i purchased a used evga ultra ftw3 3080 (no overcloked) from my close friend . i run modern games on 4k ultra setting fps is above 60 but temp is near 85c ! when i run 2k ultra temp is 81c. i tested unigine heaven in extreme setting every thing is fine temp is 82 max ,fps 82.6 score 2082 !
Last night, I tested furmark, after a minute the software crashes when the temp reaches 74, is there really a problem with the card?
my psu is coolermaster 750w psu 2 direct PCIe power plus 1 daisy chained is plugged into the GPU
in msi afterburn max tpd is 380w
What other tests can I do so that I can feel at ease because I have no problems in the game except the temp?
my full specs:
i5 12400f
ram 2*8 gb ddr4
6x 120m fan case
coolermaster td500 crystal
You must log in or register to comment.
either a driver issue or a problem with the card, both are very possible. try this https://www.guru3d.com/download/display-driver-uninstaller-download/
wrong subI’m wondering if between the Intel CPU and 3080 it’s too much power draw. Have you considered undervoltimg your GPU. You might sacrifice a few fps but it should help with power draw and temos?
my psu is coolermaster 750w psu 2 direct PCIe power plus 1 daisy chained is plugged into the GPU
I’d guess that’s your problem.
3080 is known for large power spikes afterall, so your psu just might not be handling it.
1 daisy chained is plugged into the GPU
I had similar issues with games crashing after swapping from a 2080 to a 4070 actually, even though there’s barely a difference in TDP (15 Watts on paper, though the 2080 was actually overclocked while the 4070 is not). Using just one cable to connect to the 12 pin power adapter would cause instabilities and crashes in games like Cyberpunk. The solution was to use two separate cables.
OP already uses two separate PCIE cables. The GPU has however accepts three plugs so one cable is daisy chained.
85C, unless your airflow is shit, means that it’s time for new thermal paste on the core.
Use HWInfo64 to monitor the card while you’re running furmark. See what the hotspot temperature gets up to. I bet it’s over 100C. 105C is where it will VRM throttle and ramp the fans to emergency mode, which runs them beyond the 100% that the user can set to try to keep the card from dying.
If the hotspot is more than 15C above the core, that’s another indication that the core needs a repaste, since there are spots with no paste coverage due to pump out and dry out over the time it was used by the original owner.
Try with 3 PCIE cables to see if that solves the problem.
If you must go with only two then switch the plugs around so the daisy chain ones plus into different slots. I don’t think power draw is spread equally on all three. Assuming your PSU is high quality and should be able to handle the load.