I've Built a Couple OSX CUDA Apps...

Author	Message
Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1822448 - Posted: 7 Oct 2016, 11:14:28 UTC Seems we have quite similar situation in OS vs Vendors world BTW: On Windows: M$ drivers for NV will support CUDA (AFAIK) but lack of OpenCL support. On OS X: Apple provide OpenCL support by itself (broken one though) and lack of CUDA support. In both cases vendor driver installation required. Just Mac users maybe less familiar with such situation. M$ started to provide own non-trivial (non-VGA-only) drivers not so long ago so users familiar with th need of vendor driver installation. SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1822448 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826154 - Posted: 22 Oct 2016, 18:18:01 UTC Last modified: 22 Oct 2016, 18:18:53 UTC Try to rebuild from r3548 (OpenCL NV and iGPU) SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826154 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826254 - Posted: 23 Oct 2016, 6:46:26 UTC - in response to Message 1826154. I'm getting some strange results with this build in Darwin 15.6 on a GTX 950; Build features: SETI8 Non-graphics OpenCL USE_OPENCL_NV OCL_CHIRP3 ASYNC_SPIKE FFTW SSSE3 64bit The first task ran was the quick BLC3 overflow and it looked very good; Strongly similar, Q= 99.97% The second task was the reference_work_unit_r3215.wu which was also good; Strongly similar, Q= 99.49% Then I tried a longer BLC2 overflow and it wasn't so good; Unmatched signal(s) in R1 at line(s) 340 368 384 400 416 432 448 464 491 517 534 559 592 608 625 646 678 Result : Different. I tried a few more normal length BLCs and they are also bad; Unmatched signal(s) in R1 at line(s) 383 409 431 459 481 506 522 539 560 For R1:R2 matched signals only, Q= 99.99% Result : Different. Unmatched signal(s) in R1 at line(s) 340 356 372 388 404 420 437 454 481 497 513 530 547 563 579 595 622 638 654 680 705 731 758 789 805 822 843 873 Result : Different. It would seem it's good on the Quick BLC overflows and Arecibo tasks but very bad on the normal BLCs. The matched signals are very good, but, it's not matching many signals on the normal BLC tasks. ?? ID: 1826254 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826258 - Posted: 23 Oct 2016, 7:16:47 UTC - in response to Message 1826254. 1) Why SIGNALS_ON_GPU define omitted? 2) Please post links to tasks and logs with results. SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826258 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826261 - Posted: 23 Oct 2016, 7:52:32 UTC - in response to Message 1826258. Last modified: 23 Oct 2016, 8:11:54 UTC Probably 'cause I wanted to try the 'normal' build first. I'm not fond of an App that bounces the progress bar around to the point I fear catching Hillary eyes ;) I'm trying the SoG build now...and it's using Much more CPU than the 'normal' build. How can the thing use 102.8% CPU? KWSN-Darwin-MBbench v2.1.07 Running on TomsMacPro.local at Sun Oct 23 07:24:26 2016 --------------------------------------------------- Starting benchmark run... --------------------------------------------------- Listing wu-file(s) in /testWUs : blc5_2bit_guppi_57449_43932_HIP78775_0013.26700.831.18.27.53.vlar.wu blc5_2bit_guppi_57449_45355_HIP81348_0017.14390.416.17.26.21.vlar.wu reference_work_unit_r3215.wu Listing executable(s) in /APPS : MBv8_8.18r3549_NV-SoG_ssse3_x86_64-apple-darwin Listing executable in /REF_APPs : MBv8_8.05r3344_sse41_x86_64-apple-darwin --------------------------------------------------- Current WU: blc5_2bit_guppi_57449_43932_HIP78775_0013.26700.831.18.27.53.vlar.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 8062 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3549_NV-SoG_ssse3_x86_64-apple-darwin -period_iterations_num 16 -device 0 If it works, I'll think about running it in BOINC. It should be finished with the first one soon, it seems to be taking longer than the other build. I'm getting this deja vu feeling. Now that I think about it, I remember the SoG App taking MUCH longer than it should. I believe if you go back in this thread you'll see where the SoG App took around 40 Minutes to complete a Shorty that should have finished in 7 Minutes. This is a BLC5, it's been running for 50 minutes, I've got a Bad feeling... ID: 1826261 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826272 - Posted: 23 Oct 2016, 10:14:16 UTC Last modified: 23 Oct 2016, 10:21:54 UTC Yep, that's about what 6 