Dual RTX4000 ADA XID 62 Failure Ubuntu 22.04

Graphics / Linux Linux
,
1

I can’t get my dual GPU setup running.

dmesg shows, that the second GPU gets a XID 62 failure.

"[drm] Initialized nvidia-drm 0.0.0 20160202 for 0000:01:00.0 on minor 1
[ 4.778202] [drm] [nvidia-drm] [GPU ID 0x00000600] Loading driver
[ 6.485107] NVRM: GPU at PCI:0000:06:00: GPU-8a467b97-aaf0-3ca3-a95b-82f9733dd423
[ 6.485114] NVRM: Xid (PCI:0000:06:00): 62, pid=‘’, name=, 20258698 202605cc 2026107c 2025fd86 2025c050 202633f2 00000000 00000000
[ 7.016004] e1000e 0000:00:1f.6 eno2: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
[ 10.484013] NVRM: GPU 0000:06:00.0: RmInitAdapter failed! (0x62:0xffff:2468)
[ 10.484536] NVRM: GPU 0000:06:00.0: rm_init_adapter failed, device minor number 1
[ 10.484711] [drm:nv_drm_load [nvidia_drm]] ERROR [nvidia-drm] [GPU ID 0x00000600] Failed to allocate NvKmsKapiDevice
[ 10.484899] [drm:nv_drm_register_drm_device [nvidia_drm]] ERROR [nvidia-drm] [GPU ID 0x00000600] Failed to register device
[ 10.509313] nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
[ 10.537234] nvidia-uvm: Loaded the UVM driver, major device number 508.
"
my setup:
Lenovo P3 Tower
ubuntu 22.04 Kernel: Ubuntu 6.8.0-48.48~22.04.1-generic 6.8.12
cuda 12.6
nvidia driver version: 560.30.03

GPUs are not broken, each one works fine in a single GPU environment.

any ideas, how to get this setup running?

2

Updated to 565.57.01

no change. The second GPU fails with XID 62 during boot

Here is a bug report. Would appreciate, if somebody takes a look at it.

Thanks
nvidia-bug-report.log.gz (7.9 MB)