I’ve never had these kind of issues in the early stages of MSFS. Then after a long hiatus I’ve recently came back into MSFS and I have the problem I’m going to explain.
I use Rivatuner to monitor the resources MSFS uses on my hardware and I’ve noticed that now and then, apparently randomly but more frequently in intensive areas like London or on ground in a payware airport or in places with many AI traffics, my gpu fans accelerate from 2500rpm to 3500rpm/3800rpm for a few seconds before settling down again. These spikes do not affect my fps, do not cause stutters and doesn’t seem to be caused by a spike in GPU temp as Rivatuner shows it remains pretty consistent.
I use many other videogames and I’ve never had this issue with anything else than MSFS.
PS: my settings are Vsync on, 60fps limit set inside MSFS and my GPU is a RTX2070S with the latest drivers.
The issue is that rivatuner is not showing any sign of overheating. Temperatures stays around 80/85 degrees and the cpu fan spike last only a couple of seconds
Yeah, as said before the fan keeps your hardware from overheating so nothing is shown.
If the temperatures tend to rise as load on the GPU increases the fans will spool up before the temp goes up beyond limits.
The only way to deal with this more comfortable is to set up custom speed curves to let the fans turn some rounds faster on lower temps - in general there is otherwise nothing wrong with the behaviour imho.
Compared to other games MSFS pushes hardware easily to limits, especially as the sim grows in features/content
I’m a bit worried about putting too much stress on my GPU as the fans are really revving up and really loud, but if you say that’s normal, I guess I’ll have to live with it.
GPU is designed to run flat out 100% all the time, 80-85 is somewhat toasty (I think that series of card throttle at 85), id personally set up a custom fan curve and gets those temps down to the mid 70’s (at cost of noise), but, as others have mentioned, it will run happily all day everyday like that.
Normally GPU doesn’t go as up as 80/85 degrees, but it’s very hot where I live so I guess that plays a big role. I’ll try to play with the fan curve to see if it gets better
My GPU is two years old and I don’t think dust is an issue as I’ve cleaned my computer not long ago. Could thermal paste decade in such a short period?
I had previously GPU that lasted 10 years without issues.
It’s possible the paste has degraded but not likely. It’s more likely the fans are revving to keep temps below your max. If you’re using RivaTuner for stats, you’ve probably also got MSI Afterburner installed. Load it up during a sim session and check the fan curve settings. While you’re at it, have it set to log the temps over time and see if there are quick spikes you’re not noticing.
If you haven’t already, also find and download GPU-Z and see what it reports for your card’s max temp/throttling point. My RTX 3070 is set to 83°C by default, though I can adjust that up to 90°C if I wanted to push things.
Try lowering your vsync to 30fps. I run like this all the time as I find it the best experience overall on my machine even though I can much higher framerate with vsync off. This consderably lowers the GPU workload and hence the temps.
Ok, I’m not saying this is your issue for sure. This happened to me about 6 months ago. Fans ramping up to 100% for a few seconds, then winding down. This happened abruptly- not gradually over time. My card at the time was a Gigabyte Aorus Xtreme 2080ti. I took it apart and repasted the TIM. The old application of TIM was shrunk and drier than a… Well, you get the picture. Problem solved, and card’s fan profile returned to normal. Matter of fact the card is still in use on another system now,
There have been reports last year with the fan issue cropping up on 2000 series RTX cards, due to the TIM drying out, although this can happen to any card really.
Try other GPU intensive games or apps, and see if the issue persists. If so, maybe it’s time to repaste the card’s chip. In any case it cannot hurt, considering of course care is taken to do this properly.
Failing that, keep in mind that the SU10 Beta will likely work your GPU harder, with DX12. So there’s that.
True, however this does not address the root of the problem. The OP has stated that this just started occuring to his GPU. If a new application of TIM is warranted, then undervolting will simply mask the underlying issue.
Thanks for all the suggestions. For the moment I tried modifying the GPU fan speed curve and it seemed to work. I’ll test more in the next days.
The fact that this issue is appearing only with MSFS makes me think it’s not an issue with the GPU because I use other demanding games without any issues.
