I like your idea of one video card at a time since that will drop the load from the power supply but at 93C I wonder why you don't suspect cooling.
For years I've pulled heatsinks off to refresh the compound and clean as I reassemble. Then I can check fans and leave the case cover off for a test.
Keep in mind those power supplies don't age well. That's why dropping the load is a good idea.
Trying to get a second opinion before I start replacing parts.
My computer recently started restarting in the middle of games like Crysis and even in Starcraft 2, especially during cinematics. There is no error message or blue screen. The computer just suddenly powers off and restarts, telling me that it was improperly shut down and asking if I want to start in safe mode.
Since the problem is only happening during game, I was figuring it was video card related. I already uninstalled/reinstalled drivers and still the same problem. I suspected it might be an overheating issue, its a pain to find a good temperature monitor that logs temps that I can check after a crash.
I tried a trial of HMonitor. I think I found the problem. I have two Sapphire HD4850 cards set up crossfire. The monitor is showing the first card getting as hot as 93 C. My second card stays in the 50's C. The fans are fine. The monitor even shows the fan speeds, the speed even increases to 100% as the card heats up. I can see the fan is running fine(clear case, and dust was recently blown out). I may have to buy the monitor to check temperature at time of crash, the trial will only run for a few minutes.
So, will an overheating video card cause a computer to restart like I am experiencing? There doesn't appear to be any case cooling issues, would a defect in the video card cause the overheating?
I guess there is other trouble shooting I can do like running games using one video card at a time.
Would appreciate any suggestions or advice. Even advice on a replacement video card. Its been a couple years since I have even looked at what hardware was out there.