Message boards :
Number crunching :
TestCase: blc25_2bit_guppi_57895_47387_HIP91358_0034.24610.818.23.46.191.vlar
Message board moderation
Author | Message |
---|---|
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
All supplement materials including task itself for this TestCase can be found: https://cloud.mail.ru/public/B3uZ/cQvL86wF9 I encourage testers with different OSes try to run this task offline with different available apps and provide results for comparison. Direct comparison between iGPU stock and x86 windows stock: D:\R\_bench_MB.1\Testdatas\ref>rescmpv5.exe ref-setiathome_8.20_windows_intelx86__opencl_intel_gpu_sah.exe-blc25_2bit_guppi_57895_47387_HIP91358_0 ------------- R1:R2 ------------ ------------- R2:R1 ------------ Exact Super Tight Good Bad Exact Super Tight Good Bad Spike 0 5 5 5 0 0 5 5 5 0 Autocorr 0 0 0 0 0 0 0 0 0 0 Gaussian 0 0 0 0 0 0 0 0 0 0 Pulse 0 9 10 10 0 0 9 10 10 1 Triplet 0 3 3 3 0 0 3 3 3 0 Best Spike 0 1 1 1 0 0 1 1 1 0 Best Autocorr 0 1 1 1 0 0 1 1 1 0 Best Gaussian 1 1 1 1 0 1 1 1 1 0 Best Pulse 0 1 1 1 0 0 1 1 1 0 Best Triplet 0 1 1 1 0 0 1 1 1 0 ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- 1 22 23 23 0 1 22 23 23 1 Unmatched signal(s) in R2 at line(s) 491 For R1:R2 matched signals only, Q= 98.97% Result : Weakly similar. SETI apps news We're not gonna fight them. We're gonna transcend them. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Unfortunately that very app that provides additional 11th pulse is most silent one so it's impossible to track its decisions regarding that missing pulse or check thresholds of it's detection. SETI apps news We're not gonna fight them. We're gonna transcend them. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
The 11th signal is <pulse> <peak_power>6.3387684822083</peak_power> <mean_power>0.12848031520844</mean_power> <time>2457896.0506873</time> <ra>18.632833600614</ra> <decl>-1.8170919815812</decl> <q_pix>0</q_pix> <freq>2082850157.842</freq> <detection_freq>2082851430.8309</detection_freq> <barycentric_freq>0</barycentric_freq> <fft_len>2048</fft_len> <chirp_rate>27.732474062542</chirp_rate> <rfi_checked>0</rfi_checked> <rfi_found>0</rfi_found> <reserved>0</reserved> <period>18.074654037333</period> <snr>14.173917770386</snr> <thresh>14.173913955688</thresh> <score>0</score> <len_prof>101</len_prof> <pot length=319 encoding="x-csv"> 10,17,28,36,7,12,29,18,25,17,52,51,28,13,14,38,34,7,30,22,16,8,23,23,8, 11,23,30,36,21,0,34,43,19,26,18,42,50,22,3,3,10,34,11,45,27,20,26,23,13, 22,5,43,24,41,31,12,49,49,25,9,17,1,20,25,13,26,48,22,35,255,45,48,40, 25,26,40,28,26,34,51,24,36,11,28,15,39,11,50,6,40,53,16,26,30,20,28,31, 13,36,7 </pot> </pulse>and it's found immediately before the second triplet. I'm starting a run now. |
Mike Send message Joined: 17 Feb 01 Posts: 34258 Credit: 79,922,639 RAC: 80 |