times longer would be, Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 7700 seconds :-( Hey, it gave the correct results though, even if it did take almost as long as the CPU would have taken by itself. So, why does the non-SoG build give the wrong BLC results (it actually crashed on that particular task), and why does the SoG App take 6 times as long to finish on both tasks? Running app with command : MBv8_8.18r3549_NV-SoG_ssse3_x86_64-apple-darwin -period_iterations_num 16 -device 0 7700.49 real 6451.20 user 254.82 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 7700 seconds Speed compared to default : 104 % ----------------- Comparing results Result : Strongly similar, Q= 99.96% --------------------------------------------------- Done with blc5_2bit_guppi_57449_43932_HIP78775_0013.26700.831.18.27.53.vlar.wu. blc5_2bit_guppi_57449_45355_HIP81348_0017.14390.416.17.26.21.vlar.wu does not exist. blc5_2bit_guppi_57449_45355_HIP81348_0017.14390.416.17.26.21.vlar.wu does not exist. Current WU: reference_work_unit_r3215.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 2521 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3549_NV-SoG_ssse3_x86_64-apple-darwin -period_iterations_num 16 -device 0 975.99 real 642.92 user 155.15 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 976 seconds Speed compared to default : 258 % ----------------- Comparing results Result : Strongly similar, Q= 99.51% --------------------------------------------------- Done with reference_work_unit_r3215.wu. I didn't bother running the second BLC5 task. stderr.MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin.blc5_2bit_guppi_57449_43932_HIP78775_0013.26700.831.18.27.53.vlar.wu.txt; Build features: SETI8 Non-graphics OpenCL USE_OPENCL_NV OCL_CHIRP3 ASYNC_SPIKE FFTW SSSE3 64bit System: Darwin x86_64 Kernel: 15.6.0 CPU : Intel(R) Xeon(R) CPU E5472 @ 3.00GHz ... Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.010092 Used GPU device parameters are: Number of compute units: 6 Single buffer allocation size: 128MB Total device global memory: 2048MB max WG size: 1024 local mem type: Real FERMI path used: yes LotOfMem path: no LowPerformanceGPU path: no HighPerformanceGPU path: no period_iterations_num=16 Pulse: peak=5.321574, time=45.99, period=13.06, d_freq=1228150097.09, score=1.003, chirp=-4.4361, fft_len=4k Autocorr: peak=18.36389, time=74.45, delay=5.4228, d_freq=1228144214.73, chirp=-18.592, fft_len=128k Pulse: peak=2.62872, time=45.9, period=5.577, d_freq=1228148715.01, score=1.019, chirp=-61.396, fft_len=2k Pulse: peak=6.128602, time=45.82, period=11.56, d_freq=1228147742.68, score=1.031, chirp=70.205, fft_len=128 Pulse: peak=3.743349, time=45.86, period=8.277, d_freq=1228143446.26, score=1.001, chirp=-77.161, fft_len=1024 Pulse: peak=5.850301, time=45.82, period=12.46, d_freq=1228146038.13, score=1.004, chirp=79.825, fft_len=256 MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin(31714,0x700000122000) malloc: * error for object 0x1292d8000: pointer being freed was not allocated * set a breakpoint in malloc_error_break to debug SIGABRT: abort called Crashed executable name: MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin Machine type Intel 80486 (64-bit executable) System version: Macintosh OS 10.11.6 build 15G1004 Sun Oct 23 02:56:11 2016 0 MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin 0x0000000105c03438 std::__1::__tree<std::__1::__value_type<int, PROCINFO>, std::__1::__map_value_compare<int, std::__1::__value_type<int, PROCINFO>, std::__1::less<int>, true>, std::__1::allocator<std::__1::__value_type<int, PROCINFO> > >::__insert_unique(std::__1::__value_type<int, PROCINFO> const&) + 1080 1 MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin 0x0000000105bf2c66 COPROCS::clear() + 4262 2 libsystem_platform.dylib 0x00007fff9a06152a _sigtramp + 26 3 libsystem_malloc.dylib 0x00007fff8d68e5a1 malloc_zone_malloc + 71 4 libsystem_c.dylib 0x00007fff882926df abort + 129 5 libsystem_malloc.dylib 0x00007fff8d690041 szone_size + 0 6 GeForceGLDriverWeb 0x0000000107f8c18a gldWaitForObject + 4947 7 GeForceGLDriverWeb 0x0000000107f95e1b gldExecuteKernel + 201 8 OpenCL 0x00007fff9a6844a7 OpenCL + 13479 9 OpenCL 0x00007fff9a6a10da clSetEventCallback + 5888 10 OpenCL 0x00007fff9a6a49cc clFinish + 761 11 libdispatch.dylib 0x00007fff95b0c40b _dispatch_client_callout + 8 12 libdispatch.dylib 0x00007fff95b1103b _dispatch_queue_drain + 754 13 libdispatch.dylib 0x00007fff95b17707 _dispatch_queue_invoke + 549 14 