The host has one GTX 690 (Seti only) and one HD 5970 (Milkyway only) and see that Milkyway (or BOINC?) is not assigning the ATI tasks to the HD 5970, and I have no idea how to correct this.

My project settings has ATI-only, despite this, it is assigning OpenCL tasks to the GTX 690.

BOINC reports this:
21/09/2012 19:11:56 NVIDIA GPU 0: GeForce GTX 690 (driver version 30623, CUDA version 5000, compute capability 3.0, 2048MB, 522 GFLOPS peak)
21/09/2012 19:11:56 NVIDIA GPU 1: GeForce GTX 690 (driver version 30623, CUDA version 5000, compute capability 3.0, 2048MB, 522 GFLOPS peak)
21/09/2012 19:11:56 ATI GPU 0: ATI Radeon HD5800 series (Cypress) (CAL version 1.4.1741, 1024MB, 2320 GFLOPS peak)
21/09/2012 19:11:56 ATI GPU 1: ATI Radeon HD5800 series (Cypress) (CAL version 1.4.1741, 1024MB, 2320 GFLOPS peak)

Is the only way to override default logistics to use app_info also for Milkyway?


Looking at the stderr it appears that the Milkyway science app will select any available device within the platform - in this case bot ATI and Nvidia is in this platform, and for some reason Nvidia takes presedence. It's Guessing ATI is preferred:

"Guessing preferred OpenCL vendor 'Advanced Micro Devices, Inc.'", but it's omitted.....

Stderr:
<search_application> milkyway_separation 1.02 Windows x86_64 double OpenCL </search_application>
Unrecognized XML in project preferences: max_gfx_cpu_pct
Skipping: 100
Skipping: /max_gfx_cpu_pct
Unrecognized XML in project preferences: allow_non_preferred_apps
Skipping: 1
Skipping: /allow_non_preferred_apps
Unrecognized XML in project preferences: nbody_graphics_poll_period
Skipping: 30
Skipping: /nbody_graphics_poll_period
Unrecognized XML in project preferences: nbody_graphics_float_speed
Skipping: 5
Skipping: /nbody_graphics_float_speed
Unrecognized XML in project preferences: nbody_graphics_textured_point_size
Skipping: 250
Skipping: /nbody_graphics_textured_point_size
Unrecognized XML in project preferences: nbody_graphics_point_point_size
Skipping: 40
Skipping: /nbody_graphics_point_point_size
Guessing preferred OpenCL vendor 'Advanced Micro Devices, Inc.'
Error loading Lua script 'astronomy_parameters.txt': [string "number_parameters: 4..."]:1: '<name>' expected near '4'
Error reading astronomy parameters from file 'astronomy_parameters.txt'
Trying old parameters file
Using SSE4.1 path
Found 1 platform
Platform 0 information:
Name: NVIDIA CUDA
Version: OpenCL 1.1 CUDA 4.2.1
Vendor: NVIDIA Corporation
Extensions: cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing cl_nv_d3d9_sharing cl_nv_d3d10_sharing cl_khr_d3d10_sharing cl_nv_d3d11_sharing cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll
Profile: FULL_PROFILE
Didn't find preferred platform
Using device 0 on platform 0
Found 2 CL devices
Device 'GeForce GTX 690' (NVIDIA Corporation:0x10de) (CL_DEVICE_TYPE_GPU)
Driver version: 306.23
Version: OpenCL 1.1 CUDA
Compute capability: 3.0
Max compute units: 8
Clock frequency: 1019 Mhz
Global mem size: 2147483648
Local mem size: 49152
Max const buf size: 65536
Double extension: cl_khr_fp64
Build log:
--------------------------------------------------------------------------------

ptxas : info : Compiling entry function 'probabilities' for 'sm_30'
ptxas : info : Function properties for probabilities
88 bytes stack frame, 84 bytes spill stores, 84 bytes spill loads
ptxas : info : Used 62 registers, 388 bytes cmem[0], 176 bytes cmem[2]
--------------------------------------------------------------------------------
Build log:
--------------------------------------------------------------------------------


--------------------------------------------------------------------------------
Estimated Nvidia GPU GFLOP/s: 522 SP GFLOP/s, 65 DP FLOP/s
Using a target frequency of 60.0
Using a block size of 8192 with 5 blocks/chunk
Using clWaitForEvents() for polling with initial wait of 12 ms (mode 0)
Range: { nu_steps = 640, mu_steps = 1600, r_steps = 1400 }
Iteration area: 2240000
Chunk estimate: 53
Num chunks: 55
Chunk size: 40960
Added area: 12800
Effective area: 2252800
Initial wait: 12 ms

---

Morten Ross

Re-ran the OpenCl.msi and problem was resolved.

---

Morten Ross