Ubuntu Multi-GPU Problem

[Aldur]

I have a rig with two GPUs: a GTX 1080Ti and a GTX 1080. I've been folding with this system in Win10 for quite some time. Since Win10 is no longer supported, I decided to try Linux and chose Ubuntu 24.04. The system and F@H recognize both GPUs. However, if I enable folding on the 2nd GPU (the 1080), the folding core will SIGSEG after about 5 minutes.

This is a fresh install of Ubuntu 24.04 using the nVidia 580 proprietary drivers. I started off with F@H 8.4.9, but I installed the 8.5.4 beta to see if some bug had been fixed...

I've tried:

• Folding with both GPUs enabled - doesn't work

• Folding with only the 1080 Ti enabled - this works

• Folding with only the 1080 enabled - doesn't work

I can share any logs that will be useful or other system / setup info.

Thanks in advance!

[Nicolas_orleans]

Have you tried setting up two resource group, one for each card ?

[muziqaz]

Screenshot of your FAH system settings would help. Click on the i in the circle icon and post the screenshot.

[Aldur]

Thanks for the suggestion. I hadn't tried setting up separate resource groups, so I gave it a go. Unfortunately, I get the same result. The good news is that the problem only stops the 1080 from folding now. The other resource group keeps running fine.

Code: Select all

17:14:48:I1:WU143:*********************** Log Started 2025-12-06T17:14:48Z ***********************
17:14:48:I1:WU143:*************************** Core27 Folding@home Core ***************************
17:14:48:I1:WU143: Core: Core27
17:14:48:I1:WU143: Type: 0x27
17:14:48:I1:WU143: Version: 8.2.1
17:14:48:I1:WU143: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
17:14:48:I1:WU143: Copyright: 2022 foldingathome.org
17:14:48:I1:WU143: Homepage: https://foldingathome.org/
17:14:48:I1:WU143: Date: Mar 27 2025
17:14:48:I1:WU143: Time: 04:28:24
17:14:48:I1:WU143: Revision: 84019442387121f2bed4ed52e05c53b7b7b52a72
17:14:48:I1:WU143: Branch: HEAD
17:14:48:I1:WU143: Compiler: GNU 7.5.0
17:14:48:I1:WU143: Options: -faligned-new -std=c++14 -fsigned-char -ffunction-sections
17:14:48:I1:WU143: -fdata-sections -O3 -funroll-loops -fno-pie
17:14:48:I1:WU143: -DOPENMM_VERSION="\"8.2.0\""
17:14:48:I1:WU143: Platform: linux 6.8.0-1021-azure
17:14:48:I1:WU143: Bits: 64
17:14:48:I1:WU143: Mode: Release
17:14:48:I1:WU143:Maintainers: John Chodera <john.chodera@choderalab.org> and Peter Eastman
17:14:48:I1:WU143: <peastman@stanford.edu>
17:14:48:I1:WU143: Args: -dir 09pSNEVev6qzPZeznkA7XQWeSv0qBGEdUakEEax5As0 -suffix 01
17:14:48:I1:WU143: -version 8.5.4 -lifeline 26098 -gpu-uuid
17:14:48:I1:WU143: 8f507a2c-abd8-13c8-c622-47c8c599f81d -gpu-platform cuda -gpu-vendor
17:14:48:I1:WU143: nvidia -opencl-platform 0 -opencl-device 1 -cuda-platform 0