libdispatch.dylib 0x00007fff95b0c40b _dispatch_client_callout + 8 15 libdispatch.dylib 0x00007fff95b1029b _dispatch_root_queue_drain + 1890 16 libdispatch.dylib 0x00007fff95b0fb00 _dispatch_worker_thread3 + 91 17 libsystem_pthread.dylib 0x00007fff9aa1f4de _pthread_wqthread + 1129 18 libsystem_pthread.dylib 0x00007fff9aa1d341 start_wqthread + 13 Thread 4 crashed with X86 Thread State (64-bit): rax: 0x0100001f rbx: 0x00000000 rcx: 0x70000011f4f8 rdx: 0x00000028 rdi: 0x70000011f560 rsi: 0x00000003 rbp: 0x70000011f540 rsp: 0x70000011f4f8 r8: 0x0000260f r9: 0x00000000 r10: 0x000003b0 r11: 0x00000206 r12: 0x000003b0 r13: 0x00000028 r14: 0x70000011f560 r15: 0x0000260f rip: 0x7fff8e5fef72 rfl: 0x00000206 stderr.MBv8_8.18r3549_NV-SoG_ssse3_x86_64-apple-darwin.blc5_2bit_guppi_57449_43932_HIP78775_0013.26700.831.18.27.53.vlar.wu.txt Build features: SETI8 Non-graphics OpenCL USE_OPENCL_NV OCL_ZERO_COPY SIGNALS_ON_GPU OCL_CHIRP3 FFTW SSSE3 64bit System: Darwin x86_64 Kernel: 15.6.0 CPU : Intel(R) Xeon(R) CPU E5472 @ 3.00GHz ... Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.010092 Used GPU device parameters are: Number of compute units: 6 Single buffer allocation size: 128MB Total device global memory: 2048MB max WG size: 1024 local mem type: Real FERMI path used: yes LotOfMem path: yes LowPerformanceGPU path: no HighPerformanceGPU path: no period_iterations_num=16 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 Pulse: peak=5.321574, time=45.99, period=13.06, d_freq=1228150097.09, score=1.003, chirp=-4.4361, fft_len=4k cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 Autocorr: peak=18.36389, time=74.45, delay=5.4228, d_freq=1228144214.73, chirp=-18.592, fft_len=128k cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 Pulse: peak=2.62872, time=45.9, period=5.577, d_freq=1228148715.01, score=1.019, chirp=-61.396, fft_len=2k cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 Pulse: peak=6.128602, time=45.82, period=11.56, d_freq=1228147742.68, score=1.031, chirp=70.205, fft_len=128 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 Pulse: peak=3.743349, time=45.86, period=8.277, d_freq=1228143446.26, score=1.001, chirp=-77.161, fft_len=1024 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 Pulse: peak=5.850302, time=45.82, period=12.46, d_freq=1228146038.13, score=1.004, chirp=79.825, fft_len=256 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 Pulse: peak=0.5505553, time=45.82, period=0.4381, d_freq=1228141965.11, score=1.021, chirp=-94.907, fft_len=256 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 cpu_GPUState->gaussians.index=0 Best spike: peak=23.57418, time=74.45, d_freq=1228151239.83, chirp=-23.93, fft_len=128k Best autocorr: peak=18.36389, time=74.45, delay=5.4228, d_freq=1228144214.73, chirp=-18.592, fft_len=128k Best gaussian: peak=0, mean=0, ChiSq=0, time=-2.123e+11, d_freq=0, score=-12, null_hyp=0, chirp=0, fft_len=0 Best pulse: peak=6.128602, time=45.82, period=11.56, d_freq=1228147742.68, score=1.031, chirp=70.205, fft_len=128 Best triplet: peak=0, time=-2.123e+11, period=0, d_freq=0, chirp=0, fft_len=0 Flopcounter: 4199755425882.255859 Spike count: 0 Autocorr count: 1 Pulse count: 6 Triplet count: 0 Gaussian count: 0 Time cpu in use since last restart: 6705.7 seconds ... And....What's with all the "cpu_GPUState->gaussians.index=0" entries? ID: 1826272 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826274 - Posted: 23 Oct 2016, 10:27:22 UTC - in response to Message 1826272. Hm. If crash occured about what time comparisons and valid/invalid results you talking?? App crashed <=> computation error, end of story of benching, debugging mode on... 6 times longer or whatever - doesn't matter at all. Can you locate function call in crash place? I see no my fuctions in provided crash log at all... SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826274 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826276 - Posted: 23 Oct 2016, 10:30:47 UTC - in response to Message 1826274. Check iGPU build awhile. I'll check windows NV non-SoG one meantime. SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826276 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826278 - Posted: 23 Oct 2016, 10:34:41 UTC - in response to Message 1826261. I'm getting this deja vu feeling. Now that I think about it, I remember the SoG App taking MUCH longer than it should. I believe if you go back in this thread you'll see where the SoG App took around 40 Minutes to complete a Shorty that should have