It`s strongly similar here with r_3584_SoG on my R9 380 win 7. setiathome_8.00_windows_intelx86.exe -verb -nog / blc25_2bit_guppi_57895_47387_HIP91358_0034.24610.818.23.46.191.vlar.wu : Result cached, skipping execution Stderr.txt : not found ------------ MB8_win_x86_SSE2_OpenCL_ATi_HD5_SoG_r3584.exe -v 1 / blc25_2bit_guppi_57895_47387_HIP91358_0034.24610.818.23.46.191.vlar.wu : AppName: MB8_win_x86_SSE2_OpenCL_ATi_HD5_SoG_r3584.exe AppArgs: -v 1 TaskName: blc25_2bit_guppi_57895_47387_HIP91358_0034.24610.818.23.46.191.vlar.wu Started at : 13:53:30.275 Ended at : 14:07:44.239 853.890 secs Elapsed 235.281 secs CPU time R2: .\ref\ref-setiathome_8.00_windows_intelx86.exe-blc25_2bit_guppi_57895_47387_HIP91358_0034.24610.818.23.46.191.vlar.wu.res Result : Strongly similar, Q= 99.95% [ stderr ] 13:53:30 (5988): Can't set up shared mem: -1. Will run in standalone mode. Verbose level set to:1 Running on device number: 0 Priority of worker thread raised successfully Priority of process adjusted successfully, below normal priority class used Illegal value for gpu_device_num: -1 in BOINC Client 0.0.0 WARNING: boinc_get_opencl_ids failed with code -33 OpenCL platform detected: Advanced Micro Devices, Inc. WARNING: BOINC supplied wrong platform! BOINC assigns device 0 0 slot of 64 used for this instance WARNING: BOINC failed to provide OpenCL device, using own enumeration abilities Info: CPU affinity mask used: 1; system mask is ffff Build features: SETI8 Non-graphics OpenCL USE_OPENCL_HD5xxx OCL_ZERO_COPY SIGNALS_ON_GPU OCL_CHIRP3 FFTW AMD specific USE_SSE2 x86 CPUID: AMD Ryzen 7 1800X Eight-Core Processor Cache: L1=64K L2=512K CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3 SSSE3 FMA3 SSE4.1 SSE4.2 AVX SSE4A OpenCL-kernels filename : MultiBeam_Kernels_r3584.cl ar=0.013392 NumCfft=113671 NumGauss=0 NumPulse=44234970240 NumTriplet=57204137376 Currently allocated 201 MB for GPU buffers In v_BaseLineSmooth: NumDataPoints=1048576, BoxCarLength=8192, NumPointsInChunk=32768 Windows optimized setiathome_v8 application Based on Intel, Core 2-optimized v8-nographics V5.13 by Alex Kan SSE2xj Win32 Build 3584 , Ported by : Raistmer, JDWhale SETI8 update by Raistmer OpenCL version by Raistmer, r3584 AMD HD5 version by Raistmer Number of OpenCL platforms: 1 OpenCL Platform Name: AMD Accelerated Parallel Processing Number of devices: 1 Max compute units: 28 Max work group size: 256 Max clock frequency: 985Mhz Max memory allocation: 3019898880 Cache type: Read/Write Cache line size: 64 Cache size: 16384 Global memory size: 3221225472 Constant buffer size: 65536 Max number of constant args: 8 Local memory type: Scratchpad Local memory size: 32768 Queue properties: Out-of-Order: No Name: Tonga Vendor: Advanced Micro Devices, Inc. Driver version: 1800.5 (VM) Version: OpenCL 1.2 AMD-APP (1800.5) Extensions: cl_khr_fp64 cl_amd_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_gl_sharing cl_ext_atomic_counters_32 cl_amd_device_attribute_query cl_amd_vec3 cl_amd_printf cl_amd_media_ops cl_amd_media_ops2 cl_amd_popcnt cl_khr_d3d10_sharing cl_khr_d3d11_sharing cl_khr_dx9_media_sharing cl_khr_image2d_from_buffer cl_khr_spir cl_khr_gl_event Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.013392 Used GPU device parameters are: Number of compute units: 28 Single buffer allocation size: 128MB Total device global memory: 3072MB max WG size: 256 local mem type: Real LotOfMem path: yes LowPerformanceGPU path: no HighPerformanceGPU path: no period_iterations_num=50 Spike: peak=24.57976, time=28.63, d_freq=2082851814.32, chirp=-9.9651, fft_len=128k Spike: