Bitcoin Forum

Hello,

If I am mining with my GPU 24/7 (R9 280x), at a pretty much constant temp of 65C, and have my fans running around 2000-2500 RPM will it decrease the lifespan of my GPU? How long till it might stop working?

Would water cooling it help my GPU "live longer" by keeping it at low temperatures?

How long has your GPU been running 24/7?

Thank you.

If you keep your GPU at 65C, it shouldn't shorten it's lifetime at all. However, it might decrease the life of the fans on the GPU. I used to run a 7850 24/7 and after a year or so one of the fans began to rattle. You might want to keep that in mind.

So the only thing that is in danger is the GPU fans correct? Do you have any tips to keep the GPU at lower temps, so the fans doesn't have to turn so fast but keep the hashrates as high? I set my Max allocated GPU to 60% should it be lower? Right now 60% or 100% gives me the same hashrates.

You could try adding additional fans to your system to lower the load on the GPU funs, then they could run at slower speeds. You should just make sure your VRM temperatures are in check too - some GPUs have sensors for the VRMs and their temperatures will show up in GPU-z. I'd try and keep VRM temps below 80C.

You could also try downvolting your card - that way it would take less power, and the fans could run slower too.

How about Watercooling my GPU?

Watercooling can still go to dangerous temperatures. Keep this in mind.

Plus, it would be a huge investment and if you plan to make that money back from mining it will take a very long time, you might never reach a full return.

What would you consider to be a dangerous temperature?

I have the NZXT Phantom case. So should I consider buying a side fan that would blow right on the GPU as well as am adittional top fan?

Dangerous temps are generally seen as above 85 degrees for long periods of time.

how long is bad at 85c?

I would personally say periods of more than 6 hours.

Only 6 hours?  :o What capacitors do they use those days?

Protip: not all components/brands are the same.

I have mined at constant temps around 82 degree celsius and GPUs survived for more than 4 months , so 65c is not bad at 2000-2500RPM.

just to state the obvious, the nominal temperature of the gpu shouldn't damage it.
I ran for about a year my 780ti between 78 and 82°C without any damage at all (well the pcb shows at some place shows some slight change in color...). If you go above the standard temp of operation yes you will damage it
 

65 it's pretty normal for winter, just don't go above 80 and don't set fan to 100% or other stupid high % like that, leave them at 50%

I have been running 10 280xs for 13 months now varrying from 70-88c.  Those blake algos are hot...ya so I wrecked some fans (lookin at you MSI) Other than that all good no failures.

your GPU temperature is excellent. So you shouldn't worry about lifespan.

Oh really?  You've got proper answers here but I do wanna show off this... If you're into it, and you can get a couple of el-cheapo cpu water-cooling units, closed loop ones, you can get some very effective cooling happening, which translates to some ripping overclocking.  Here's my setup. (http://i633.photobucket.com/albums/uu53/acatphoto/Tech/BadGirl02.jpg)  The hardest thing to cool was the VRM's which were running 20 deg hotter than the core, note the fan array and heat-sink. (http://i633.photobucket.com/albums/uu53/acatphoto/Tech/BG_Detail01.jpg)

So how does it run?  I can't even get CLOSE to 55 deg on the 7950!  CRANKED!

http://i633.photobucket.com/albums/uu53/acatphoto/Tech/Benchmarks/7950-1200-1600.jpg

http://s633.photobucket.com/user/acatphoto/media/Tech/7950-1240-1740-Firestrike.jpg

I picked up an R9-290 a couple of weeks ago for just over $200 Can.  Buddy must have thought he was getting rid of a lemon because it was IDLING at 80C.  After slapping a block on I now have a 290 that'll do 1150/1500 and stay under 60C.  Sweet!

i used to have a HD7850 working for feb 2014 to sept 2014 at 60 to 65c and nothing happened wrong just the fans strted to make a bit more noise nothing that oil would not solve.

i have old HD5800 and HD5970 cards at constatnt temperature of 85C for a year! and still running at full speed. i think that will never end, so, at 65C it will running forever. Also you got to know, that gpu's needs that temp to run smootly, it is working temperature.
 i have tested mine to run at very low weather temperature outside = -15, and i was strugling to start mining, but after sometime, when they reached 50C gpu start mining normally.
So cold is not always better for gpu's.

I have 4x 7970 that have been running for 2 years, first fan broke yesterday.

When they where mining LTC the temp was about 75c, now with XMR they stay around 55c.

I would stay below 80c and keep fans running on as low a RPM as possible at that temp.

Undervolting can do wonders for fan speed and GPU lifespan.