finished in 7 Minutes. Does Mac have any CPU/GPU profiling/monitoring tools? What CPU/GPU activity during such run? SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826278 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826282 - Posted: 23 Oct 2016, 11:16:28 UTC - in response to Message 1826274. Last modified: 23 Oct 2016, 11:35:57 UTC Hm. If crash occured about what time comparisons and valid/invalid results you talking?? App crashed <=> computation error, end of story of benching, debugging mode on... 6 times longer or whatever - doesn't matter at all. Can you locate function call in crash place? I see no my fuctions in provided crash log at all... Hmmmm, I have no idea how the benchmark can keep going after the App crashes. However, it seems the non-SoG App crashed both times. Here's the other crash; stderr.MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin.blc5_2bit_guppi_57449_45355_HIP81348_0017.14390.416.17.26.21.vlar.wu.txt 02:56:11 (31842): Can't set up shared mem: -1. Will run in standalone mode. Build features: SETI8 Non-graphics OpenCL USE_OPENCL_NV OCL_CHIRP3 ASYNC_SPIKE FFTW SSSE3 64bit Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.012361 Used GPU device parameters are: Number of compute units: 6 Single buffer allocation size: 128MB Total device global memory: 2048MB max WG size: 1024 local mem type: Real FERMI path used: yes LotOfMem path: no LowPerformanceGPU path: no HighPerformanceGPU path: no period_iterations_num=16 MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin(31842,0x7000001a5000) malloc: * error for object 0x117fc8000: pointer being freed was not allocated * set a breakpoint in malloc_error_break to debug SIGABRT: abort called Crashed executable name: MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin Machine type Intel 80486 (64-bit executable) System version: Macintosh OS 10.11.6 build 15G1004 Sun Oct 23 02:57:12 2016 0 MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin 0x0000000102707438 std::__1::__tree<std::__1::__value_type<int, PROCINFO>, std::__1::__map_value_compare<int, std::__1::__value_type<int, PROCINFO>, std::__1::less<int>, true>, std::__1::allocator<std::__1::__value_type<int, PROCINFO> > >::__insert_unique(std::__1::__value_type<int, PROCINFO> const&) + 1080 1 MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin 0x00000001026f6c66 COPROCS::clear() + 4262 2 libsystem_platform.dylib 0x00007fff9a06152a _sigtramp + 26 3 libsystem_malloc.dylib 0x00007fff8d68e5a1 malloc_zone_malloc + 71 4 libsystem_c.dylib 0x00007fff882926df abort + 129 5 libsystem_malloc.dylib 0x00007fff8d690041 szone_size + 0 6 GeForceGLDriverWeb 0x0000000104b4c18a gldWaitForObject + 4947 7 GeForceGLDriverWeb 0x0000000104b55e1b gldExecuteKernel + 201 8 OpenCL 0x00007fff9a6844a7 OpenCL + 13479 9 OpenCL 0x00007fff9a6a10da clSetEventCallback + 5888 10 OpenCL 0x00007fff9a6a49cc clFinish + 761 11 libdispatch.dylib 0x00007fff95b0c40b _dispatch_client_callout + 8 12 libdispatch.dylib 0x00007fff95b1103b _dispatch_queue_drain + 754 13 libdispatch.dylib 0x00007fff95b17707 _dispatch_queue_invoke + 549 14 libdispatch.dylib 0x00007fff95b0c40b _dispatch_client_callout + 8 15 libdispatch.dylib 0x00007fff95b1029b _dispatch_root_queue_drain + 1890 16 libdispatch.dylib 0x00007fff95b0fb00 _dispatch_worker_thread3 + 91 17 libsystem_pthread.dylib 0x00007fff9aa1f4de _pthread_wqthread + 1129 18 libsystem_pthread.dylib 0x00007fff9aa1d341 start_wqthread + 13 Thread 5 crashed with X86 Thread State (64-bit): rax: 0x0100001f rbx: 0x00000000 rcx: 0x7000001a24f8 rdx: 0x00000028 rdi: 0x7000001a2560 rsi: 0x00000003 rbp: 0x7000001a2540 rsp: 0x7000001a24f8 r8: 0x00004203 r9: 0x00000000 r10: 0x000003b0 r11: 0x00000206 r12: 0x000003b0 r13: 0x00000028 r14: 0x7000001a2560 r15: 0x00004203 rip: 0x7fff8e5fef72 rfl: 0x00000206 I've got a testData folder full of files, I suppose I'll clear them out and try the App again and see how it goes. I'm pretty sure this is the result from the above crash; Done with blc5_2bit_guppi_57449_43932_HIP78775_0013.26700.831.18.27.53.vlar.wu. Current WU: blc5_2bit_guppi_57449_45355_HIP81348_0017.14390.416.17.26.21.vlar.