Hi, My GPU a MSI R9 290 4GB Gaming is overheating and thus also underclocking itself quite a lot. It will reach 95C in WoWS for instance and even reach 85C in hearthstone and about 51C when completely idle. Watching youtube will get it up to 60-65C. This has been a problem ever since the Win10 Amd drivers were released i august. Before that everything seemed fine and it was near silent when doing light tasks. I have had the card for a bit over a year now. I have tried downgrading but it won't let me install older drivers for some reason it only wants to install the catalyst software and not the actual driver I was using before. No idea what is happening really.
What I have tried: - Reinstalling drivers, trying to downgrade and beta drivers and nothing helps. - Running the computer with the side open and it still reaches insane temps. - Underclocking the card to 500MHz on the GPU and 650MHz on the memory to no avail. Though underclocking the memory at least had some impact, 1-2C~ but it still throttles when doing other stuff. - Removing ram and seeing if that was the culprit but nope. - Reinstalled Win10 without any changes.
Components: CPU: AMD 8350 Memory: 4*4 Corsair 1800MHz vengence LP GPU: MSI R9 290 Gaming 4GB Motherboard: Gigabyte 990FXA-UD3 PSU: 550W Seasonic, can't remember the exact model name but I think it's an M12 Case: Fractal R4 with 3*140mm fans.
Any help appriciated. Willing to try almost anything at this point.
I would get a monitoring program such as gpu-z or msi afterburner to check the fan speeds while you are playing games. If it is staying at a low fan speed you can use afterburner to force a fan speed or make a custom curve. Also see if there is a bunch of dust blocking airflow and clean it out.
I have tried checking for a BIOS with the MSI app for updating stuff but that can't find any new versions of anything but MSI afterburner.
And I do have MSI afterburner installed which is what I use to underclock the card and general monitoring. I have saved logs so that I could backtrack a while to see what the graphs were showing. And leaving the side open which should negate the airflow thing a bit as well didn't change anything really.
Running 15.12 since I update immediately hoping they would have fixed something by some miracle. The fans are running at 22% when idle. I think they should be at 18% while idle if I remember correctly.
Tried changing the thermal paste? Check if the cooler is fit properly, sometimes when handling a card you can "loosen" the thermal paste by grabbing the cooler, making tons of air pockets between the silicon and heatsink.
As it as new a year ago and I haven't touched it since and that the problems appeared over night more or less I have a hard time believing that it would be something like that. Don't really like fiddling with the GPU like that either. Have swapped a cooler on an 6870 before and there were a lot of screws etc. I would rather in that case hand it in to where ever I bought it and tell them that it isn't working as it should and play stupid. Only problem with that is that I would be without a computer for a couple of weeks most likely >_<
There is a few things that can make a card behave like this, I already stated one and I had a card do that on me once.
There might be a failed temperature sensor on the die, which I find extremely unlikely since its solid state and all that. The heat pipe magic fluid might have escaped the tubes from a sudden leakage, which is also extremely unlikely since you got many of them, if 1 failed it wouldn't throttle the card that hard. There might be a software/firmware issue, also extremely unlikely.
Tried patching the bios version of the card? Check the version with GPU-Z and download a update from MSI if they have one. With GPU-Z you can also monitor tons of other stuff with your card, like voltage for example.
If you make a backup of your original factory bios and then start to flash you can at the end still flash the original back on to the gpu. And as far i know they cant tell if a change in that area was made.
I had a similar problem bought a used gigabyte r9 290 WF3 OC it crashed amd drivers like every 30 minutes when i realised ti had a mining bios on i searched on gigabyte's website for the original bios i found it flashed it an the gpu was stable. And after a month or so i send the card back to RMA because i was stupid and thought a r9 290 OCed to 1140 and 50% power delivery was overheating..... no shit xD they just send back the same card and wrote the gpu reached 80 degrees.
What is the point of running a 50% fan speed curve when the card has been near silent for a year and then more or less over night got louder. The point isn't to keep the card running cool it's to get it back to where it should be.
I have tried looking for BIOS's on MSI's homepage but can't find any and their live update thingy says everything is up to date etc. Can't really find any more information than that, I'm suppose to run the latest MSI BIOS atm.