peak=26.24196, time=28.63, d_freq=2082851814.31, chirp=-9.9715, fft_len=128k Spike: peak=26.43271, time=28.63, d_freq=2082851814.33, chirp=-9.974, fft_len=128k Spike: peak=24.47405, time=28.63, d_freq=2082851814.3, chirp=-9.9778, fft_len=128k Spike: peak=26.40807, time=28.63, d_freq=2082851814.32, chirp=-9.9803, fft_len=128k Pulse: peak=6.793147, time=45.86, period=16.42, d_freq=2082844438.43, score=1.05, chirp=-11.275, fft_len=1024 D: threshold 0.4664439; unscaled peak power: 0.4865191 exceeds threshold for 4.304% Triplet: peak=11.12709, time=57.39, period=9.51, d_freq=2082849725.7, chirp=-12.106, fft_len=64 Pulse: peak=8.704633, time=45.82, period=19.08, d_freq=2082851200.04, score=1.045, chirp=24.819, fft_len=128 D: threshold 0.07261781; unscaled peak power: 0.07552418 exceeds threshold for 4.002% Pulse: peak=6.338768, time=45.9, period=18.07, d_freq=2082851430.83, score=1, chirp=27.732, fft_len=2k D: threshold 0.9428869; unscaled peak power: 0.9428873 exceeds threshold for 3.793e-005% Triplet: peak=11.88211, time=42.75, period=12.03, d_freq=2082851225.44, chirp=41.314, fft_len=512 Pulse: peak=9.735943, time=45.9, period=23.53, d_freq=2082848053.15, score=1.011, chirp=57.015, fft_len=2k D: threshold 1.360308; unscaled peak power: 1.37325 exceeds threshold for 0.9514% Pulse: peak=4.374365, time=45.9, period=9.425, d_freq=2082848075.21, score=1.011, chirp=58.226, fft_len=2k D: threshold 0.6579975; unscaled peak power: 0.6641021 exceeds threshold for 0.9277% Pulse: peak=2.363577, time=45.86, period=4.049, d_freq=2082845504.88, score=1.055, chirp=69.008, fft_len=1024 D: threshold 0.2087702; unscaled peak power: 0.2167602 exceeds threshold for 3.827% Pulse: peak=5.350348, time=45.99, period=12.71, d_freq=2082844237.46, score=1.009, chirp=-71.865, fft_len=4k D: threshold 1.393561; unscaled peak power: 1.404622 exceeds threshold for 0.7938% Pulse: peak=1.285314, time=45.82, period=1.599, d_freq=2082853384.18, score=1.005, chirp=-73.852, fft_len=256 D: threshold 0.03585184; unscaled peak power: 0.03594291 exceeds threshold for 0.254% Pulse: peak=4.573022, time=45.84, period=10.38, d_freq=2082850064.25, score=1.006, chirp=74.154, fft_len=512 D: threshold 0.1696465; unscaled peak power: 0.1704526 exceeds threshold for 0.4752% Triplet: peak=11.29756, time=60.89, period=17.26, d_freq=2082852207.64, chirp=-84.445, fft_len=256 Pulse: peak=2.282089, time=45.86, period=4.267, d_freq=2082854852.39, score=1.018, chirp=-89.06, fft_len=1024 D: threshold 0.1977848; unscaled peak power: 0.2001909 exceeds threshold for 1.217% Pulse: peak=3.863604, time=45.9, period=8.143, d_freq=2082848821.45, score=1.058, chirp=98.708, fft_len=2k D: threshold 0.6301774; unscaled peak power: 0.6590711 exceeds threshold for 4.585% Best spike: peak=26.43271, time=28.63, d_freq=2082851814.33, chirp=-9.974, fft_len=128k Best autocorr: peak=16.24012, time=62.99, delay=3.3796, d_freq=2082850121.5, chirp=3.7698, fft_len=128k Best gaussian: peak=0, mean=0, ChiSq=0, time=-2.124e+011, d_freq=0, score=-12, null_hyp=0, chirp=0, fft_len=0 Best pulse: peak=3.863604, time=45.9, period=8.143, d_freq=2082848821.45, score=1.058, chirp=98.708, fft_len=2k Best triplet: peak=11.88211, time=42.75, period=12.03, d_freq=2082851225.44, chirp=41.314, fft_len=512 