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 8289 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin -sbs 192 -period_iterations_num 16 -device 0 Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 60 seconds Speed compared to default : 13815 % ----------------- Comparing results ------------- R1:R2 ------------ ------------- R2:R1 ------------ Exact Super Tight Good Bad Exact Super Tight Good Bad Spike 0 0 0 0 12 0 0 0 0 0 Autocorr 0 0 0 0 2 0 0 0 0 0 Gaussian 0 0 0 0 0 0 0 0 0 0 Pulse 0 0 0 0 7 0 0 0 0 0 Triplet 0 0 0 0 2 0 0 0 0 0 Best Spike 0 0 0 0 1 0 0 0 0 0 Best Autocorr 0 0 0 0 1 0 0 0 0 0 Best Gaussian 0 0 0 0 1 0 0 0 0 0 Best Pulse 0 0 0 0 1 0 0 0 0 0 Best Triplet 0 0 0 0 1 0 0 0 0 0 ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- 0 0 0 0 28 0 0 0 0 0 Unmatched signal(s) in R1 at line(s) 340 356 372 388 404 420 437 454 481 497 513 530 547 563 579 595 622 638 654 680 705 731 758 789 805 822 843 873 Result : Different. --------------------------------------------------- Done with blc5_2bit_guppi_57449_45355_HIP81348_0017.14390.416.17.26.21.vlar.wu. Current WU: reference_work_unit_r3215.wu It looks as though the first task just about finished before it crashed. This should be the result from the other crash; Current WU: blc5_2bit_guppi_57449_43932_HIP78775_0013.26700.831.18.27.53.vlar.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 8062 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin -sbs 192 -period_iterations_num 16 -device 0 Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 1242 seconds Speed compared to default : 649 % Can you locate function call in crash place? I'm not sure what to look for, can you give an example. ID: 1826282 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826302 - Posted: 23 Oct 2016, 13:30:29 UTC - in response to Message 1826276. Last modified: 23 Oct 2016, 13:46:50 UTC Check iGPU build awhile. I'll check windows NV non-SoG one meantime. The nVidia non-SoG App keeps crashing at differing points. On the last try the tasks crashed at; Current WU: blc5_2bit_guppi_57449_43932_HIP78775_0013.26700.831.18.27.53.vlar.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 8062 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin -period_iterations_num 10 -device 0 79.28 real 20.34 user 12.57 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 79 seconds & Current WU: blc5_2bit_guppi_57449_45355_HIP81348_0017.14390.416.17.26.21.vlar.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 8289 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin -period_iterations_num 10 -device 0 358.30 real 80.22 user 54.49 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 358 seconds Always with the same reason; period_iterations_num=10 Spike: peak=24.87894, time=66.57, d_freq=1154537601.45, chirp=-1.7059, fft_len=16k Spike: peak=25.13841, time=66.57, d_freq=1154537601.47, chirp=-2.0308, fft_len=16k Spike: peak=24.60172, time=17.18, d_freq=1154537341.87, chirp=-3.7647, fft_len=128k Spike: peak=24.60887, time=17.18, d_freq=1154537341.87, chirp=-3.7698, fft_len=128k Spike: peak=24.22487, time=17.18, d_freq=1154537341.85, chirp=-3.771, fft_len=128k Autocorr: peak=18.03834, time=40.09, delay=1.0461, d_freq=1154537413.81, chirp=5.3107, fft_len=128k Autocorr: peak=18.36849, time=40.09, delay=1.0461, d_freq=1154537413.91, chirp=5.3132, fft_len=128k MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin(38011,0x70000008c000) malloc: * error for object 0x1183f0000: pointer being freed was not allocated * set a breakpoint in malloc_error_break to debug SIGABRT: abort called Crashed executable name: MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin The Intel build seems to be working even though it is running on the Exact same GTX 950; Build features: SETI8 Non-graphics OpenCL USE_OPENCL_INTEL OCL_CHIRP3 ASYNC_SPIKE FFTW SSSE3 64bit System: Darwin x86_64 Kernel: 15.6.0 Current WU: blc2_2bit_guppi_57403_HIP11048_0006.17091.831.22.45.71.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 4797 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin -device 0 903.57 real 147.08 user 296.94 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 904 seconds Speed compared to default : 530 % ----------------- Comparing results Result : Strongly similar, Q= 99.89% --------------------------------------------------- Done with blc2_2bit_guppi_57403_HIP11048_0006.17091.831.22.45.71.wu. Current WU: reference_work_unit_r3215.