Message boards :
Number crunching :
Try using "ATI Stream Client core" selection to select OpenCL core
Message board moderation
Author | Message |
---|---|
![]() Send message Joined: 11 Feb 14 Posts: 114 Credit: 7,646,283 RAC: 87 |
Yes, this seems to work, at least on the iGPU of my Ryzen CPU. dnetc_r72_1714231152_13_819_0, run time 2 hours 15 min 54 sec with automatic core selection: <core_client_version>7.24.1</core_client_version> <![CDATA[ <stderr_txt> 17:58:00 (5640): wrapper v1.5 build 20 for AMD/ATI OpenCL starting (BOINC Wrapper v7.13.26016) 17:58:00 (5640): device: OpenCL: AMD/ATI GPU 0: AMD Radeon(TM) Graphics (driver version 3444.0 (PAL,HSAIL), device version OpenCL 2.0 AMD-APP (3444.0), 12367MB, 12367MB available, 2048 GFLOPS peak) 17:58:00 (5640): checkpoint interval: 0h15m00s00 (task 1146600 GFLOPS, 0h09m19s86 per packet) 17:58:00 (5640): wrapper: running dnetc521-win32-x86-opencl.exe (-ini dnetc.ini -runoffline -multiok=1) - attempt 1/10 dnetc v2.9112-521-CTR-16021318 for OpenCL on Win32 (WindowsNT 6.2). Using email address (distributed.net ID) '<censored> ' [Apr 27 15:58:01 UTC] Automatic processor type detection did not recognize the processor (tag: "gfx90c (AMD Radeon(TM) Graphics)") [Apr 27 15:58:20 UTC] RC5-72: using core #2 (CL 2-pipe). [Apr 27 15:58:20 UTC] RC5-72: Switched to random subspace 1343 (stored) [Apr 27 15:58:20 UTC] RC5-72: Loaded 5D:EAE55B49:00000000:64*2^32 [Apr 27 15:58:20 UTC] RC5-72: 12 packets (755.00 stats units) remain in in.r72 Projected ideal time to completion: 0.02:31:00.00 [Apr 27 15:58:20 UTC] RC5-72: 0 packets are in out.r72 [Apr 27 15:58:20 UTC] 1 cruncher has been started. [Apr 27 16:07:13 UTC] RC5-72: Completed 5D:EAE55B49:00000000 (64.00 stats units) 0.00:08:52.60 - [516,096,999 keys/s] (...) After running the benchmark integrated in dnetc521-win32-x86-opencl.exe: dnetc_r72_1715073438_13_823_0 run time 38 min 59 sec using core #1 set via the settings for ATI Steam: <core_client_version>7.24.1</core_client_version> <![CDATA[ <stderr_txt> 11:52:34 (10888): wrapper v1.5 build 20 for AMD/ATI OpenCL starting (BOINC Wrapper v7.13.26016) 11:52:34 (10888): device: OpenCL: AMD/ATI GPU 0: AMD Radeon(TM) Graphics (driver version 3444.0 (PAL,HSAIL), device version OpenCL 2.0 AMD-APP (3444.0), 12367MB, 12367MB available, 2048 GFLOPS peak) 11:52:34 (10888): checkpoint interval: 0h15m00s00 (task 1152200 GFLOPS, 0h09m22s59 per packet) 11:52:34 (10888): wrapper: running dnetc521-win32-x86-opencl.exe (-ini dnetc.ini -runoffline -multiok=1) - attempt 1/10 dnetc v2.9112-521-CTR-16021318 for OpenCL on Win32 (WindowsNT 6.2). Using email address (distributed.net ID) '<censored> ' [May 07 09:52:35 UTC] RC5-72: using core #1 (CL 1-pipe). [May 07 09:52:35 UTC] RC5-72: Switched to random subspace 1343 (stored) [May 07 09:52:35 UTC] RC5-72: Loaded 5E:0D31F809:00000000:64*2^32 [May 07 09:52:35 UTC] RC5-72: 12 packets (759.00 stats units) remain in in.r72 [May 07 09:52:35 UTC] RC5-72: 0 packets are in out.r72 [May 07 09:52:35 UTC] 1 cruncher has been started. [May 07 09:55:35 UTC] RC5-72: Completed 5E:0D31F809:00000000 (64.00 stats units) 0.00:02:59.34 - [1,532,685,269 keys/s] (...)As you see, no automatic detection and using the selected core instantly. To run the benchmark, suspend GPU processing in BOINC (not CPU if you run CPU tasks), open command line, navigate to your moowrap.net directory and execute dnetc521-win32-x86-opencl.exe -bench Try and report your results here, that's probably as close to OpenCL core selection as we can get here. ![]() |
Send message Joined: 19 Sep 14 Posts: 3 Credit: 89,256 RAC: 5 |
