Hello,
Since some days I get a problem with one machine, the screen went black after some time (2 … 10 minutes).
The only typical message in the logs is :
NVRM: GPU 0000:07:00.0: GPU has fallen off the bus.
I wonder about another message :
caller _nv000720rm+0x1ad/0x200 [nvidia] mapping multiple BARs
I did not see how to join the log file made through “nvidia-bug-report.sh”.
Since the 710 is quite old, maybe the fan is broken/clogged so it’s overheating and shutting down. Please use nvidia-settings or nvidia-smi to monitor temperatures/fan speed.
Otherwise, there might be a general board incompatibilty (unlikely since according to your description it worked before) or the 710 simply broken.
is a sign the driver resetted the gpu. Does not necessarily mean anything but if that’s happening more than once/twice right after reboot, it’s a sign for the gpu to be broken.
Hello,
When looking at temp I saw a problem with the fan which was limited to 40%, I put it to the maximum (more than 76 is not possible) but all went nice and the screen is on sine 3 hours.
But I don’t understand why the fan has “fallen” to 40%.
Ill was OK with the card until last week.
As that card is old I think to upgrade, anyone knows if I can put a Gt730 to replace the old GT710 ?
Usually, the fans only ramp up when temperature is beyond 60 or 70°C so if the 710 gets unstable while at 53°C, I guess it’s broken. Either the sensor or the gpu.
Hello,
My problem is that the “big” pcie slot is used by a card, only “small” pcie slot free and the 710/730 are using such ports.
Is there any “modern” card using that kind of slot ?
Is that an x8 (half size) or x4 (quarter size) slot? The only half-modern card where some models have an x8 connector is the GT1030 but that’s very limited (no video encoders, 2 monitors only).