I'm not sure whether this goes in nVidia or water cooling. Apologies if I guess wrong, please move if needed.
My 660ti got hit by a non-negligible amount of water during the recent installation of my new CPU/MB/RAM due to faulty/lazy/improper leak testing. I waited for everything to dry out afterwards while I continued to leak test and fix issues properly. On first fire up afterwards, the display had display glitches along the bottom tenth of the screen. I immediately shut everything down and waited longer. The next day I fired it up again and got a beep code indicating faulty graphics RAM. I removed the card from the slot, disconnected power, and installed my 750ti in an available PCIe slot, leaving the card hanging. After a few days, I took out the 750ti and reinstalled the 660ti. And it worked!! I played games on it, browsed the web, no issues. I probably used it for about 3 hours before shutting the rig down and going to bed.
The next morning it wouldn't POST.
I removed the 660ti again but left it hanging without power again so I wouldn't have to modify my loop just yet. That lasted a few days as I started to evaluate my options. Yesterday afternoon I decided to give it one more try. I removed the 750ti and installed the 660ti again.
And it worked.
I've run Heaven for an hour (no problem), I ran GPUgrid tasks on it overnight (tasks validated), I gamed with it for a few hours yesterday and today (no crashes or glitches). It seems to run just fine. I'm terrified to shut it down overnight at the moment though.
So, the question is: if I shut it down tonight and it doesn't come up tomorrow, what does that mean? Is it something that might be fixable somehow? Or is it just plain toast and I should forget about it? I inspected the PCB and there are no burn marks or scorches, no smoking gun to speak of. The whole thing has me slightly puzzled and I'd appreciate any thoughts on the matter.
My 660ti got hit by a non-negligible amount of water during the recent installation of my new CPU/MB/RAM due to faulty/lazy/improper leak testing. I waited for everything to dry out afterwards while I continued to leak test and fix issues properly. On first fire up afterwards, the display had display glitches along the bottom tenth of the screen. I immediately shut everything down and waited longer. The next day I fired it up again and got a beep code indicating faulty graphics RAM. I removed the card from the slot, disconnected power, and installed my 750ti in an available PCIe slot, leaving the card hanging. After a few days, I took out the 750ti and reinstalled the 660ti. And it worked!! I played games on it, browsed the web, no issues. I probably used it for about 3 hours before shutting the rig down and going to bed.
The next morning it wouldn't POST.
I removed the 660ti again but left it hanging without power again so I wouldn't have to modify my loop just yet. That lasted a few days as I started to evaluate my options. Yesterday afternoon I decided to give it one more try. I removed the 750ti and installed the 660ti again.
And it worked.
I've run Heaven for an hour (no problem), I ran GPUgrid tasks on it overnight (tasks validated), I gamed with it for a few hours yesterday and today (no crashes or glitches). It seems to run just fine. I'm terrified to shut it down overnight at the moment though.
So, the question is: if I shut it down tonight and it doesn't come up tomorrow, what does that mean? Is it something that might be fixable somehow? Or is it just plain toast and I should forget about it? I inspected the PCB and there are no burn marks or scorches, no smoking gun to speak of. The whole thing has me slightly puzzled and I'd appreciate any thoughts on the matter.