplier.

On a side note, what aftermarket cooler are you running?

 

[utisz]

On a side note, what aftermarket cooler are you running?

I'm using a Corsair h90.

One question: I was using manual settings for the voltage. As far as I can understand, 0.025 you mean is for "offset" setting? I have never experimented with offset yet, everything I red so far pointed me to the direction where I have to set the fixed voltage first...

 

[utisz]

1. Higher voltages lead to higher heat, and heat will lessen the life of all electronics. Might even lead to failure. Personally, I would not go over the manufacturer's limits and recommendations.
.

What if even at very high voltage everything's cool?

 

[Ubrales]

1. Higher voltages lead to higher heat, and heat will lessen the life of all electronics. Might even lead to failure. Personally, I would not go over the manufacturer's limits and recommendations.
.

What if even at very high voltage everything's cool?

It normally don't work that way. Higher voltages will lead to higher temps everything else being the same.

 

[utisz]

It normally don't work that way. Higher voltages will lead to higher temps everything else being the same.

I know, what if I can cool it down, I mean. Lets say, down to 65 degrees in 1.75v, if this was possible, would it be safe?

What I'm trying to understand here is - is the temperature the only thing that harms CPU in a high voltage situation? Or, even if we can cool everything down, does the voltage (by itself) harm the cpu?

 

[footbrake4life]

On a side note, what aftermarket cooler are you running?

I'm using a Corsair h90.

One question: I was using manual settings for the voltage. As far as I can understand, 0.025 you mean is for "offset" setting? I have never experimented with offset yet, everything I red so far pointed me to the direction where I have to set the fixed voltage first...

I mean when you get to a point where you have to up your voltage only raise it by .025 at a time. An example: my core voltage was at 1.35v but I had to raise it, I didn't want to jump to 1.4v because that may fry my CPU. So I bumped mine up to 1.375v and it was stable. So if you jump your voltage by .025 every time you have to raise it there is less likely of a chance that you will over volt and ruin your CPU.

 

[Ubrales]

It normally don't work that way. Higher voltages will lead to higher temps everything else being the same.

I know, what if I can cool it down, I mean. Lets say, down to 65 degrees in 1.75v, if this was possible, would it be safe?

What I'm trying to understand here is - is the temperature the only thing that harms CPU in a high voltage situation? Or, even if we can cool everything down, does the voltage (by itself) harm the cpu?

If you can cool it down then it is ok as long as you do not exceed the manufacturer's voltage limit. The manufacturer has set the voltage limits based on overall design considerations of which the temp is one factor. The temp is easily measurable compared to other design parameters.

Use of dice to cool down the CPU under severe OC is more for the lab and not for general use in a PC.