For some reason it hangs when i specify core via "ATI Stream Client core selection" at dnetc v2.9112-521-CTR-16021317 for OpenCL on Win32 (WindowsNT 6.2). Using email address (distributed.net ID) 'redacted' Here is my manual benchmark for NVIDIA GeForce GTX 1650 dnetc v2.9112-521-CTR-16021317 for OpenCL on Win32 (WindowsNT 6.2). [Jun 27 20:13:47 UTC] RC5-72: using core #0 (CL ANSI 1-pipe). [Jun 27 20:13:50 UTC] RC5-72: Benchmark for core #0 (CL ANSI 1-pipe) 0.00:00:01.39 [3,193,795,124 keys/sec] [Jun 27 20:13:50 UTC] RC5-72: using core #1 (CL 1-pipe). [Jun 27 20:13:53 UTC] RC5-72: Benchmark for core #1 (CL 1-pipe) 0.00:00:01.12 [4,117,914,608 keys/sec] [Jun 27 20:13:53 UTC] RC5-72: using core #2 (CL 2-pipe). [Jun 27 20:13:55 UTC] RC5-72: Benchmark for core #2 (CL 2-pipe) 0.00:00:01.14 [3,943,453,137 keys/sec] [Jun 27 20:13:55 UTC] RC5-72: using core #3 (CL 4-pipe). [Jun 27 20:13:57 UTC] RC5-72: Benchmark for core #3 (CL 4-pipe) 0.00:00:01.14 [3,952,754,344 keys/sec] [Jun 27 20:13:57 UTC] RC5-72 benchmark summary : Default core : #-1 (undefined) 0 keys/sec Fastest core : #1 (CL 1-pipe) 4,117,914,608 keys/sec [Jun 27 20:13:57 UTC] Compare and share your rates in the speeds database at http://www.distributed.net/speed/ (benchmark rates are for a single processor core) For CUDA dnetc v2.9110-519-CTR-11041423 for CUDA 3.1 on Win32 (WindowsNT 6.2). [Jun 27 20:04:38 UTC] RC5-72: using core #0 (CUDA 1-pipe 64-thd). [Jun 27 20:04:42 UTC] RC5-72: Benchmark for core #0 (CUDA 1-pipe 64-thd) 0.00:00:01.98 [2,221,661,107 keys/sec] [Jun 27 20:05:09 UTC] RC5-72: using core #1 (CUDA 1-pipe 128-thd). [Jun 27 20:05:13 UTC] RC5-72: Benchmark for core #1 (CUDA 1-pipe 128-thd) 0.00:00:01.95 [2,227,471,603 keys/sec] [Jun 27 20:05:38 UTC] RC5-72: using core #2 (CUDA 1-pipe 256-thd). [Jun 27 20:05:42 UTC] RC5-72: Benchmark for core #2 (CUDA 1-pipe 256-thd) 0.00:00:01.90 [2,324,563,844 keys/sec] [Jun 27 20:06:01 UTC] RC5-72: using core #3 (CUDA 2-pipe 64-thd). [Jun 27 20:06:05 UTC] RC5-72: Benchmark for core #3 (CUDA 2-pipe 64-thd) 0.00:00:01.98 [2,203,381,523 keys/sec] [Jun 27 20:06:26 UTC] RC5-72: using core #4 (CUDA 2-pipe 128-thd). [Jun 27 20:06:30 UTC] RC5-72: Benchmark for core #4 (CUDA 2-pipe 128-thd) 0.00:00:01.92 [2,329,517,830 keys/sec] [Jun 27 20:06:54 UTC] RC5-72: using core #5 (CUDA 2-pipe 256-thd). [Jun 27 20:06:57 UTC] RC5-72: Benchmark for core #5 (CUDA 2-pipe 256-thd) 0.00:00:01.87 [2,375,854,918 keys/sec] [Jun 27 20:07:23 UTC] RC5-72: using core #6 (CUDA 4-pipe 64-thd). [Jun 27 20:07:26 UTC] RC5-72: Benchmark for core #6 (CUDA 4-pipe 64-thd) 0.00:00:01.92 [2,292,168,845 keys/sec] [Jun 27 20:07:43 UTC] RC5-72: using core #7 (CUDA 4-pipe 128-thd). [Jun 27 20:07:47 UTC] RC5-72: Benchmark for core #7 (CUDA 4-pipe 128-thd) 0.00:00:01.90 [2,329,162,412 keys/sec] [Jun 27 20:08:06 UTC] RC5-72: using core #8 (CUDA 4-pipe 256-thd). [Jun 27 20:08:09 UTC] RC5-72: Benchmark for core #8 (CUDA 4-pipe 256-thd) 0.00:00:01.92 [2,327,490,786 keys/sec] [Jun 27 20:08:30 UTC] RC5-72: using core #9 (CUDA 1-pipe 64-thd busy wait). [Jun 27 20:08:34 UTC] RC5-72: Benchmark for core #9 (CUDA 1-pipe 64-thd busy wait) 0.00:00:01.95 [2,271,437,027 keys/sec] [Jun 27 20:09:56 UTC] RC5-72: using core #10 (CUDA 1-pipe 64-thd sleep 100us). [Jun 27 20:10:14 UTC] RC5-72: Benchmark for core #10 (CUDA 1-pipe 64-thd sleep 100us) 0.00:00:16.18 [229,614,809 keys/sec] [Jun 27 20:11:17 UTC] RC5-72: using core #11 (CUDA 1-pipe 64-thd sleep dynamic). [Jun 27 20:11:21 UTC] RC5-72: Benchmark for core #11 (CUDA 1-pipe 64-thd sleep dynamic) 0.00:00:02.01 [2,194,360,995 keys/sec] |