Spike count: 5 Autocorr count: 0 Pulse count: 11 Triplet count: 3 Gaussian count: 0 Wallclock time elapsed since last restart: 848.7 seconds Fftlength=32,pass=3:Tune: sum=61201.1(ms); min=9.72(ms); max=67.64(ms); mean=52.26(ms); s_mean=54.61; sleep=45(ms); delta=457; N=1171; usual Fftlength=32,pass=4:Tune: sum=54274.2(ms); min=8.078(ms); max=72.07(ms); mean=51.69(ms); s_mean=55.69; sleep=45(ms); delta=450; N=1050; usual Fftlength=32,pass=5:Tune: sum=20989.6(ms); min=6.439(ms); max=52.45(ms); mean=28.67(ms); s_mean=47.4; sleep=45(ms); delta=891; N=732; usual Fftlength=64,pass=3:Tune: sum=31931.1(ms); min=4.961(ms); max=69.57(ms); mean=35.13(ms); s_mean=61.68; sleep=60(ms); delta=912; N=909; usual Fftlength=64,pass=4:Tune: sum=31155.6(ms); min=3.826(ms); max=74.19(ms); mean=36.83(ms); s_mean=62.97; sleep=60(ms); delta=825; N=846; usual Fftlength=64,pass=5:Tune: sum=14965.7(ms); min=3.318(ms); max=39.69(ms); mean=18.8(ms); s_mean=27.95; sleep=30(ms); delta=907; N=796; usual Fftlength=128,pass=3:Tune: sum=27973.9(ms); min=2.559(ms); max=66.39(ms); mean=27.92(ms); s_mean=38.03; sleep=30(ms); delta=1057; N=1002; usual Fftlength=128,pass=4:Tune: sum=15554.4(ms); min=2.259(ms); max=41.35(ms); mean=17.06(ms); s_mean=21.47; sleep=15(ms); delta=1023; N=912; usual Fftlength=128,pass=5:Tune: sum=9786.61(ms); min=1.734(ms); max=27.06(ms); mean=11.22(ms); s_mean=13.23; sleep=15(ms); delta=983; N=872; usual Fftlength=256,pass=3:Tune: sum=19708.9(ms); min=1.267(ms); max=33.37(ms); mean=19.27(ms); s_mean=31.03; sleep=30(ms); delta=1106; N=1023; usual Fftlength=256,pass=4:Tune: sum=14314.4(ms); min=1.02(ms); max=24.25(ms); mean=14.99(ms); s_mean=21.83; sleep=15(ms); delta=1066; N=955; usual Fftlength=256,pass=5:Tune: sum=10781.6(ms); min=0.8739(ms); max=17.53(ms); mean=11.83(ms); s_mean=16.91; sleep=15(ms); delta=1022; N=911; usual Fftlength=512,pass=3:Tune: sum=42788.1(ms); min=0.6415(ms); max=62.84(ms); mean=28.74(ms); s_mean=35.45; sleep=30(ms); delta=1544; N=1489; high_perf Fftlength=512,pass=4:Tune: sum=1185.67(ms); min=0.5127(ms); max= 11(ms); mean=4.491(ms); s_mean=10.55; sleep=0(ms); delta=1523; N=264; usual Fftlength=512,pass=5:Tune: sum=877.612(ms); min=0.4421(ms); max=8.382(ms); mean=3.626(ms); s_mean=8.098; sleep=0(ms); delta=1501; N=242; usual Fftlength=1024,pass=3:Tune: sum=44523.4(ms); min=0.3305(ms); max=37.21(ms); mean=16.33(ms); s_mean=16.96; sleep=15(ms); delta=2754; N=2727; high_perf Fftlength=1024,pass=4:Tune: sum=283.876(ms); min=0.267(ms); max=5.483(ms); mean=2.167(ms); s_mean=4.441; sleep=0(ms); delta=2743; N=131; usual Fftlength=1024,pass=5:Tune: sum=209.26(ms); min=0.2324(ms); max=4.24(ms); mean=1.758(ms); s_mean=3.905; sleep=0(ms); delta=2731; N=119; usual Fftlength=2048,pass=3:Tune: sum=45293.9(ms); min=3.792(ms); max=53.01(ms); mean=8.567(ms); s_mean=8.493; sleep=0(ms); delta=1; N=5287; high_perf Fftlength=4096,pass=3:Tune: sum=45620.5(ms); min=1.836(ms); max=46.56(ms); mean=4.315(ms); s_mean=4.207; sleep=0(ms); delta=1; N=10573; high_perf Fftlength=8192,pass=3:Tune: sum=58914.3(ms); min=2.705(ms); max=3.485(ms); mean=2.786(ms); s_mean=2.785; sleep=0(ms); delta=1; N=21145; usual class Gaussian_transfer_not_needed: total=0, N=0, <>=0, min=0 max=0 class