I've recently made my i7 870 stable at 4Ghz using linX, occt, prim95, younameit.
My psu is a corsair 650tx which is a quality brand.
I've been overclocking my PNY GTX 460 now and got to 840 mhz core clock stable with 1.037 volts using OCCT gpu test on 1920x1080.
The problem is that when I raise it to 850 mhz core, my monitor turns off (orange light on) but my computer is still on and running. This happens even if I raise the voltage to 1.062 which is a big jump from 1.032 just for 10 mhz.
My monitor is a Asus VH242H 24".
Another weird thing is that I can bring the memory clock from default 1850 to 2000 mhz and core 840 mhz w/ 1.037 volts and STILL remain stable in OCCT for about 30 min with temperature maxing at 79 degrees and 70% fan speed.
hmm yea that's what I was afraid of, was hoping I could at least get to the FTW version's stock level haha. It didn't happen during my full playthrough of crysis AND warhead at 850 core and 2000 memory nor during other games so I figured it still had some headroom...
From what I can see so far, I think i'll be stuck at 840 mhz and 2050 memory at 1.037 volts
Edit: If anyone feels it's some other issue and not the card's limit, please feel free to add
Oh also during OCCT the temperature gets to around 85 and there are 0 errors about 5 minutes into it before the monitor turns off. (I also hypothesize that the gpu downclocks because the FPS in OCCT goes from 300+ to 60ish when I run it again without a fresh restart of the computer. If this information clarifies some things I hope it does.
I just found something new out. In Furmark when I tried extreme burning mode, with the stable settings, the graphics card downclocked to 405 mhz core clock right around the time it hit 84 degrees... is this coincidence or is there some kinda fail safe that is turning off my monitor and downclocking the card?
In NVIDIA control panel I have the power mode in global 3D settings set to maximum performance.