17:14:48:I1:WU143: -cuda-device 1 -gpu 1
17:14:48:I1:WU143:************************************ libFAH ************************************
17:14:48:I1:WU143: Date: Mar 27 2025
17:14:48:I1:WU143: Time: 04:23:18
17:14:48:I1:WU143: Revision: c7d2824a47eb025fa8cda8968c7a5e971585d90c
17:14:48:I1:WU143: Branch: HEAD
17:14:48:I1:WU143: Compiler: GNU 7.5.0
17:14:48:I1:WU143: Options: -faligned-new -std=c++11 -fsigned-char -ffunction-sections
17:14:48:I1:WU143: -fdata-sections -O3 -funroll-loops -fno-pie
17:14:48:I1:WU143: Platform: linux 6.8.0-1021-azure
17:14:48:I1:WU143: Bits: 64
17:14:48:I1:WU143: Mode: Release
17:14:48:I1:WU143:************************************ CBang *************************************
17:14:48:I1:WU143: Version: 1.7.2
17:14:48:I1:WU143: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
17:14:48:I1:WU143: Org: Cauldron Development LLC
17:14:48:I1:WU143: Copyright: Cauldron Development LLC, 2003-2024
17:14:48:I1:WU143: Homepage: https://cauldrondevelopment.com/
17:14:48:I1:WU143: License: LGPL-2.1-or-later
17:14:48:I1:WU143: Date: Mar 27 2025
17:14:48:I1:WU143: Time: 04:22:53
17:14:48:I1:WU143: Revision: f1cd4c791e8c40a35dcfeab3ab85d910949cc0cb
17:14:48:I1:WU143: Branch: HEAD
17:14:48:I1:WU143: Compiler: GNU 7.5.0
17:14:48:I1:WU143: Options: -faligned-new -std=c++11 -fsigned-char -ffunction-sections
17:14:48:I1:WU143: -fdata-sections -O3 -funroll-loops -fno-pie -fPIC
17:14:48:I1:WU143: Platform: linux 6.8.0-1021-azure
17:14:48:I1:WU143: Bits: 64
17:14:48:I1:WU143: Mode: Release
17:14:48:I1:WU143:************************************ System ************************************
17:14:48:I1:WU143: CPU: Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz
17:14:48:I1:WU143: CPU ID: GenuineIntel Family 6 Model 94 Stepping 3
17:14:48:I1:WU143: CPUs: 8
17:14:48:I1:WU143: Memory: 31.28GiB
17:14:48:I1:WU143:Free Memory: 25.64GiB
17:14:48:I1:WU143:Digital signatures verified
17:14:48:I1:WU143:Folding@home GPU Core27 Folding@home Core
17:14:48:I1:WU143:Version 8.2.1
17:14:48:I1:WU143: GPU info: Platform: OpenCL: NVIDIA CUDA
17:14:48:I1:WU143: GPU info: PlatformIndex: 0
17:14:48:I1:WU143: GPU info: Device: NVIDIA GeForce GTX 1080 Ti
17:14:48:I1:WU143: GPU info: DeviceIndex: 0
17:14:48:I1:WU143: GPU info: Vendor: 0x10de
17:14:48:I1:WU143: GPU info: PCI: 01:00:00
17:14:48:I1:WU143: GPU info: Compute: 3.0
17:14:48:I1:WU143: GPU info: Driver: 580.95
17:14:48:I1:WU143: GPU info: GPU: true
17:14:48:I1:WU143: GPU info: Platform: OpenCL: NVIDIA CUDA
17:14:48:I1:WU143: GPU info: PlatformIndex: 0
17:14:48:I1:WU143: GPU info: Device: NVIDIA GeForce GTX 1080
17:14:48:I1:WU143: GPU info: DeviceIndex: 1
17:14:48:I1:WU143: GPU info: Vendor: 0x10de
