How To Overclock

[mtdewcowboy]

I know this comes up all the time but what I would like to know is what is the best way to overclock for optimal use not max freq.
Do you just do the multiplier or the cpu/nb or is it both
I am just trying to head around the ins & outs of maxing out my computer

 

[trents]

Whether for optimal use (whatever you mean by that) or max frequency the answer is the same: Usually, the most effective approach to overclocking is to use a combination of increasing the FSB frequency and increasing the multiplier.

 

[RGone]

Optimal Overclock ?

Optimal the adjective might be optimally defined as:
1. >> Being the best or of the greatest value, often within certain parameters or restrictions.

2. >> Having the greatest degree or best result obtained or obtainable under specific conditions.

Also might be said as the most favorable conditions being met. There are favorable conditions for plant growth. Temperature. Light. Moisture. Taken just one at a time a plant may get too much light. It might have tempertures too high or low. Or the plant may be in soil that has too much moisture for good growth.

Now if we begin to think like that in terms of overclocking, there will be a great difference in the end result we strive for that is almost totally different than the apparent mind-set of most all first time visitors asking how to overclock their AMD system.

You can tell by the constant struggle to raise cpu mhz that most are not interested in reaching anything resembling an optimal cpu speed. Most are far from understanding that most of us that have overclocked in this century and also in the last century, know that the maxed-out cpu speed we obtain is seldom the speed we would maintain day in and day out or for 24/7 use.

Optimal cpu speed would be reached within certain limits. Easily to mind lately is for sure the amount of heat generated by our speed. Of course the amount of voltage necessary to bring about any semblance of stability, which is a great contributor to the output of heat.

I will insert a thought of my own here. I can quite easily remember if you could not run 8+ hours of Prime 95 Blend; there was no way under the sun that your sytem could be considered stable. Now we glibly support a 2 hour run of Prime 95 Blend as just plenty of stress time in order to stamp our system as stable.

It seems without much thought or consideration we can widen our chance of an error by suggesting that a shorter time under stress is just fine for labeling our systems as stable. Reminds me of a saying I have heard and not liked for years. Oh heck it is close enough for ********** work. Going back to the idea of something being optimal, with conditions, then maybe that is good enough.

Now since this question of optimality is being asked in the AMD CPU forum section; a quick scenario comes to mind. What might be the optimal speed of say a BullDozer FX-8120 or FX-8150? Well you know what? There is an answer to that question that has begun to surface amongst the optimal thinkers in this forum section. It is ~ 4.3Ghz.

4.3Ghz? The heck you say! Well you did say running my system optimally did you not? Well 4.3Ghz is a speed point at which the amount of Vcore, the output of heat that must be dealt with and the diminishing performance increase for added cpu speed all seem to hit an optimal point. Faster than 4.3Ghz and the performance increase begins to level off for each increment of faster cpu speed. The output of heat from a FX-8120 or FX-8150 is considerally more managable at 4.3Ghz. Also the voltage to stabilize an FX-8120 or FX-8150 is reduced drastically when not pushing cpu speed beyond roughly 4.3Ghz. The heck you say. That sounds nearly like the optimal place to run those two cpus. Well you might say that.

I speculate that we don't deal with overclocking questions in an optimal manner in this AMD CPU forum section. I can honestly say that I know of only one person that was most likely helped in an optimal manner since I know the person doing the helping put the one requesting help in overclocking, on the 'path' to an overclock but then released that helpee to find his own optimal overclock. He did not hand-hold the requester of help to a Max overclock like he was training someone for the benchmark team. Max it man Max it. And no after those that help in the AMD CPU forum section spend all that effort to try and help each user to his MAX overclock; I wager not even 0.05% of those helped to their MAX overclock ever join the benching team.

So when you begin to think in terms of being in an optimal condition or state, there are many things that must be considered. An older adage comes to my mind. Nothing done to excess is perhaps the best course of action.

 

[juane414]

Optimal the adjective might be optimally defined as:
1. >> Being the best or of the greatest value, often within certain parameters or restrictions.

2. >> Having the greatest degree or best result obtained or obtainable under specific conditions.

Also might be said as the most favorable conditions being met. There are favorable conditions for plant growth. Temperature. Light. Moisture. Taken just one at a time a plant may get too much light. It might have tempertures too high or low. Or the plant may be in soil that has too much moisture for good growth.

Now if we begin to think like that in terms of overclocking, there will be a great difference in the end result we strive for that is almost totally different than the apparent mind-set of most all first time visitors asking how to overclock their AMD system.

You can tell by the constant struggle to raise cpu mhz that most are not interested in reaching anything resembling an optimal cpu speed. Most are far from understanding that most of us that have overclocked in this century and also in the last century, know that the maxed-out cpu speed we obtain is seldom the speed we would maintain day in and day out or for 24/7 use.

Optimal cpu speed would be reached within certain limits. Easily to mind lately is for sure the amount of heat generated by our speed. Of course the amount of voltage necessary to bring about any semblance of stability, which is a great contributor to the output of heat.