Gaussian_transfer_needed: total=0, N=0, <>=0, min=0 max=0 class Gaussian_skip1_no_peak: total=0, N=0, <>=0, min=0 max=0 class Gaussian_skip2_bad_group_peak: total=0, N=0, <>=0, min=0 max=0 class Gaussian_skip3_too_weak_peak: total=0, N=0, <>=0, min=0 max=0 class Gaussian_skip4_too_big_ChiSq: total=0, N=0, <>=0, min=0 max=0 class Gaussian_skip6_low_power: total=0, N=0, <>=0, min=0 max=0 class Gaussian_new_best: total=0, N=0, <>=0, min=0 max=0 class Gaussian_report: total=0, N=0, <>=0, min=0 max=0 class Gaussian_miss: total=0, N=0, <>=0, min=0 max=0 class PC_triplet_find_hit: total=53562, N=53562, <>=1, min=1 max=1 class PC_triplet_find_miss: total=1020, N=1020, <>=1, min=1 max=1 class PC_pulse_find_hit: total=42191, N=42191, <>=1, min=1 max=1 class PC_pulse_find_miss: total=18, N=18, <>=1, min=1 max=1 class PC_pulse_find_early_miss: total=13, N=13, <>=1, min=1 max=1 class PC_pulse_find_2CPU: total=0, N=0, <>=0, min=0 max=0 class PoT_transfer_not_needed: total=53550, N=53550, <>=1, min=1 max=1 class PoT_transfer_needed: total=1033, N=1033, <>=1, min=1 max=1 class SleepQuantum: total=0, N=0, <>=0, min=0 max=0 GPU device sync requested... ...GPU device synched 14:07:42 (5988): called boinc_finish(0) [ /stderr ] With each crime and every kindness we birth our future. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
It`s strongly similar here with r_3584_SoG on my R9 380 win 7.Yes, you have it: Pulse: peak=6.338768, time=45.9, period=18.07, d_freq=2082851430.83, score=1, chirp=27.732, fft_len=2k D: threshold 0.9428869; unscaled peak power: 0.9428873 exceeds threshold for 3.793e-005%Just scrapes over the threshold? |
Mike Send message Joined: 17 Feb 01 Posts: 34258 Credit: 79,922,639 RAC: 80 |
It`s strongly similar here with r_3584_SoG on my R9 380 win 7.Similar to which? His provided res file with seti800 stock. With each crime and every kindness we birth our future. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
Yes, I asked before you'd posted your edit - see my edit. My post with the signal details came from the provided seti800 stock file too. |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
Unfortunately that very app that provides additional 11th pulse is most silent one so it's impossible to track its decisions regarding that missing pulse or check thresholds of it's detection. My Apps from AKv8 don't find it. My CUDA Special Apps do find it. SSE4.1xjf OS X 64bit Build 3344 Spike count: 5 Autocorr count: 0 Pulse count: 10 Triplet count: 3 Gaussian count: 0 NV OpenCL version by Raistmer, r3709 Spike count: 5 Autocorr count: 0 Pulse count: 10 Triplet count: 3 Gaussian count: 0 setiathome v8 enhanced x41p_zi3v, Cuda 9.00 special Spike count: 5 Autocorr count: 0 Pulse count: 11 Triplet count: 3 Gaussian count: 0 |
Old man Send message Joined: 19 Sep 07 Posts: 29 Credit: 3,025,264 RAC: 0 |
Where i unpack it in linux? |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