17:14:48:I1:WU143: GPU info: PCI: 02:00:00
17:14:48:I1:WU143: GPU info: Compute: 3.0
17:14:48:I1:WU143: GPU info: Driver: 580.95
17:14:48:I1:WU143: GPU info: GPU: true
17:14:48:I1:WU143: GPU info: Platform: CUDA
17:14:48:I1:WU143: GPU info: PlatformIndex: 0
17:14:48:I1:WU143: GPU info: Device: NVIDIA GeForce GTX 1080 Ti
17:14:48:I1:WU143: GPU info: DeviceIndex: 0
17:14:48:I1:WU143: GPU info: Vendor: 0x10de
17:14:48:I1:WU143: GPU info: PCI: 01:00:00
17:14:48:I1:WU143: GPU info: Compute: 6.1
17:14:48:I1:WU143: GPU info: Driver: 13.0
17:14:48:I1:WU143: GPU info: GPU: true
17:14:48:I1:WU143: GPU info: Platform: CUDA
17:14:48:I1:WU143: GPU info: PlatformIndex: 0
17:14:48:I1:WU143: GPU info: Device: NVIDIA GeForce GTX 1080
17:14:48:I1:WU143: GPU info: DeviceIndex: 1
17:14:48:I1:WU143: GPU info: Vendor: 0x10de
17:14:48:I1:WU143: GPU info: PCI: 02:00:00
17:14:48:I1:WU143: GPU info: Compute: 6.1
17:14:48:I1:WU143: GPU info: Driver: 13.0
17:14:48:I1:WU143: GPU info: GPU: true
17:14:48:I1:WU143: Checkpoint write interval: 100000 steps (2%) [50 total]
17:14:48:I1:WU143: JSON viewer frame write interval: 50000 steps (1%) [100 total]
17:14:48:I1:WU143: XTC frame write interval: 10000 steps (0.2%) [500 total]
17:14:48:I1:WU143: TRR frame write interval: disabled
17:14:48:I1:WU143: Global context and integrator variables write interval: disabled
17:14:48:I1:WU143:There are 4 platforms available.
17:14:48:I1:WU143:Platform 0: Reference
17:14:48:I1:WU143:Platform 1: CPU
17:14:48:I1:WU143:Platform 2: OpenCL
17:14:48:I1:WU143: opencl-device 1 specified
17:14:48:I1:WU143:Platform 3: CUDA
17:14:48:I1:WU143: cuda-device 1 specified
17:14:50:I1:WU143:Attempting to create CUDA context:
17:14:50:I1:WU143: Configuring platform CUDA
17:14:53:I1:WU143: Using CUDA on CUDA Platform and gpu 1
17:14:53:I1:WU143: GPU info: Platform: CUDA
17:14:53:I1:WU143: GPU info: PlatformIndex: 0
17:14:53:I1:WU143: GPU info: Device: NVIDIA GeForce GTX 1080
17:14:53:I1:WU143: GPU info: DeviceIndex: 1
17:14:53:I1:WU143: GPU info: Vendor: 0x10de
17:14:53:I1:WU143: GPU info: PCI: 02:00:00
17:14:53:I1:WU143: GPU info: Compute: 6.1
17:14:53:I1:WU143: GPU info: Driver: 13.0
17:14:53:I1:WU143: GPU info: GPU: true
17:14:53:I1:WU143:Completed 0 out of 5000000 steps (0%)
17:14:53:I1:WU143:Checkpoint completed at step 0
17:15:13:I1:WU143:An exception occurred at step 15596: Particle coordinate is NaN. For more information, see https://github.com/openmm/openmm/wiki/Frequently-Asked-Questions#nan
17:15:13:I1:WU143:ERROR:98: Attempting to restart from last good checkpoint by restarting core.
17:15:13:I1:WU143:Folding@home Core Shutdown: CORE_RESTART
17:15:13:I1:WU143:Core returned CORE_RESTART (98)
17:15:13:I1:WU143:Retry #1 in 2 secs