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 2110 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin -device 0 344.60 real 110.29 user 92.96 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 345 seconds Speed compared to default : 611 % ----------------- Comparing results Result : Strongly similar, Q= 99.51% This is good...much better than the Q= 19.x% I was getting with the r3541 build. ID: 1826302 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826317 - Posted: 23 Oct 2016, 15:34:26 UTC - in response to Message 1826278. Last modified: 23 Oct 2016, 16:01:26 UTC I'm getting this deja vu feeling. Now that I think about it, I remember the SoG App taking MUCH longer than it should. I believe if you go back in this thread you'll see where the SoG App took around 40 Minutes to complete a Shorty that should have finished in 7 Minutes. Does Mac have any CPU/GPU profiling/monitoring tools? What CPU/GPU activity during such run? You have the Activity monitor which gives basic info on CPU load and memory use. As far as I know there isn't anything that gives detailed GPU info. Looking at the task run this morning with MBv8_8.18r3549_NV-SoG_ssse3_x86_64-apple-darwin the Run times were; GPU: 7700 seconds CPU: 6705.7 seconds That's much higher than the non-SoG App which just finished the First Full BLC without crashing; Current WU: blc5_2bit_guppi_57449_45695_HIP81348_OFF_0018.29291.831.17.26.236.vlar.wu --------------------------------------------------- Running default app with command : MBv8_8.05r3344_sse41_x86_64-apple-darwin 7331.44 real 7313.16 user 15.37 sys Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 7332 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3548_NV_ssse3_x86_64-apple-darwin -period_iterations_num 10 -device 0 2071.73 real 466.85 user 268.64 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 2072 seconds Speed compared to default : 353 % ----------------- Comparing results Result : Strongly similar, Q= 99.93% Kinda slow, but, much better than the 7700 seconds it took the SoG App. If you could fix the crashing maybe a few settings might speed it up. I changed the settings on the Intel App and it's working a little faster. So far the Intel App is working well, but, I don't have an Intel GPU to test it on; GPU not found: type=intel_gpu, opencl_device_index=-1, device_num=2 WARNING: boinc_get_opencl_ids failed with code -1 OpenCL platform detected: Apple WARNING: BOINC supplied wrong platform! Number of OpenCL devices found : 3 BOINC assigns slot on device #3 of 3 devices. WARNING: BOINC failed to provide OpenCL device, using own enumeration abilities Build features: SETI8 Non-graphics OpenCL USE_OPENCL_INTEL OCL_CHIRP3 ASYNC_SPIKE FFTW SSSE3 64bit But, it does work on the GTX 950, and I haven't had any problems yet. It finished that Problem Quick Overflow BLC3 without any trouble, and even had the correct stderr. ID: 1826317 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826330 - Posted: 23 Oct 2016, 16:40:25 UTC - in response to Message 1826302. Result : Strongly similar, Q= 99.51%[/pre] This is good...much better than the Q= 19.x% I was getting with the r3541 build. So, you have no target hardware for iGPU build ? Could you post binary for someone with iGPU-capable Mac could do test. Also, on Windows I use OCL_SYNCHED define for iGPU build. This allows to considerably reduce CPU consumption. Try to provide another iGPU build with such define enabled for external testing. SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826330 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826337 - Posted: 23 Oct 2016, 17:32:41 UTC Last modified: 23 Oct 2016, 17:33:48 UTC Build options line for Windows NV non-SoG build: USE_JSPF;OCL_ZERO_COPY;USE_OPENCL_NV;OCL_CHIRP3;USE_OPENCL;ATI_OS_WIN;SETI7;SETI8;USE_FFTW; WIN32;_WIN32;_MT;NDEBUG;_WINDOWS;CLIENT;_CONSOLE;USE_I386_OPTIMIZATIONS;USE_I386_XEON;USE_SSE3 Will see if it causes any crashes. SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826337 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826359 - Posted: 23 Oct 2016, 20:51:07 UTC - in response to Message 1826337. It passed PG set 3 times w/o crash. SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826359 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826417 - Posted: 24 Oct 2016, 4:38:33 UTC - in response to Message 1826359. Last modified: 24 Oct 2016, 4:39:07 UTC It's nice it doesn't crash on your Windows machine, too bad it crashes on my Mac. You know, I've compiled dozens of these Apps in the past and none of them crashed, the r3541 from a couple days ago doesn't crash either. The OCL_SYNCHED doesn't appear to work very well on my Mac either. It actually Slows down the App and uses More CPU time. The top two tasks are not using OCL_SYNCHED, the last two tasks are using the App with OCL_SYNCHED; KWSN-Darwin-MBbench v2.1.07 Running on TomsMacPro.local at Mon Oct 24 03:42:33 2016 --------------------------------------------------- Starting benchmark run... --------------------------------------------------- Listing wu-file(s) in /testWUs : 11au16aa.28481.85822.12.39.56.wu reference_work_unit_r3215.wu Listing executable(s) in /APPS : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin Listing executable in /REF_APPs : MBv8_8.05r3344_sse41_x86_64-apple-darwin --------------------------------------------------- Current WU: 11au16aa.28481.85822.12.39.56.