I remember looking into this way back when I got a GTX 480 around launch; what I found at the time was that Nvidia set up their drivers to go into full 3D clocks (P0 state) when more than one monitor was connected. I found it excessive, but I soon dropped the second monitor so the idea faded. Now that I'm back to two, I have questions.
1. Does anyone know of a workaround to prevent this? Possibly putting the GPU into P8 instead of P0 so at least it's only marginally higher clocked.
2. Are Kepler cards also going into P0 when more than one monitor is connected? Or have they implemented some proper power-saving?
3. Just for testing, I went into NVCP and disabled the second monitor. My clocks dropped to P8, then P12, but then the following happened. Any reason why it would keep... "oscillating"?

1. Does anyone know of a workaround to prevent this? Possibly putting the GPU into P8 instead of P0 so at least it's only marginally higher clocked.
2. Are Kepler cards also going into P0 when more than one monitor is connected? Or have they implemented some proper power-saving?
3. Just for testing, I went into NVCP and disabled the second monitor. My clocks dropped to P8, then P12, but then the following happened. Any reason why it would keep... "oscillating"?










