Hi all, I recently added an EVGA RTX 3080 FTW3 in one of my rigs, it’s running fine but after 5-6 hours the fan speed indication changes to “err”, nevertheless the fans continue to work at the same speed.
Anyone had similar experience or a solution?
1 Like
Try 120 seconds timeout for oc settings.
If that does not help reduce mem a bit and try.
@tom-i thanks for your answer, I tried with lower mem 2500-2600 because when the error occurs, if I restart the miner I get “NVIDIA OC failed”
=== GPU 5, 0b:00.0 NVIDIA GeForce RTX 3080 10018 MB, PL: 100 W, 380 W, 400 W === 14:10:57
SET POWER LIMIT: 220.0 W [Unknown Error]
(exitcode=123)
SET CLOCKS: 1150 MHz
Max Perf mode: 4 (auto)
Attribute ‘GPUGraphicsClockOffset’ was already set to 0
ERROR: Error assigning value 85 to attribute ‘GPUTargetFanSpeed’
(Rig2:0[fan:10]) as specified in assignment
‘[fan:10]/GPUTargetFanSpeed=85’ (Unknown Error).
ERROR: Error assigning value 85 to attribute ‘GPUTargetFanSpeed’
(Rig2:0[fan:11]) as specified in assignment
‘[fan:11]/GPUTargetFanSpeed=85’ (Unknown Error).
ERROR: Error assigning value 0 to attribute ‘GPUMemoryTransferRateOffset’
(Rig2:0[gpu:5]) as specified in assignment
‘[gpu:5]/GPUMemoryTransferRateOffset[4]=0’ (Unknown Error).
Attribute ‘GPUFanControlState’ (Rig2:0[gpu:5]) assigned value 1.
(exitcode=100)
I currently have 35s delay which is enough to build the DAG, 120 seems like an overkill although I’ll definitely try it.
You need to restart your rig when it happens to be able to set new settings for this card…
Yes obviously I do, I just mentioned the error i case it would give any clues.
this error ist the conesquence of failed oc settings. only reboot will give you a new try.
1 Like
I got the same problem but with an FE 3090. However it only happened after a long while of running. The fan is still running and it’s still hashing. I imagine the only way to fix it is to reboot the whole rig - which I have not done yet.
Actually, now that looking at my status, it seems this card stopped mining for a short time and my hiveOS watchdog automatically restarted t-rex to get it running again. This happened twice and has been running fine ever since (about 2 wks ago). I think maybe the fan error must have glitched during this time for whatever unknown reason…
A simply fix is power off your rig, switch PCI slot of risers, then restart.
I get the same issue with 3 different style cards(3080-ti, 3090-gigabyte, 3090-FE). I’ve tried rebooting the rig, lowering MEM and adjusting power settings. I haven’t tried turning off and changing PCIe slots or adding seconds to the OC time adjustment. Does anyone know what truly causes this or is it a guessing game for us all still? lol Thanks!
I didn’t realize it affected other 3090s. It appears for my 3090 FE after hashing continuously for a long while (X number of days?)
I suspect maybe it’s a linux driver issue but not really sure.
But in any case it doesn’t affect hashing operations. I just ignore it now if it shows Err.
I’m running Linux so it could be. I’ve got the 3090 Gigabyte that throws that Error and a 3080ti and 3090(both FE) that both throw the error code too. I’ve heard it’s a memory issue and that causes the yellow “failed OC setting” error code. To my knowledge it’s only on the Nvidia Cards.
I have another EVGA 3090 card but I run this on a Windows platform – not HiveOS (another possibly Linux driver issue that won’t let me change PL to 105% so I can’t get the overclock I need), and have not really noticed anything related to fan errors on that
This topic was automatically closed 416 days after the last reply. New replies are no longer allowed.