I've reviewed the FAQ for the Particle coordinate is NaN, but that isn't really any help. Also, I don't think the problem I'm seeing is WU dependent. It happens every time I start folding on the 1080 within about 5 minutes of starting.

[Aldur]

Here are my system settings:

[muziqaz]

NaN errors are not your fault.

[Aldur]

I dumped that WU and tried again. This log is more like what I normally see with a SIGINT.

Code: Select all

20:02:39:I1:WU158:Requesting WU assignment for user ALDUR team 40051
20:02:39:I1:WU158:Received WU assignment 2CQnDuu1ROkE0tI1WF_-F4PBunnX_lCcO1h3fc4l1t0
20:02:39:I1:WU158:Downloading WU
20:02:43:I1:WU158:DOWNLOAD 100% 220B of 220B
20:02:44:I1:WU158:DOWNLOAD 100% 17.64MiB of 17.64MiB
20:02:44:I1:WU158:Received WU P16773 R93 C8 G83
20:02:44:I3:WU158:Started FahCore on PID 44990
20:02:45:I1:WU158:*********************** Log Started 2025-12-06T20:02:45Z ***********************
20:02:45:I1:WU158:*************************** Core27 Folding@home Core ***************************
20:02:45:I1:WU158: Core: Core27
20:02:45:I1:WU158: Type: 0x27
20:02:45:I1:WU158: Version: 8.2.1
20:02:45:I1:WU158: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
20:02:45:I1:WU158: Copyright: 2022 foldingathome.org
20:02:45:I1:WU158: Homepage: https://foldingathome.org/
20:02:45:I1:WU158: Date: Mar 27 2025
20:02:45:I1:WU158: Time: 04:28:24
20:02:45:I1:WU158: Revision: 84019442387121f2bed4ed52e05c53b7b7b52a72
20:02:45:I1:WU158: Branch: HEAD
20:02:45:I1:WU158: Compiler: GNU 7.5.0
20:02:45:I1:WU158: Options: -faligned-new -std=c++14 -fsigned-char -ffunction-sections
20:02:45:I1:WU158: -fdata-sections -O3 -funroll-loops -fno-pie
20:02:45:I1:WU158: -DOPENMM_VERSION="\"8.2.0\""
20:02:45:I1:WU158: Platform: linux 6.8.0-1021-azure
20:02:45:I1:WU158: Bits: 64
20:02:45:I1:WU158: Mode: Release
20:02:45:I1:WU158:Maintainers: John Chodera <john.chodera@choderalab.org> and Peter Eastman
20:02:45:I1:WU158: <peastman@stanford.edu>
20:02:45:I1:WU158: Args: -dir 2CQnDuu1ROkE0tI1WF_-F4PBunnX_lCcO1h3fc4l1t0 -suffix 01
20:02:45:I1:WU158: -version 8.5.4 -lifeline 26098 -gpu-uuid
20:02:45:I1:WU158: 8f507a2c-abd8-13c8-c622-47c8c599f81d -gpu-platform cuda -gpu-vendor
20:02:45:I1:WU158: nvidia -opencl-platform 0 -opencl-device 1 -cuda-platform 0
20:02:45:I1:WU158: -cuda-device 1 -gpu 1
20:02:45:I1:WU158:************************************ libFAH ************************************
20:02:45:I1:WU158: Date: Mar 27 2025
20:02:45:I1:WU158: Time: 04:23:18
20:02:45:I1:WU158: Revision: c7d2824a47eb025fa8cda8968c7a5e971585d90c
20:02:45:I1:WU158: Branch: HEAD
20:02:45:I1:WU158: Compiler: GNU 7.5.0
20:02:45:I1:WU158: Options: -faligned-new -std=c++11 -fsigned-char -ffunction-sections
20:02:45:I1:WU158: -fdata-sections -O3 -funroll-loops -fno-pie
20:02:45:I1:WU158: Platform: linux 6.8.0-1021-azure
20:02:45:I1:WU158: Bits: 64
20:02:45:I1:WU158: Mode: Release
20:02:45:I1:WU158:************************************ CBang *************************************
20:02:45:I1:WU158: Version: 1.7.2
20:02:45:I1:WU158: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
20:02:45:I1:WU158: Org: Cauldron Development LLC
20:02:45:I1:WU158: Copyright: Cauldron Development LLC, 2003-2024
20:02:45:I1:WU158: Homepage: https://cauldrondevelopment.com/
20:02:45:I1:WU158: License: LGPL-2.1-or-later
20:02:45:I1:WU158: Date: Mar 27 2025
20:02:45:I1:WU158: Time: 04:22:53
20:02:45:I1:WU158: Revision: 84019442387121f2bed4ed52e05c53b7b7b52a72
