Poor Performance w/ Nvidia GTX 980 Ti FE.

Jun 5, 2018
1
0
10
Hello,

I'm having poor scores in firestrike (~13,200) and userbenchmark with my GTX 980 Ti founders edition card. I bought it cheap from a buddy. It was a bit dusty, but was running about 83 C under load, which I know is normal. It works fine for my needs but I know it should be performing better.

I just opened it up to get more dust out of the heatsink/shroud. I added new thermal paste but didn't change the thermal pads, which looked like they should be replaced, I just didn't think about it when I opened it up. The temps are now about 87 C. I thinking it's probably due to the old thermal pads. ANYWAY, i'm still getting the same scores.

I checked that the cables are plugged in correctly, it's in the correct PCI-e slot, drivers are up to date and the GPU is at default settings.

I'm not really sure what to do next. If you have any solutions or suggestions, please let me know.

My specs are:
Intel 7700k @ stock
Asus Strix Z270 E Mobo
Corsair H100i V2
16 gb of g.skill 16gb 3000 ram
Samsung evo 850 250 g
Seagate Barracuda 2 TB HD
EVGA Supernova 650 G2
Meshify C case w/ Noctua chromax fans.
2 x 140 in the front
1 x 120 in the rear
and the 240 rad exhausting air on top.


SOLUTION UPDATE: I did adjust the power management setting to high performance and changed the Nvidia Control Panel to maximum performance mode. It did help performance. I'm now getting a ~14,200 in fire strike. Also, the temps are down to 83 C under load, like it was before. I wasn't thermal throttling before.

I downloaded MSI afterburner and that's helped with cooling down to about 74 C.

I'll mess around more but it definitely helped with performance but I think I should be getting better scores but I could be wrong.

 
Solution
I would check:
_ whether windows is running in High performance power plan: with PCI Express link state power management set to off as well in the advanced power setting
_ With GPU-z whether card works at is PCIe 16. V 3.0 under load (Run GPUz render test) and to determine whether your card is getting thermo-throttled (under sensor tab) and Hwmonitor to check for same values.
_ In Nvidia Control Panel that card is set to work in maximum performance mode (also try adaptive mode as some users reporting better paradoxically better results)
_ Improve card cooling: Use MSi after burner to insure card is run at 100% power and at improved more abrupt fan curve to maintain the card below 75 or better below 65 degrees, and add an extra...

R0GG

Distinguished
I would check:
_ whether windows is running in High performance power plan: with PCI Express link state power management set to off as well in the advanced power setting
_ With GPU-z whether card works at is PCIe 16. V 3.0 under load (Run GPUz render test) and to determine whether your card is getting thermo-throttled (under sensor tab) and Hwmonitor to check for same values.
_ In Nvidia Control Panel that card is set to work in maximum performance mode (also try adaptive mode as some users reporting better paradoxically better results)
_ Improve card cooling: Use MSi after burner to insure card is run at 100% power and at improved more abrupt fan curve to maintain the card below 75 or better below 65 degrees, and add an extra fan blowing on the card PCB or underneath card to help with cooling +++
 
Solution