I was monitoring the temps of my GF 6600 while running it at a known unstable value of 450 MHz. The last read out from RivaTuner is 73 deg C, before the card fails, the screen goes blank, and the system freezes. So, first question, is this a typical temperature for a GPU to fail, and is it normal for the card to be fine right up until the critical temperature and then suddenly stop without warning? All the stuff I read seems to suggest that you see some artifacts on screen as overheating begins, giving you warning long before failure.
Second, how can I improve the overclocking without actually spending any more money? It's a chaintech card, and the actual cooler on it is decent. Not amazing, but not bad. I also replaced their badly applied thermal goop with silver, and I put heatsinks on the RAM. Short of buying a new, beefier GPU cooler, is there anything else to be done? I bought a 6600 instead of a 6600GT intentionally, since I wasn't willing to pay for the extra performance, but I'd like to squeeze what I can out of this card.
Second, how can I improve the overclocking without actually spending any more money? It's a chaintech card, and the actual cooler on it is decent. Not amazing, but not bad. I also replaced their badly applied thermal goop with silver, and I put heatsinks on the RAM. Short of buying a new, beefier GPU cooler, is there anything else to be done? I bought a 6600 instead of a 6600GT intentionally, since I wasn't willing to pay for the extra performance, but I'd like to squeeze what I can out of this card.