Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I know how to overclock my CPU (mess with multiplier, and bus speed)... However, I've noticed that it seems a bit more complicated with GPU's.

How and where do I start? I've noticed that I can adjust the GPU clock speed in my BIOS.

Card I'm overclocking:

I found that memory bus speed is (Mem Speed * Bus width) / 8. So obviously a good way to overclock the memory bandwidth is to adjust the memory speed.

Now, GPU speed is 550 Mhz. How do I find its speed as well? Do I multiply it by the bus width (128)?

What is ideal GPU speed relative to memory bandwidth?

share|improve this question
OC'ing your GPU compared to your CPU is easy. Just go up in small numbers like Sathya said do some stress testing (3D Mark). If your system crashes or gets artifacts back your OC off a bit and try again. – ubiquibacon Jan 11 '11 at 4:30

What is ideal GPU speed relative to memory bandwidth?

There's no "ideal" speed. It depends - on the card, the cooling system used, whether it's factory overclocked, the yield of the chip, the power supply in action, the ambient temperatures et al.

The "only" way is to use a software like RivaTuner and increment the core & memory clocks in 5-10Mhz steps, use FurMurk & check for temperatures and artifacts. Once you think you've hit a decent clock speed, fire up 3D Mark 06 and let it run for couple of hours, or till system crashes - whichever happens first.

IF it crashes, reduce the clocks, rinse & repeat

share|improve this answer

You must log in to answer this question.