PErhaps into separate directory or in KWSN bench. Not too familiar with KWSN bench layout under Linux. SETI apps news We're not gonna fight them. We're gonna transcend them. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Thanks for participation, I think the case IS closed. Task is above threshold by neglectible small value. So, it's general example of borderline discrepance. That's why I added threshold display in stderr for almost all apps - to relatively fast analysis. If only x86 CPU would have it too... SETI apps news We're not gonna fight them. We're gonna transcend them. |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
I ran it overnight since there was no work anyway. Both the r3711 and r3712 apps didn't find the 11th pulse. shmget in attach_shmem: Invalid argument 00:33:50 (12530): Can't set up shared mem: -1. Will run in standalone mode. Not using mb_cmdline.txt-file, using commandline options. Build features: SETI8 Non-graphics FFTW FFTOUT JSPF SSE4.1 64bit System: Linux x86_64 Kernel: 4.10.0-40-generic CPU : AMD Ryzen 7 1800X Eight-Core Processor 16 core(s), Speed : 3942.569 MHz L1 : 64 KB, Cache : 512 KB Features : FPU TSC PAE APIC MTRR MMX SSE SSE2 HT PNI SSSE3 SSE4A SSE4_1 SSE4_2 AVX AVX2 ar=0.013392 NumCfft=113671 NumGauss=0 NumPulse=44234970240 NumTriplet=57204137376 In v_BaseLineSmooth: NumDataPoints=1048576, BoxCarLength=8192, NumPointsInChunk=32768 Linux optimized setiathome_v8 application Version info: SSE4.1xjf (Intel, Core 2-optimized v8-nographics) V5.13 by Alex Kan SSE4.1xjf Linux64 Build 3711 , Ported by : Raistmer, JDWhale, Urs Echternacht Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.013392 Spike: peak=24.5798, time=28.63, d_freq=2082851814.32, chirp=-9.9651, fft_len=128k Spike: peak=26.24197, time=28.63, d_freq=2082851814.31, chirp=-9.9715, fft_len=128k Spike: peak=26.43267, time=28.63, d_freq=2082851814.33, chirp=-9.974, fft_len=128k Spike: peak=24.47411, time=28.63, d_freq=2082851814.3, chirp=-9.9778, fft_len=128k Spike: peak=26.40804, time=28.63, d_freq=2082851814.32, chirp=-9.9803, fft_len=128k Pulse: peak=6.793138, time=45.86, period=16.42, d_freq=2082844438.43, score=1.05, chirp=-11.275, fft_len=1024 D: threshold 0.4664439; unscaled peak power: 0.4865185 exceeds threshold for 4.304% Triplet: peak=11.12711, time=57.39, period=9.51, d_freq=2082849725.7, chirp=-12.106, fft_len=64 Pulse: peak=8.704662, time=45.82, period=19.08, d_freq=2082851200.04, score=1.045, chirp=24.819, fft_len=128 D: threshold 0.07261784; unscaled peak power: 0.07552444 exceeds threshold for 4.003% Triplet: peak=11.88214, time=42.75, period=12.03, d_freq=2082851225.44, chirp=41.314, fft_len=512 Pulse: peak=9.735926, time=45.9, period=23.53, d_freq=2082848053.15, score=1.01, chirp=57.015, fft_len=2k D: threshold 1.360308; unscaled peak power: 1.373248 exceeds threshold for 0.9513% Pulse: peak=4.374367, time=45.9, period=9.425, d_freq=2082848075.21, score=1.011, chirp=58.226, fft_len=2k D: threshold 0.6579973; unscaled peak power: 0.6641018 exceeds threshold for 0.9278% Pulse: peak=2.363575, time=45.86, period=4.049, d_freq=2082845504.88, score=1.055, chirp=69.008, fft_len=1024 D: threshold 0.2087702; unscaled peak power: 0.2167601 exceeds threshold for 3.827% Pulse: peak=5.350352, time=45.99, period=12.71, d_freq=2082844237.46, score=1.009, chirp=-71.865, fft_len=4k D: threshold 1.393561; unscaled peak power: 1.404623 exceeds threshold for 0.7938% Pulse: peak=1.285312, time=45.82, period=1.599, d_freq=2082853384.18, score=1.005, chirp=-73.852, fft_len=256 