Hii there,

I started BOINC up after a long time (yes, I updated the thing), but it only used my GTX960, leaving my GTX750Ti out of the fight?

How can I get my GTX750Ti to also do some lifting?
I already enabled this thing in my cc_config.xml, but it only enables it to be used for BOINC, but BOINC doesn't use it.

<use_all_gpus>1</use_all_gpus>

In the Tasks list it shows this:



Here is my log:

9/7/2016 1:17:15 PM |  | Starting BOINC client version 7.6.22 for windows_x86_64
9/7/2016 1:17:15 PM |  | log flags: file_xfer, sched_ops, task, coproc_debug
9/7/2016 1:17:15 PM |  | Libraries: libcurl/7.45.0 OpenSSL/1.0.2d zlib/1.2.8
9/7/2016 1:17:15 PM |  | Data directory: D:\BOINC-Data
9/7/2016 1:17:15 PM |  | Running under account Finlay
9/7/2016 1:17:15 PM |  | [coproc] launching child process at D:\BOINC\boinc.exe
9/7/2016 1:17:15 PM |  | [coproc] relative to directory D:\BOINC-Data
9/7/2016 1:17:15 PM |  | [coproc] with data directory "D:\BOINC-Data"
9/7/2016 1:17:16 PM |  | CUDA: NVIDIA GPU 0: GeForce GTX 960 (driver version 368.81, CUDA version 8.0, compute capability 5.2, 4096MB, 3067MB available, 2618 GFLOPS peak)
9/7/2016 1:17:16 PM |  | CUDA: NVIDIA GPU 1: GeForce GTX 750 Ti (driver version 368.81, CUDA version 8.0, compute capability 5.0, 2048MB, 1909MB available, 1539 GFLOPS peak)
9/7/2016 1:17:16 PM |  | OpenCL: NVIDIA GPU 0: GeForce GTX 960 (driver version 368.81, device version OpenCL 1.2 CUDA, 4096MB, 3067MB available, 2618 GFLOPS peak)
9/7/2016 1:17:16 PM |  | OpenCL: NVIDIA GPU 1: GeForce GTX 750 Ti (driver version 368.81, device version OpenCL 1.2 CUDA, 2048MB, 1909MB available, 1539 GFLOPS peak)
9/7/2016 1:17:16 PM |  | OpenCL: Intel GPU 0: Intel(R) HD Graphics 4600 (driver version 10.18.14.4264, device version OpenCL 1.2, 1426MB, 1426MB available, 64 GFLOPS peak)
9/7/2016 1:17:16 PM |  | OpenCL CPU: Intel(R) Core(TM) i5-4460  CPU @ 3.20GHz (OpenCL driver vendor: Intel(R) Corporation, driver version 4.2.0.148, device version OpenCL 1.2 (Build 148))
9/7/2016 1:17:16 PM |  | OpenCL CPU: Intel(R) Core(TM) i5-4460  CPU @ 3.20GHz (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1800.5 (sse2,avx), device version OpenCL 1.2 AMD-APP (1800.5))
9/7/2016 1:17:16 PM |  | [coproc] NVIDIA library reports 2 GPUs
9/7/2016 1:17:16 PM |  | [coproc] calInit() returned 1
9/7/2016 1:17:16 PM |  | Host name: Finlay-PC
9/7/2016 1:17:16 PM |  | Processor: 4 GenuineIntel Intel(R) Core(TM) i5-4460  CPU @ 3.20GHz [Family 6 Model 60 Stepping 3]
9/7/2016 1:17:16 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 smep bmi2
9/7/2016 1:17:16 PM |  | OS: Microsoft Windows 8.1: Professional x64 Edition, (06.03.9600.00)
9/7/2016 1:17:16 PM |  | Memory: 15.90 GB physical, 18.78 GB virtual
9/7/2016 1:17:16 PM |  | Disk: 465.76 GB total, 253.78 GB free
9/7/2016 1:17:16 PM |  | Local time is UTC +2 hours
9/7/2016 1:17:16 PM |  | VirtualBox version: 5.1.2
9/7/2016 1:17:16 PM |  | Config: don't use GPUs while Overwatch.exe is running
9/7/2016 1:17:16 PM |  | Config: use all coprocessors
9/7/2016 1:17:17 PM | Milkyway@Home | URL http://milkyway.cs.rpi.edu/milkyway/; Computer ID 702989; resource share 100
9/7/2016 1:17:17 PM |  | General prefs: from http://setiathome.berkeley.edu/ (last modified 02-Sep-2016 09:41:03)
9/7/2016 1:17:17 PM |  | Host location: none
9/7/2016 1:17:17 PM |  | General prefs: using your defaults
9/7/2016 1:17:17 PM |  | Reading preferences override file
9/7/2016 1:17:17 PM |  | Preferences:
9/7/2016 1:17:17 PM |  | max memory usage when active: 12212.31MB
9/7/2016 1:17:17 PM |  | max memory usage when idle: 14654.77MB
9/7/2016 1:17:17 PM |  | max disk usage: 10.00GB
9/7/2016 1:17:17 PM |  | (to change preferences, visit a project web site or select Preferences in the Manager)

I hope somebody can help me, as it's about 1500GFLOPS being "thrown away" :C

Cheers

...but BOINC doesn't use it.

Actually, when looking at the log, BOINC is using it. If it weren't you'd see a message on the least GPU Nvidia GPU # (not used) and that's not showing in your case.

What you don't show is whether there is more work in cache for the GPU(s), whether work is being asked for them. Not sure if this is still the case, but Milkyway allowed only for a small cache, and only after you've run the cache down can you ask for more work. It also uses double precision in the calculations, something the Nvidia GPUs aren't very strong in, meaning that a potential single precision 1305 GFLOPS is just 40.8 double precision GFLOPS on a 750 Ti (referencing value found on Nvidia GPU comparison, Geforce 700 series).

I'd wait for the next work request and if that doesn't fix things, to ask at the project first if this GPU is even supported. Or if they allow for BOINC to have different GPUs in one system.

---