I'm noticing that my graphics card's performance isn't "up to par" as its supposed to be. I have MSI Afterburner installed but certainly don't want to make my GPU FUBAR. Not necessarily trying to benchmark, just get better performance. When using afterburner, what's the highest recommended clock for each setting? They're currently at:
Core clock: 805 MHz
Shader clock: 1610 MHz
Memory clock: 2100 MHz
Fan speed: 40
I"m using the PNY GeForce GTX 580 Enthusiast Edition Overclocked with an AMD Phenom II X4 840 @ 3.2 GHz (Looking to overclock my CPU as well). My main concern is that when I'm on a game like Mount and Blade, my graphics are strangely low, at around 50-75 fps. With the 580, I should be getting at least 500. A friend of mine uses the Radeon HD 5670 and they're getting 200 fps. What could the problem be?
What are the factory settings for the card ? The thing with overclocking the video card is that there is a limit to what you can go to and the higher the overclock the faster the fan has to spin to cool the extra heat that is produced by the overclock. In you case do you have a side mounted fan that blows directly on the video card. If not then that would be a good place to put one to help out the fan that is on the card. In using afterburner when you increase the clock on the core are you increasing the voltage as well. If you go much higher with the core clock you will have to increase the voltage as well. You may be able to go to 825 on the core but I'm sure that the fan will be at 100%. Increasing the voltage is very risky and should be done in the smallest increments as possible , it doesn't need very much.
To give you an example of what some core clocks are I have a Evga GTX 580 Hydro Copper 2 (that is the one that has the water block on it) and the core clock comes at 850mhz from the factory and the reason is because it is cooled by water which is better than fans , but the point is that the core is clocked at 850 and it runs fine. So the core can go higher it's just a matter of cooling it and having the right voltage also.
Nvidia cards have the shader clock set at 2 X the core clock, and they can be increased without a lot of concern. The problem is with the memory clock, which seems to be very sensitive to tweaking. I was able to OC my GTX 470 fro its factory 605 MHz to 900 MHz without a hitch. But when I tried to ramp up the memory it caused a lot of crashes. I now run my dusl GTX 470 in sli at 650 MHz, without cranking the memory.
I also ramped up the voltage on the card when it was a single GPU. I have now backed it down to factory.
Play with the core/shader clocks to see if you can get what you desire. Ramp up the memory only when you have a stable core OC. Play with the voltage only if your results are not to your satisfaction.