PC Graphics

[franky97]

I've bought and MSI R9 290 Gaming OC 4GB and it can seem to be able to run below 95 °C under load. Is it a serious problem? And I think the card is overheating at this point because I suffer from heavy stutter and frame latency... Even on Unigine Valley and with the dual fans at full speed... please help

[elstaci]

If you bought the card used and it is not under Warranty, I would suggest replacing the GPU's Thermal Paste which is probably not working as it should be.

According to this website the Maximum Operating Temperature for the R9 290 is 94c.  So you are getting close to maximum or past the maximum temperature which will cause the GPU card to throttle (slow down) to prevent overheating and being damaged.

[franky97]

Yeah it was a used card I bought, I dismounted it, cleaned it and replace the thermal paste but it keeps climbing to 95 °C like 3 minutes after I lauched even Skyrim I don't understand...

[elstaci]

Since it is a Used GPU card, it is possible it is either a fake or the vBIOS was modified by the original owner.

One way to check is by downloading and installing a free app called GPU-Z.

Run it and upload a image of the GPU-Z and lets compare the vBIOS version and your GPU card specs from what is posted by the website of GPU-Z.  If the specs don't correspond to what is posted for your GPU card, it will be a good indication that it may be fake or the vBIOS was modified or replaced by the original owner.

Do you know the manufacturer of the GPU card? Hopefully GPU-Z will show it also if your don't know. 

Here is my GPU card from GPU-Z:

[franky97]

Thank you for the advice I downloaded GPU-Z here is a screenshot of what it shows me. On Techpowerup it says it's the last bios made for the card.

[elstaci]

Thanks, Seems like your GPU card is legitimate. GPU-Z specs corresponds with TechPowerUP website specs for your GPU Card.

Try running a diagnostic Stress test on your GPU and see what it shows as far as temperatures and voltages.

I personally have a preference for OCCT but there are others that are just as good or better at stress testing the gpu card.

Not sure if AMD Wattman is compatible with R9 290, but if it is you can, Configure the GPU fans to run at its best speed depending on the GPU's temperature.

Note: before doing any stress test, try installing a previous AMD Driver and see if your GPU card still overheats. Try downloading the AMD Driver for your GPU Card from MSI Support and see if it runs cooler. If it does, then try installing a new AMD driver until you find out which driver is causing your R9 to overheat. If it still overheats with the older AMD Driver then try stress testing it and observe the GPU fans that they are running at maximum speed when it starts getting close to 70c.

[franky97]

I looked on MSI support, downloaded the only driver but it was an amd driver from July 2018, I installed it, I even pushed my GPU fan on max before lauching unigine valley, same story... 95°C in les than a minute and horrible frame latency and stutter... I don't know what to do... I'd maybe try put on different thermal compound... Thank you so much for trying to help me

[elstaci]

I beginning to suspect your GPU card may be defective.  Possibly the GPU sensors are not reporting the correct temperature.  Unless the GPU Cowl Heatsink is not making good contact with the GPU chip or it is dirty.

If the heat sink is installed correctly and the GPU Fan is running at Maximum and the GPU card is still overheating, it beginning to sound like a hardware issue since no matter what driver you install it still tends to overheat. It shouldn't reach such high temperatures while gaming. The GPU card is throttling (Slowing down it speed) to prevent the temperature from going past its max operating temperature of 94c.

Leave one side panel off your computer case off and see if it still overheats. It is possible it could be poor air circulation in the computer case.

What are the GPU temperatures when just browsing or when the computer is at idle?  It should be in the 40's to lower 50's C.

EDIT: Need to mention something, when the GPU Card is running hot or overheating are both the GPU FANS WORKING at MAXIMUM? If only one is working that would cause the GPU to overheat under heavy loads but not under normal or idle situations. Visually check to make sure both Fans are running at maximum. Also check to see, safely, if the GPU card is blowing hot air or is physically hot to the touch when it is running very hot. If it is fairly cool when it is running a temperature over 90c then it is a good indication a defective GPU Card.

You can always confirm this by installing the GPU Card in another compatible computer and see if the exact same thing happens.

[pokester]

Absolutely try the new thermal compound. The dry or shrink. Sometimes loosing complete contact with the heatsink. Even pads on the memory and vrms can can cause issues. So check it all out. Carefully remove the heatsink. A lot of times if you don't mess them up removing the sink you can reuse the thermal pads on the memory and vrms by just adding a little compound to them. Completely redo the compound on the GPU and evenly put torque the heatsink back in place. Back sure to wiggle it in on the gpu a bit before securing to make sure you have a good bond on the compound. Or you could go online and order a whole set of thermal pads for it all.

Hopefully that is all it is and not a failing GPU. Good Luck!

Great advice BTW elstaci! As usual!