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 3630 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin -sbs 128 -period_iterations_num 10 -device 2 470.81 real 81.76 user 138.66 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 471 seconds Speed compared to default : 770 % ----------------- Comparing results Result : Strongly similar, Q= 98.12% --------------------------------------------------- Done with 11au16aa.28481.85822.12.39.56.wu. Current WU: reference_work_unit_r3215.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 2521 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin -sbs 128 -period_iterations_num 10 -device 2 327.30 real 105.53 user 87.96 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 327 seconds Speed compared to default : 770 % ----------------- Comparing results Result : Strongly similar, Q= 99.51% --------------------------------------------------- KWSN-Darwin-MBbench v2.1.07 Running on TomsMacPro.local at Mon Oct 24 03:58:28 2016 --------------------------------------------------- Starting benchmark run... --------------------------------------------------- Listing wu-file(s) in /testWUs : 11au16aa.28481.85822.12.39.56.wu reference_work_unit_r3215.wu Listing executable(s) in /APPS : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin Listing executable in /REF_APPs : MBv8_8.05r3344_sse41_x86_64-apple-darwin --------------------------------------------------- Current WU: 11au16aa.28481.85822.12.39.56.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 3630 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin -sbs 128 -period_iterations_num 10 -device 2 593.23 real 112.91 user 340.85 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 593 seconds Speed compared to default : 612 % ----------------- Comparing results Result : Strongly similar, Q= 98.12% --------------------------------------------------- Done with 11au16aa.28481.85822.12.39.56.wu. Current WU: reference_work_unit_r3215.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 2521 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin -sbs 128 -period_iterations_num 10 -device 2 405.23 real 126.03 user 194.89 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 406 seconds Speed compared to default : 620 % ----------------- Comparing results Result : Strongly similar, Q= 99.51% --------------------------------------------------- Done with reference_work_unit_r3215.wu. It's a shame the App compiled with the -DUSE_OPENCL_NV tag doesn't work as well as the one compiled with -DUSE_OPENCL_INTEL. That is the only difference in the two configure lines, yet, one works well while the other crashes. ID: 1826417 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826441 - Posted: 24 Oct 2016, 7:43:15 UTC I've posted both versions of the Mac Intel App for anyone with an Intel iGPU that wishes to test them. I also included the Preset KWSN-OSX-bench-MB_v2.1.07 Package for easy testing. Just move the desired App into the APPS folder and run the benchmark App from the terminal. It will run the CPU App first and then the iGPU App and compare the results. Mac Intel iGPU Test Apps I'm not going to hold my breath, the last time I tried this the App worked fine on my DeskTop GPU but gave the same Inconclusive results when run on an iGPU. Good Luck. It should work on your Desktop GPUs just fine though, it works on my nVidia GPUs. Too bad I can't get the 'real' nVidia OpenCL App to work. ID: 1826441 ·