20:02:45:I1:WU158: Branch: HEAD
20:02:45:I1:WU158: Compiler: GNU 7.5.0
20:02:45:I1:WU158: Options: -faligned-new -std=c++14 -fsigned-char -ffunction-sections
20:02:45:I1:WU158: -fdata-sections -O3 -funroll-loops -fno-pie
20:02:45:I1:WU158: -DOPENMM_VERSION="\"8.2.0\""
20:02:45:I1:WU158: Platform: linux 6.8.0-1021-azure
20:02:45:I1:WU158: Bits: 64
20:02:45:I1:WU158: Mode: Release
20:02:45:I1:WU158:Maintainers: John Chodera <john.chodera@choderalab.org> and Peter Eastman
20:02:45:I1:WU158: <peastman@stanford.edu>
20:02:45:I1:WU158: Args: -dir 2CQnDuu1ROkE0tI1WF_-F4PBunnX_lCcO1h3fc4l1t0 -suffix 01
20:02:45:I1:WU158: -version 8.5.4 -lifeline 26098 -gpu-uuid
20:02:45:I1:WU158: 8f507a2c-abd8-13c8-c622-47c8c599f81d -gpu-platform cuda -gpu-vendor
20:02:45:I1:WU158: nvidia -opencl-platform 0 -opencl-device 1 -cuda-platform 0
20:02:45:I1:WU158: -cuda-device 1 -gpu 1
20:02:45:I1:WU158:************************************ libFAH ************************************
20:02:45:I1:WU158: Date: Mar 27 2025
20:02:45:I1:WU158: Time: 04:23:18
20:02:45:I1:WU158: Revision: c7d2824a47eb025fa8cda8968c7a5e971585d90c
20:02:45:I1:WU158: Branch: HEAD
20:02:45:I1:WU158: Compiler: GNU 7.5.0
20:02:45:I1:WU158: Options: -faligned-new -std=c++11 -fsigned-char -ffunction-sections
20:02:45:I1:WU158: -fdata-sections -O3 -funroll-loops -fno-pie
20:02:45:I1:WU158: Platform: linux 6.8.0-1021-azure
20:02:45:I1:WU158: Bits: 64
20:02:45:I1:WU158: Mode: Release
20:02:45:I1:WU158:************************************ CBang *************************************
20:02:45:I1:WU158: Version: 1.7.2
20:02:45:I1:WU158: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
20:02:45:I1:WU158: Org: Cauldron Development LLC
20:02:45:I1:WU158: Copyright: Cauldron Development LLC, 2003-2024
20:02:45:I1:WU158: Homepage: https://cauldrondevelopment.com/
20:02:45:I1:WU158: License: LGPL-2.1-or-later
20:02:45:I1:WU158: Date: Mar 27 2025
20:02:45:I1:WU158: Time: 04:22:53
20:02:45:I1:WU158: Revision: f1cd4c791e8c40a35dcfeab3ab85d910949cc0cb
20:02:45:I1:WU158: Branch: HEAD
20:02:45:I1:WU158: Compiler: GNU 7.5.0
20:02:45:I1:WU158: Options: -faligned-new -std=c++11 -fsigned-char -ffunction-sections
20:02:45:I1:WU158: -fdata-sections -O3 -funroll-loops -fno-pie -fPIC
20:02:45:I1:WU158: Platform: linux 6.8.0-1021-azure
20:02:45:I1:WU158: Bits: 64
20:02:45:I1:WU158: Mode: Release
20:02:45:I1:WU158:************************************ System ************************************
20:02:45:I1:WU158: CPU: Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz
20:02:45:I1:WU158: CPU ID: GenuineIntel Family 6 Model 94 Stepping 3
20:02:45:I1:WU158: CPUs: 8
20:02:45:I1:WU158: Memory: 31.28GiB
20:02:45:I1:WU158:Free Memory: 24.35GiB
20:02:45:I1:WU158: OS Version: 6.14
20:02:45:I1:WU158:Has Battery: false
20:02:45:I1:WU158: On Battery: false
20:02:45:I1:WU158: Hostname: Mimi
20:02:45:I1:WU158: UTC Offset: -6
20:02:45:I1:WU158: PID: 44990
20:02:45:I1:WU158: CWD: /var/lib/fah-client/work
20:02:45:I1:WU158: Exec: /var/lib/fah-client/cores/openmm-core-27/centos-7.9.2009-64bit/release/fahcore-27-centos-7.9.2009-64bit-release-8.2.1/FahCore_27
20:02:45:I1:WU158:************************************ OpenMM ************************************
20:02:45:I1:WU158: Version: 8.2.0
20:02:45:I1:WU158:********************************************************************************
20:02:45:I1:WU158:Project: 16773 (Run 93, Clone 8, Gen 83)
20:02:45:I1:WU158:Reading tar file core.xml
