Re: Is my PSU responsible for freezes? Possible bad caps?
I concur.
Looks like the video card is just starting to fail. I have a few "finicky" video cards like that myself.
Exactly.
To be precise, anything over 60°C is already getting dangerous. Tin-whisker growth is one reason (hence why it can take a few years for a reflow to go bad again). But also, Lead-free solder is just too damn brittle.
Think about why CPUs almost never fail - it's because people frown when they see temperatures above 60°C on their CPUs. And 70°C? Most people will panic and tell you that is way too much. Funny how they don't think the same about their GPUs, though. I guess nVidia and ATI did a good job of convincing people that it is okay to have your GPU running at scalding temperatures.
I think that's pretty much a myth nowadays. If the PSU can't supply enough power, it would either shut down (under-voltage and/or over-power protection) or something will blow up / burn. In most cases, even out-of-spec ripple won't crash a modern motherboard, as virtually all of the secondary voltage rails are regulated (be it linear or buck). Out-of-spec ripple used to be a problem only back in the days when you had things like SDR RAM and Southbridge powered directly from the 3.3V rail and the like.
I concur.
Looks like the video card is just starting to fail. I have a few "finicky" video cards like that myself.
Exactly.
To be precise, anything over 60°C is already getting dangerous. Tin-whisker growth is one reason (hence why it can take a few years for a reflow to go bad again). But also, Lead-free solder is just too damn brittle.
Think about why CPUs almost never fail - it's because people frown when they see temperatures above 60°C on their CPUs. And 70°C? Most people will panic and tell you that is way too much. Funny how they don't think the same about their GPUs, though. I guess nVidia and ATI did a good job of convincing people that it is okay to have your GPU running at scalding temperatures.
I think that's pretty much a myth nowadays. If the PSU can't supply enough power, it would either shut down (under-voltage and/or over-power protection) or something will blow up / burn. In most cases, even out-of-spec ripple won't crash a modern motherboard, as virtually all of the secondary voltage rails are regulated (be it linear or buck). Out-of-spec ripple used to be a problem only back in the days when you had things like SDR RAM and Southbridge powered directly from the 3.3V rail and the like.
Comment