Raistmer Volunteer developer Volunteer tester Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121	Message 1826443 - Posted: 24 Oct 2016, 7:46:45 UTC - in response to Message 1826417. It's nice it doesn't crash on your Windows machine, too bad it crashes on my Mac. That means crash in OS-specific part. Though I made single change in source code - disabled debug output you saw in SoG. I'll committ that change soon so try again with new rev. SETI apps news We're not gonna fight them. We're gonna transcend them. ID: 1826443 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826508 - Posted: 24 Oct 2016, 16:05:22 UTC - in response to Message 1826443. I switched the Mac over to running the Intel build on Main. So far all the tasks have validated on the first try. It's easy to tell which tasks are the OpenCL ones, just look for the ones with Run times over 1000 seconds, http://setiathome.berkeley.edu/result.php?resultid=5236104866. I ran down most of the Arecibo tasks before making the switch, so, the OpenCL App is running BLC tasks for now. I see no one has downloaded the test Package yet. It would be very helpful if someone were to try this App on their Intel iGPU and see if it produces First Time Validations. Keeping from having to send extra tasks to validate Inconclusive results means much more work can be accomplished with the same number of machines. Mac Intel iGPU Test Apps ID: 1826508 ·

TBar Volunteer tester Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768	Message 1826717 - Posted: 25 Oct 2016, 23:41:28 UTC So far, one person has downloaded the App. I decided to test the current Intel iGPU app at Beta against the new El Capitan version. The App at Beta has the same 'accuracy' problem as the other MB OpenCL nVidia Apps in El Capitan & Sierra. So basically, this new version is the only MB OpenCL App in the Known Universe that gives the correct results on an nVidia GPU in El Capitan 11.4+. Surely it's worth some time to see how it works on an iGPU. The below test was run in Darwin 15.6 on a GTX 950. As you can see, the 'Q' level on the older App is dismal. KWSN-Darwin-MBbench v2.1.07 Running on TomsMacPro.local at Tue Oct 25 15:08:19 2016 --------------------------------------------------- Starting benchmark run... --------------------------------------------------- Listing wu-file(s) in /testWUs : 11au16aa.28481.85822.12.39.56.wu reference_work_unit_r3215.wu Listing executable(s) in /APPS : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin setiathome_8.10_x86_64-apple-darwin__opencl_intel_gpu_sah Listing executable in /REF_APPs : MBv8_8.05r3344_sse41_x86_64-apple-darwin --------------------------------------------------- Current WU: 11au16aa.28481.85822.12.39.56.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 3630 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin -sbs 128 -period_iterations_num 8 -device 2 467.72 real 80.70 user 132.45 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 468 seconds Speed compared to default : 775 % ----------------- Comparing results Result : Strongly similar, Q= 98.12% --------------------------------------------------- Running app with command : setiathome_8.10_x86_64-apple-darwin__opencl_intel_gpu_sah -sbs 128 -period_iterations_num 8 -device 2 472.09 real 67.94 user 111.96 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 472 seconds Speed compared to default : 769 % ----------------- Comparing results Unmatched signal(s) in R1 at line(s) 776 Unmatched signal(s) in R2 at line(s) 373 For R1:R2 matched signals only, Q= ???? Result : Weakly similar. --------------------------------------------------- Done with 11au16aa.28481.85822.12.39.56.wu. Current WU: reference_work_unit_r3215.wu --------------------------------------------------- Skipping default app MBv8_8.05r3344_sse41_x86_64-apple-darwin, displaying saved result(s) Elapsed Time: â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 2521 seconds --------------------------------------------------- Running app with command : MBv8_8.18r3550_Intel_ssse3_x86_64-apple-darwin -sbs 128 -period_iterations_num 8 -device 2 325.90 real 104.35 user 83.90 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 326 seconds Speed compared to default : 773 % ----------------- Comparing results Result : Strongly similar, Q= 99.51% --------------------------------------------------- Running app with command : setiathome_8.10_x86_64-apple-darwin__opencl_intel_gpu_sah -sbs 128 -period_iterations_num 8 -device 2 328.09 real 98.16 user 69.22 sys Elapsed Time : â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 328 seconds Speed compared to default : 768 % ----------------- Comparing results Unmatched signal(s) in R1 at line(s) 499 526 580 607 634 694 720 Unmatched signal(s) in R2 at line(s) 482 509 526 569 595 649 676 703 763 789 For R1:R2 matched signals only, Q= 8.044% Result : Weakly similar. --------------------------------------------------- ID: 1826717 ·

©2024 University of California

SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.