D: threshold 0.03585184; unscaled peak power: 0.03594288 exceeds threshold for 0.254% Pulse: peak=4.573032, time=45.84, period=10.38, d_freq=2082850064.25, score=1.006, chirp=74.154, fft_len=512 D: threshold 0.1696466; unscaled peak power: 0.1704529 exceeds threshold for 0.4753% Triplet: peak=11.29759, time=60.89, period=17.26, d_freq=2082852207.64, chirp=-84.445, fft_len=256 Pulse: peak=2.282089, time=45.86, period=4.267, d_freq=2082854852.39, score=1.018, chirp=-89.06, fft_len=1024 D: threshold 0.1977848; unscaled peak power: 0.2001909 exceeds threshold for 1.217% Pulse: peak=3.863591, time=45.9, period=8.143, d_freq=2082848821.45, score=1.058, chirp=98.708, fft_len=2k D: threshold 0.6301777; unscaled peak power: 0.6590698 exceeds threshold for 4.585% Best spike: peak=26.43267, time=28.63, d_freq=2082851814.33, chirp=-9.974, fft_len=128k Best autocorr: peak=16.24013, time=62.99, delay=3.3796, d_freq=2082850121.5, chirp=3.7698, fft_len=128k Best gaussian: peak=0, mean=0, ChiSq=0, time=-2.124e+11, d_freq=0, score=-12, null_hyp=0, chirp=0, fft_len=0 Best pulse: peak=3.863591, time=45.9, period=8.143, d_freq=2082848821.45, score=1.058, chirp=98.708, fft_len=2k Best triplet: peak=11.88214, time=42.75, period=12.03, d_freq=2082851225.44, chirp=41.314, fft_len=512 Spike count: 5 Autocorr count: 0 Pulse count: 10 Triplet count: 3 Gaussian count: 0 Time cpu in use since last restart: 2564.5 seconds 01:16:38 (12530): called boinc_finish(0) shmget in attach_shmem: Invalid argument 01:16:40 (8249): Can't set up shared mem: -1. Will run in standalone mode. Not using mb_cmdline.txt-file, using commandline options. Build features: SETI8 Non-graphics FFTW FFTOUT JSPF AVX2 64bit System: Linux x86_64 Kernel: 4.10.0-40-generic CPU : AMD Ryzen 7 1800X Eight-Core Processor 16 core(s), Speed : 3942.569 MHz L1 : 64 KB, Cache : 512 KB Features : FPU TSC PAE APIC MTRR MMX SSE SSE2 HT PNI SSSE3 SSE4A SSE4_1 SSE4_2 AVX AVX2 ar=0.013392 NumCfft=113671 NumGauss=0 NumPulse=44234970240 NumTriplet=57204137376 In v_BaseLineSmooth: NumDataPoints=1048576, BoxCarLength=8192, NumPointsInChunk=32768 Linux optimized setiathome_v8 application Version info: AVX2jf (Intel, Core 2-optimized v8-nographics) V5.13 by Alex Kan AVX2jf Linux64 Build 3712 , Ported by : Raistmer, JDWhale, Urs Echternacht Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.013392 Spike: peak=24.5798, time=28.63, d_freq=2082851814.32, chirp=-9.9651, fft_len=128k Spike: peak=26.24198, time=28.63, d_freq=2082851814.31, chirp=-9.9715, fft_len=128k Spike: peak=26.43267, time=28.63, d_freq=2082851814.33, chirp=-9.974, fft_len=128k Spike: peak=24.47411, time=28.63, d_freq=2082851814.3, chirp=-9.9778, fft_len=128k Spike: peak=26.40804, time=28.63, d_freq=2082851814.32, chirp=-9.9803, fft_len=128k Pulse: peak=6.79314, time=45.86, period=16.42, d_freq=2082844438.43, score=1.05, chirp=-11.275, fft_len=1024 D: threshold 0.4664439; unscaled peak power: 0.4865187 exceeds threshold for 4.304% Triplet: peak=11.12711, time=57.39, period=9.51, d_freq=2082849725.7, chirp=-12.106, fft_len=64 Pulse: peak=8.704659, time=45.82, period=19.08, d_freq=2082851200.04, score=1.045, chirp=24.819, fft_len=128 D: threshold 0.07261784; unscaled peak