Question GPU Idle & load temp.

[quadrax]

Hi there.

I am experienced as PC user, but still have one question which is hardly debated on internet and also it's interesting for me. Got a MSI GTX 1660 Ti, which is great GPU, but after a time i decided to set-up my own Fan Curve profile (i really didn't want wear fans out quickier, because of start on / off spinning constantly, if i am playing not demanding games, but in some situations, it can raise temp above the Temp limit, so fans kicks on).

Got a curve which i set for 20% fan speed (0-50°C) then it rise up higher. So my idle temps are (actually at ambient temp 24°C) on GPU : 34°C and in Load it's about 62-63°C. ( Note : 20% fan speed is minimum, because less than this, FANS are turning on / off, or stays off.)

The curious part is for me, what / or how much variety in temperature fluctuation can harm PCB or soldering with this temp fluctuations. For example i play Doom for a 30 minutes, temp raise at maximum 63°C, but then i decide to quit and just for a hour looking on youtube / or go away from pc. If this happens in a cycles (more often). Can do this in long term (under warranty condition and duration) any damage to a card?

Maybe i am just paranoid, but i would like to get a proper answer or reliable informations about this, or source. I am getting lost to try any information about it.

( Not counting Heat or Voltage as killing factors, because that's what i already know ). BTW Got undervolted that gpu from 1.043 to 0.931mV and seems like it's stable at same clocks.

I just find information these fluctuations (like i mentioned) are not great for electronics, but in which range and so on?

Thanks for reply.

 

[Barty1884]

The "cycles" like that are not going to harm anything. At the end of the day, whether 30 mins on, 30 mins off (rinse and repeat) or 24hours on, 30 mins off, the card is perfectly capable of handling that.

Any GPU should be able to run at it's max rated speed, 24/7 for it's warranty period, with stock fan curves out of the box. Obviously the lower the better in terms of temperatures ,but running at ~60'C is not going to be a problem.

 

[quadrax]

Allright, thanks for the answer. My main question was right about that "Temp cycles" between game playing and then doing something less demanding on gpu. And if these cycles repeats ( it was just a raw example). So now i am then more satisfied with that if it's not going to be any trouble so far (with that example in my case - 34°C idle and then load to 63°C) or so.

Thanks once more time.

What you think could be more harmful for GPU - which range with these cycles ( or more often cycling ) ? I am asking, because i would be more wiser for the future as i can't find any evidence of that.

 

[quadrax]

bump. Just a quick recapitulation - How much range (of heat cycling) can damage GPU in long terms? And how much quickly? Lets say for 20 min, 63°C and then in 5 mins GPU getting lower temp to mid 30°, and then again (for example i came back to a pc) and start again playing.

Thanks!

 

[quadrax]

last time bump, then i give up. 😀