20:42:53 (15380): device: OpenCL: NVIDIA GPU 0: NVIDIA GeForce GTX 1080 (driver version 560.81, device version OpenCL 3.0 CUDA, 8192MB, 8192MB available, 9070 GFLOPS peak)

20:42:53 (15380): device: OpenCL: NVIDIA GPU 1 (not used): NVIDIA GeForce GTX 1050 Ti (driver version 560.81, device version OpenCL 3.0 CUDA, 4096MB, 4096MB available, 2274 GFLOPS peak)

A bit further down is: [Aug 28 18:42:55 UTC] Automatic processor type detection did not
recognize the processor (tag: "NVIDIA GeForce GTX 1080 ")
[Aug 28 18:43:12 UTC] RC5-72: using core #2 (CL 2-pipe).

Does not recognize, but yet seems that it is using it instead of the 1050 (not used)

I don't understand this conflicting information.

All I know is that the task completes just fine.

One other thing...why does it say CPU time 834.750000 when it has been using GPU's?

So how is it that 2 GPU's are tied up with their control CPU's but only 1 seems to be in use?

How do I break this? Just write a app_config (?) file to force it to one of the two GPU's?

I saw some old threads on this topic, but they are a bit dated.
What new current advice do you guys have?

SOLVED

Edited the app_config.xml to 1 CPU 1 GPU and now it runs two tasks simultaneously and separately.
Run times are a little longer, but not significant.