GPU's Running Very Hot in-game and Idle

[emze2822]

I have two GTX 970's in my system (Specs here: https://pcpartpicker.com/list/DgWHvK) and lately i've noticed them running unplayably hot in-game, and pretty toasty on idle. I've been monitoring my temps with MSI Afterburner and saw GPU 1 temps shoot up to 91C on 100% fan speed while GPU 2 was 60C also at 100% fan speed. (Cards are not overclocked. 0 Core voltage, 100 power limit, 0 core clock, 0 memory clock) Both cards display the same usage but both dont reflect nearly the same temps. GPU 1 temp at idle is 51C on 100% fan speed and GPU 2 is 40C also at 100% (GPU 1: 60C and GPU 2: 42C at 40% fan speed) Some games like PUBG are playable but have concerningly borderline temps, I have to limit GPU temps to 85 otherwise they will exceed 90, which bottlenecks performance quite a bit, while games like Rust glitch out and crash within minutes. Keep in mind that I've lowered the all the graphics settings down to potato quality and i'm still having huge temp problems. I just got Rust and PUBG so these are the only results I know of for these cards.
I've taken both cards out, cleaned them even though they weren't dusty at all, renewed thermal paste, and switched the positions of the cards. It seems that whichever GPU 1 is, that's the card that's going to vastly overheat, possibly because i'm running 3 monitors off of it, but even then i'm not sure because i'm only gaming at 1920x1080.
These cards are 2 and 3 years old (i lost track of which one is eldest, but they behave identically in whichever spot they're in, for example, one card shows the same results as GPU 1 as the other would as GPU 1) and haven't been running hardcore games, and if so, not for long. I'm not a huge gamer so i cant sit at my pc all day and play so my cards shouldn't be worn down at all. I've checked that both cards are being detected and are SLI'ed correctly.
Could it be the viruses solely affecting temps? Can it just be old cards not able to keep up anymore? Any suggestions would help, and if any other information is needed just ask. Thank you

 

[dudio]

You can try running MalwareBytes (free) and confirm that you don't have malware using the GPU (bitcoin miner, etc)

 

[dvd03cr]

is the ambient temperature has changed where the pc is located? when u change the thermal paste u do drill screws well?

 

[jr9]

- Bitcoin malware generally doesn't run unless the PC is idle. It wouldn't cause overheating any more than the card running at 100% load would either. Malware scans are good to do though.

- Temperature differences with cards in SLI is normal, the master card generally runs hotter.

- Going past 90c can be observed in nVIDIA cards. 95c is what I consider "hot" and only at 100c do I consider a graphics card badly overheating. Revert your GPU temp limit settings as it sounds like you are gimping your graphics cards. In stress tests and gaming I've seen multiple nVIDIA cards peak at 93-94c in an open air test bench. They run hot.

- Blow the cards out with compressed air along with your case. Ensure all fans are working and there is good airflow with no obstructions.

- Do not remove the heatsink from a graphics card unless you are very sure what you are doing. This is a last resort repair I would only do if the card was out of warranty and experiences serious overheating over 100c or crashes.