I added a second GPU card (GTX 1050 TI) a couple of weeks ago and it has been running fine along with the original card (BTX 750 TI)just fine until this morning -- WUs only run in one card and the other one is idle - and not always the same one. Restarting BOINC or rebooting might cause a different card to run, but never both at once.

Here is a copy of my cc_config file:
<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<exclude_gpu>
<url>asteroidsathome.net/boinc</url>
<device_num>1</device_num>
<type>NVIDIA</type>
</exclude_gpu>
<exclude_gpu>
<url>https://www.gpugrid.net</url>
<device_num>1</device_num>
<type>NVIDIA</type>
</exclude_gpu>
</options>
</cc_config>

Here is a copy of the first part of my log file at startup showing that both cards are recognized and that the cc_config file is read:

5/29/2020 1:00:19 PM | | Starting BOINC client version 7.16.5 for windows_x86_64
5/29/2020 1:00:19 PM | | Libraries: libcurl/7.47.1 OpenSSL/1.0.2s zlib/1.2.8
5/29/2020 1:00:19 PM | | Data directory: C:\ProgramData\BOINC
5/29/2020 1:00:19 PM | | Running under account Larry
5/29/2020 1:00:20 PM | | CUDA: NVIDIA GPU 0: GeForce GTX 1050 Ti (driver version 445.87, CUDA version 11.0, compute capability 6.1, 4096MB, 3376MB available, 2235 GFLOPS peak)
5/29/2020 1:00:20 PM | | CUDA: NVIDIA GPU 1: GeForce GTX 750 Ti (driver version 445.87, CUDA version 11.0, compute capability 5.0, 2048MB, 1680MB available, 1388 GFLOPS peak)
5/29/2020 1:00:20 PM | | OpenCL: NVIDIA GPU 0: GeForce GTX 1050 Ti (driver version 445.87, device version OpenCL 1.2 CUDA, 4096MB, 3376MB available, 2235 GFLOPS peak)
5/29/2020 1:00:20 PM | | OpenCL: NVIDIA GPU 1: GeForce GTX 750 Ti (driver version 445.87, device version OpenCL 1.2 CUDA, 2048MB, 1680MB available, 1388 GFLOPS peak)
5/29/2020 1:00:20 PM | Milkyway@Home | Found app_info.xml; using anonymous platform
5/29/2020 1:00:20 PM | | Windows processor group 0: 4 processors
5/29/2020 1:00:20 PM | | Host name: LEPC2
5/29/2020 1:00:20 PM | | Processor: 4 GenuineIntel Intel(R) Core(TM) i3-7350K CPU @ 4.20GHz [Family 6 Model 158 Stepping 9]
5/29/2020 1:00:20 PM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 vmx tm2 pbe fsgsbase bmi1 hle smep bmi2
5/29/2020 1:00:20 PM | | OS: Microsoft Windows 10: Professional x64 Edition, (10.00.18363.00)
5/29/2020 1:00:20 PM | | Memory: 15.95 GB physical, 18.32 GB virtual
5/29/2020 1:00:20 PM | | Disk: 222.51 GB total, 148.62 GB free
5/29/2020 1:00:20 PM | | Local time is UTC -7 hours
5/29/2020 1:00:20 PM | | No WSL found.
5/29/2020 1:00:20 PM | Asteroids@home | Config: excluded GPU. Type: NVIDIA. App: all. Device: 1
5/29/2020 1:00:20 PM | GPUGRID | Config: excluded GPU. Type: NVIDIA. App: all. Device: 1
5/29/2020 1:00:20 PM | | Config: use all coprocessors
5/29/2020 1:00:20 PM | Einstein@Home | General prefs: from Einstein@Home (last modified ---)
5/29/2020 1:00:20 PM | Einstein@Home | Computer location: home
5/29/2020 1:00:20 PM | | General prefs: using separate prefs for home
5/29/2020 1:00:20 PM | | Reading preferences override file
5/29/2020 1:00:20 PM | | Preferences:

Any ideas?!?

____________

From what I can see here, you have excluded GPU number 1(a number which seems not to be a permanent pointer to a certain GPU in your machine) from both Asteroids@home and GPUGRID.net.

I would suggest removing these lines from your config:

From what I can see here, you have excluded GPU number 1(a number which seems not to be a permanent pointer to a certain GPU in your machine) from both Asteroids@home and GPUGRID.net.

I would suggest removing these lines from your config:

<exclude_gpu>
<url>https://www.gpugrid.net</url>
<device_num>1</device_num>
<type>NVIDIA</type>
</exclude_gpu>

I would suggest removing these lines from your config:

<exclude_gpu>
<url>https://www.gpugrid.net</url>
<device_num>1</device_num>
<type>NVIDIA</type>
</exclude_gpu>

Those lines affect only GPUGRID.net, not Milkyway.
Take a look at "<exclude_gpu>" in https://boinc.berkeley.edu/wiki/Client_configuration

Good thing you've got it working now!

Those lines affect only GPUGRID.net, not Milkyway.
Take a look at "<exclude_gpu>" in https://boinc.berkeley.edu/wiki/Client_configuration

While the exclude gpu statement was perfectly fine, you forgot you have other BOINC mechanisms working. The one that stopped the previous behaviour is the REC mechanism that tries to balance the estimated credit among all your projects.

You simply got out of balance on one project versus the others and the project that needed to run was one that had an exclusion to run on the needed hardware.

So it stopped running the other tasks to free up the needed authorized gpu and the other gpu went idle.

It would have sorted itself out eventually once the REC debt had been paid off on the fallen behind project if you had done nothing. BOINC does automatically manage itself very well . . . . . if you let it and have patience. But people are not patient and don't understand how BOINC works under the hood mostly. Best is to just let BOINC run without interfering in its mechanism even though all the controls to manipulate it are very tantalizing.

Thanks.