20:02:45:I1:WU158:Reading tar file integrator.xml
20:02:45:I1:WU158:Reading tar file state.xml
20:02:45:I1:WU158:Reading tar file system.xml
20:02:45:I1:WU158:Digital signatures verified
20:02:45:I1:WU158:Folding@home GPU Core27 Folding@home Core
20:02:45:I1:WU158:Version 8.2.1
20:02:45:I1:WU158: GPU info: Platform: OpenCL: NVIDIA CUDA
20:02:45:I1:WU158: GPU info: PlatformIndex: 0
20:02:45:I1:WU158: GPU info: Device: NVIDIA GeForce GTX 1080 Ti
20:02:45:I1:WU158: GPU info: DeviceIndex: 0
20:02:45:I1:WU158: GPU info: Vendor: 0x10de
20:02:45:I1:WU158: GPU info: PCI: 01:00:00
20:02:45:I1:WU158: GPU info: Compute: 3.0
20:02:45:I1:WU158: GPU info: Driver: 580.95
20:02:45:I1:WU158: GPU info: GPU: true
20:02:45:I1:WU158: GPU info: Platform: OpenCL: NVIDIA CUDA
20:02:45:I1:WU158: GPU info: PlatformIndex: 0
20:02:45:I1:WU158: GPU info: Device: NVIDIA GeForce GTX 1080
20:02:45:I1:WU158: GPU info: DeviceIndex: 1
20:02:45:I1:WU158: GPU info: Vendor: 0x10de
20:02:45:I1:WU158: GPU info: PCI: 02:00:00
20:02:45:I1:WU158: GPU info: Compute: 3.0
20:02:45:I1:WU158: GPU info: Driver: 580.95
20:02:45:I1:WU158: GPU info: GPU: true
20:02:45:I1:WU158: GPU info: Platform: CUDA
20:02:45:I1:WU158: GPU info: PlatformIndex: 0
20:02:45:I1:WU158: GPU info: Device: NVIDIA GeForce GTX 1080 Ti
20:02:45:I1:WU158: GPU info: DeviceIndex: 0
20:02:45:I1:WU158: GPU info: Vendor: 0x10de
20:02:45:I1:WU158: GPU info: PCI: 01:00:00
20:02:45:I1:WU158: GPU info: Compute: 6.1
20:02:45:I1:WU158: GPU info: Driver: 13.0
20:02:45:I1:WU158: GPU info: GPU: true
20:02:45:I1:WU158: GPU info: Platform: CUDA
20:02:45:I1:WU158: GPU info: PlatformIndex: 0
20:02:45:I1:WU158: GPU info: Device: NVIDIA GeForce GTX 1080
20:02:45:I1:WU158: GPU info: DeviceIndex: 1
20:02:45:I1:WU158: GPU info: Vendor: 0x10de
20:02:45:I1:WU158: GPU info: PCI: 02:00:00
20:02:45:I1:WU158: GPU info: Compute: 6.1
20:02:45:I1:WU158: GPU info: Driver: 13.0
20:02:45:I1:WU158: GPU info: GPU: true
20:02:45:I1:WU158: Checkpoint write interval: 100000 steps (4%) [25 total]
20:02:45:I1:WU158: JSON viewer frame write interval: 25000 steps (1%) [100 total]
20:02:45:I1:WU158: XTC frame write interval: 25000 steps (1%) [100 total]
20:02:45:I1:WU158: TRR frame write interval: disabled
20:02:45:I1:WU158: Global context and integrator variables write interval: disabled
20:02:45:I1:WU158:There are 4 platforms available.
20:02:45:I1:WU158:Platform 0: Reference
20:02:45:I1:WU158:Platform 1: CPU
20:02:45:I1:WU158:Platform 2: OpenCL
20:02:45:I1:WU158: opencl-device 1 specified
20:02:45:I1:WU158:Platform 3: CUDA
20:02:45:I1:WU158: cuda-device 1 specified
20:02:50:I1:WU158:Attempting to create CUDA context:
20:02:50:I1:WU158: Configuring platform CUDA
20:02:52:I1:WU158: Using CUDA on CUDA Platform and gpu 1
20:02:52:I1:WU158: GPU info: Platform: CUDA
20:02:52:I1:WU158: GPU info: PlatformIndex: 0
20:02:52:I1:WU158: GPU info: Device: NVIDIA GeForce GTX 1080
20:02:52:I1:WU158: GPU info: DeviceIndex: 1
20:02:52:I1:WU158: GPU info: Vendor: 0x10de
20:02:52:I1:WU158: GPU info: PCI: 02:00:00
20:02:52:I1:WU158: GPU info: Compute: 6.1
20:02:52:I1:WU158: GPU info: Driver: 13.0
20:02:52:I1:WU158: GPU info: GPU: true
20:02:52:I1:WU158:Completed 0 out of 2500000 steps (0%)
20:02:52:I1:WU158:Checkpoint completed at step 0
20:03:09:I1:WU158:Caught signal SIGINT(2)
20:03:09:I1:WU158:Exiting, please wait. . .
20:03:09:I1:WU158:Folding@home Core Shutdown: INTERRUPTED
20:03:09:I1:WU158:Core returned INTERRUPTED (102)

