GPU Error always the same mother PCIE (changing card too)

[LuPuS2k]

Hello…i got a problem RANDOMLY, sometimes at every hour, some times after 24 hours, sometimes after 4…it’s really random but ALWAYS the same SLOT of the mother…
I change the CARD, the RISER, the cable, the PSU, everything and the problem persist (and the miner restart and restart every time.
The error y see in the syslog is this

Apr 15 19:09:43 RIG_1 kernel: [81838.839514][ T1119] NVRM: Xid (PCI:0000:0f:00): 69, pid=31144, Class Error: ChId 0014, Class 0000c7c0, Offset 00001b0c, Data ffffffff, ErrorCode 0000000c
Apr 15 19:09:47 RIG_1 kernel: [81843.393835][ T2615] DTS: killing sk:0000000024b5de99 (127.0.0.1:55168 -> 127.0.0.1:4058) state 6
Apr 15 19:09:47 RIG_1 kernel: [81843.393838][ T2615] DTS: killing sk:00000000e9073b99 (127.0.0.1:55140 -> 127.0.0.1:4058) state 6
Apr 15 19:09:47 RIG_1 kernel: [81843.393840][ T2615] DTS: killing sk:00000000bcb19766 (127.0.0.1:55160 -> 127.0.0.1:4058) state 6
Apr 15 19:09:47 RIG_1 kernel: [81843.393841][ T2615] DTS: killing sk:000000008d9e0bc1 (127.0.0.1:55148 -> 127.0.0.1:4058) state 6
Apr 15 19:09:47 RIG_1 kernel: [81843.393842][ T2615] DTS: killing sk:00000000b21c9371 (127.0.0.1:55134 -> 127.0.0.1:4058) state 6
Apr 15 19:09:47 RIG_1 kernel: [81843.393844][ T2615] DTS: killing sk:0000000060e58ee7 (127.0.0.1:55152 -> 127.0.0.1:4058) state 6

And in the miner i get this error:
20210415 15:18:24 TREX: Can't find nonce with device [ID=7, GPU #7], cuda exception in [synchronize, 51], unspecified launch failure, try to reduce overclock to stabilize GPU state

Anyone have the same error or can help me to fix this?

Thanks!

[rivendil]

Hi, I’m getting the same error. Running a rig of 6x 1660s all overclocked with mem -1004. Were you able to solve it?

[joewilert]

I’m having exactly the same problem, only difference is that in my rig the #3 slot is the lucky one. Have you ever find a way around it? (I’m running 6x GTX 1660Ss)

[hdreis]

Hi, I’m getting same error. My rig has 3 1660s and sometimes the nbminer crash and throw this error. Someone fix this problem?

[ksuhiyp]

I am mining RVN on 3070 ti PNY, OC is set to 100 core 2000 mem 251W TDP. getting this error periodically on a random manner.