Milkyway: OpenCL for Nvidia available for testing

[RSS]

The OpenCL application for Nvidia GPUs is ready for testing for Windows and Linux x86_64. I'm particularly interested in the performance / responsiveness tradeoff on mid-low range GPUs. Many thanks to cncguru for donating his GTX 480. If I hadn't had it, it would be about 30% slower than it is. http://milkyway.cs.rpi.edu/milkyway/download/test/milkyway_separation_0.48_x86_64-pc-linux-gnu__cuda_opencl.tar.gz http://milkyway.cs.rpi.edu/milkyway/download/test/milkyway_separation_0.48_windows_intelx86__cuda_op encl.zip Extract these to the project directory. On Windows this is something like C:\ProgramData\BOINC\projects\milkyway.cs.rpi.edu_ milkyway On Ubuntu for me, this is /var/lib/boinc-client/projects/milkyway.cs.rpi.edu_milkyway

More... (http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=2096)

[trigggl]

I assume a 9800 is still out of the running. :-w

[Fire$torm]

I assume a 9800 is still out of the running. :-w

Not sure. According to nVidia's OpenCL Programming Guide for the CUDA Architecture, (http://www.nvidia.com/content/cudazone/download/OpenCL/NVIDIA_OpenCL_ProgrammingGuide.pdf) Appendix A.3:

All compute devices supports the cl_khr_byte_addressable_store extension.

Devices of compute capability 1.1 and higher support the cl_khr_global_int32_base_atomics, cl_khr_global_int32_extended_atomics, cl_khr_local_int32_base_atomics, and cl_khr_local_int32_extended_atomics extensions.

Does this answer your question, I hope?

[trigggl]

Not sure. According to nVidia's OpenCL Programming Guide for the CUDA Architecture, (http://www.nvidia.com/content/cudazone/download/OpenCL/NVIDIA_OpenCL_ProgrammingGuide.pdf) Appendix A.3:

All compute devices supports the cl_khr_byte_addressable_store extension.

Devices of compute capability 1.1 and higher support the cl_khr_global_int32_base_atomics, cl_khr_global_int32_extended_atomics, cl_khr_local_int32_base_atomics, and cl_khr_local_int32_extended_atomics extensions.

Does this answer your question, I hope?
Last I heard, you need double precision. I don't think I've got that.

Thu 02 Dec 2010 11:45:19 PM CST Milkyway@home Message from server: Your NVIDIA GPU lacks the needed compute capability (1.3, required for double precision math

Just to be sure, since I didn't realize I needed to download the new app, I installed the downloaded app, and of course I got this.

Device doesn't support double precision
Device failed capability check
Failed to calculate integral 0
23:58:37 (24536): called boinc_finish

[Maxwell]

You need double precision for the native apps. However, what they are referring to in that post aren't native - you need to run the app_info file to get them to work, and download the files linked to in the original post. So, I'll echo Fire$torm's resounding "maybe"...

[trigggl]

You need double precision for the native apps. However, what they are referring to in that post aren't native - you need to run the app_info file to get them to work, and download the files linked to in the original post. So, I'll echo Fire$torm's resounding "maybe"...

See last thing I posted above. It still needs double precision

[Fire$torm]

Last I heard, you need double precision. I don't think I've got that.

Thu 02 Dec 2010 11:45:19 PM CST Milkyway@home Message from server: Your NVIDIA GPU lacks the needed compute capability (1.3, required for double precision math

Just to be sure, since I didn't realize I needed to download the new app, I installed the downloaded app, and of course I got this.

DUH. My error. You are correct. MilkyWay is always Double Precision. I was tired when I posted and IIRC I was thinking (Really, is that what you call it?) nVidia question = non-MW project problem.