I saw that the system info says the 1080 doesn't support HIP. I tried disabling that for the 1080, but it didn't make any difference.

Any suggestions?

[muziqaz]

HIP, if not supported, will not be used, ticked or un-ticked. For nvidia it is either CUDA or OpenCL. You can try unticking CUDA on both and see how it reacts. If it folds on both at the same time, tick CUDA on 1080ti, leave it unticked on 1080.
I believe this is OS or even BIOS level issue, as FAH sees everything correctly and is using everything correctly. Maybe something to do with PCIe slots. Try swapping the GPUs over in PCIe slots, and if this still doesn't work, connect your monitor output cable to the secondary card.
I'm just stalling here, until people who actually use nV show up

[Aldur]

Thanks for the suggestions. I appreciate it!

I tried disabling CUDA on both GPUs. Same result. The 1080 Ti happily folds. The 1080 pretends like it's going to start for about 5 minutes and then fails with the logs reflecting SIGINT (2) (I can guarantee nobody pressed CTRL-C...).

I then tried splitting the CPU cores up among the two resource groups. I get the same result. What's interesting is the CPUs in the resource group with the 1080 ALSO stop folding with the logs showing SIGINT (2). When I had all of the hardware in the same resource group, only the 1080 failed...

I suspect it is something OS related (or OS client specific) since this machine will fold properly when I boot into Windows.

[Nicolas_orleans]

Historically Ubuntu+Nvidia driver have been bad with multi-gpu, not working with GPUs of different generations. Last time I managed to get a multi gpu setup it was with 2xGTX770+1xGTX750Ti. But now the cards are 3 slots wide anyway so multi GPU is not that common (without risers)
Out of curiosity did it fail under Wayland or X11 ? I think I had to mess with xorg.conf to get the 3 cards working.

[muziqaz]

As a side note, GPU resource groups should not have any CPU cores assigned. GPUs need free CPU threads for full performance, and not an assigned CPU threads. So each GPU resource groups should have cpu slider set to 0. And if you are folding on a CPU in the separate resource group, it should be set to 2 less threads/cores than the total available. Those 2 free cores/threads would then be free to service 2 GPUs.

This does not solve the problem here, but just thought it is worth mentioning, since I see you tried assigning CPUs to the GPU groups