I will insert a thought of my own here. I can quite easily remember if you could not run 8+ hours of Prime 95 Blend; there was no way under the sun that your sytem could be considered stable. Now we glibly support a 2 hour run of Prime 95 Blend as just plenty of stress time in order to stamp our system as stable.

It seems without much thought or consideration we can widen our chance of an error by suggesting that a shorter time under stress is just fine for labeling our systems as stable. Reminds me of a saying I have heard and not liked for years. Oh heck it is close enough for ********** work. Going back to the idea of something being optimal, with conditions, then maybe that is good enough.

Now since this question of optimality is being asked in the AMD CPU forum section; a quick scenario comes to mind. What might be the optimal speed of say a BullDozer FX-8120 or FX-8150? Well you know what? There is an answer to that question that has begun to surface amongst the optimal thinkers in this forum section. It is ~ 4.3Ghz.

4.3Ghz? The heck you say! Well you did say running my system optimally did you not? Well 4.3Ghz is a speed point at which the amount of Vcore, the output of heat that must be dealt with and the diminishing performance increase for added cpu speed all seem to hit an optimal point. Faster than 4.3Ghz and the performance increase begins to level off for each increment of faster cpu speed. The output of heat from a FX-8120 or FX-8150 is considerally more managable at 4.3Ghz. Also the voltage to stabilize an FX-8120 or FX-8150 is reduced drastically when not pushing cpu speed beyond roughly 4.3Ghz. The heck you say. That sounds nearly like the optimal place to run those two cpus. Well you might say that.

I speculate that we don't deal with overclocking questions in an optimal manner in this AMD CPU forum section. I can honestly say that I know of only one person that was most likely helped in an optimal manner since I know the person doing the helping put the one requesting help in overclocking, on the 'path' to an overclock but then released that helpee to find his own optimal overclock. He did not hand-hold the requester of help to a Max overclock like he was training someone for the benchmark team. Max it man Max it. And no after those that help in the AMD CPU forum section spend all that effort to try and help each user to his MAX overclock; I wager not even 0.05% of those helped to their MAX overclock ever join the benching team.

So when you begin to think in terms of being in an optimal condition or state, there are many things that must be considered. An older adage comes to my mind. Nothing done to excess is perhaps the best course of action.

Aristotle? Is that you?

 

[mtdewcowboy]

Thanks for the insight Rgone
4.2 to 4.3 is about the max I was thinking

 

[mtdewcowboy]

Whether for optimal use (whatever you mean by that) or max frequency the answer is the same: Usually, the most effective approach to overclocking is to use a combination of increasing the FSB frequency and increasing the multiplier.

Do you do them evenly or is it less multi and more fsb or the other way around?
Sorry if I seem to ask too many ?'s but I want to learn the basics and then find my way on my own

 

[RGone]

Yo "mtdewcowboy", you have two stars by your screen name.
How to put up Sig at OCF.

New Shortcut method for putting a Signature with your system information following your every post so people can know what is in the case that they are trying to assist with. You can use something like what is shown in my signature as a good template of needed information Thank you.

How much of which is usually determined by the system itself. Also by ram and where you want your CPU_NB to hit frequency wise. For me I use mutlipier 19.5 and the FSB determines how far the overclock will actually go. I go 200 up to about 240 FSB and let my ram speed and everything else rise as the cpu speed does. But how much of which is really a personal thing I would think.

 

[mtdewcowboy]

Sorry I posted using phone so sig don't show
Maybe with this edit it will?

 

[mtdewcowboy]

Here is my specs

 

[mtdewcowboy]

Yo "mtdewcowboy", you have two stars by your screen name.
How to put up Sig at OCF.

New Shortcut method for putting a Signature with your system information following your every post so people can know what is in the case that they are trying to assist with. You can use something like what is shown in my signature as a good template of needed information Thank you.

How much of which is usually determined by the system itself. Also by ram and where you want your CPU_NB to hit frequency wise. For me I use mutlipier 19.5 and the FSB determines how far the overclock will actually go. I go 200 up to about 240 FSB and let my ram speed and everything else rise as the cpu speed does. But how much of which is really a personal thing I would think.

Now the light bulb just went on
Thanks for the insight it really let me look outside the box to see the whole picture

 

[RGone]

That forum app they are pushing ought to make a note that the post was from a phone like some forums do. I would not have mentioned it but I glanced at sig location and nothing just blank space.

Yes I really enjoy this CHV since I hardly have to change anything to go from 4.1Ghz to 4.8Ghz and just raise the FSB/CPU frequency. Oh a voltage bump here and there but otherwise so far ahead of the lesser boards. Your Sabertooth should do pretty darn close to the same once you spend some time just playing in the bios and testing the result.

C_D that posts in here as well, swears his Sabertooth is better than his CHV at many things. So you should be in good board land.

RGone...ster.