power: 0.07552443 exceeds threshold for 4.003% Triplet: peak=11.88214, time=42.75, period=12.03, d_freq=2082851225.44, chirp=41.314, fft_len=512 Pulse: peak=9.735925, time=45.9, period=23.53, d_freq=2082848053.15, score=1.01, chirp=57.015, fft_len=2k D: threshold 1.360308; unscaled peak power: 1.373248 exceeds threshold for 0.9512% Pulse: peak=4.374367, time=45.9, period=9.425, d_freq=2082848075.21, score=1.011, chirp=58.226, fft_len=2k D: threshold 0.6579973; unscaled peak power: 0.6641018 exceeds threshold for 0.9277% Pulse: peak=2.363575, time=45.86, period=4.049, d_freq=2082845504.88, score=1.055, chirp=69.008, fft_len=1024 D: threshold 0.2087702; unscaled peak power: 0.2167601 exceeds threshold for 3.827% Pulse: peak=5.35035, time=45.99, period=12.71, d_freq=2082844237.46, score=1.009, chirp=-71.865, fft_len=4k D: threshold 1.393561; unscaled peak power: 1.404623 exceeds threshold for 0.7938% Pulse: peak=1.285311, time=45.82, period=1.599, d_freq=2082853384.18, score=1.005, chirp=-73.852, fft_len=256 D: threshold 0.03585184; unscaled peak power: 0.03594288 exceeds threshold for 0.2539% Pulse: peak=4.573032, time=45.84, period=10.38, d_freq=2082850064.25, score=1.006, chirp=74.154, fft_len=512 D: threshold 0.1696465; unscaled peak power: 0.1704529 exceeds threshold for 0.4753% Triplet: peak=11.29759, time=60.89, period=17.26, d_freq=2082852207.64, chirp=-84.445, fft_len=256 Pulse: peak=2.282091, time=45.86, period=4.267, d_freq=2082854852.39, score=1.018, chirp=-89.06, fft_len=1024 D: threshold 0.1977847; unscaled peak power: 0.2001909 exceeds threshold for 1.217% Pulse: peak=3.863591, time=45.9, period=8.143, d_freq=2082848821.45, score=1.058, chirp=98.708, fft_len=2k D: threshold 0.6301777; unscaled peak power: 0.6590698 exceeds threshold for 4.585% Best spike: peak=26.43267, time=28.63, d_freq=2082851814.33, chirp=-9.974, fft_len=128k Best autocorr: peak=16.24012, time=62.99, delay=3.3796, d_freq=2082850121.5, chirp=3.7698, fft_len=128k Best gaussian: peak=0, mean=0, ChiSq=0, time=-2.124e+11, d_freq=0, score=-12, null_hyp=0, chirp=0, fft_len=0 Best pulse: peak=3.863591, time=45.9, period=8.143, d_freq=2082848821.45, score=1.058, chirp=98.708, fft_len=2k Best triplet: peak=11.88214, time=42.75, period=12.03, d_freq=2082851225.44, chirp=41.314, fft_len=512 Spike count: 5 Autocorr count: 0 Pulse count: 10 Triplet count: 3 Gaussian count: 0 Time cpu in use since last restart: 2578.9 seconds 02:00:19 (8249): called boinc_finish(0) Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
Task is above threshold by neglectible small value. So, it's general example of borderline discrepance.So, is that what leads to the score value being calculated as exactly "1", and is that the determining factor for reporting the signal? That's what I was curious about when the task originally showed up in my Inconclusive list. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Well, score for Pulse signal calculated as: pi.score=snr/thresh; where snr (signal to noise ratio) is: float _snr = (tmp_max-avg)/(cur_thresh-avg); So yes, value of 1 or smth very close to 1 is good sign of very small exceed of threshold. Thanks for bringing that to attention, that could help in determination of such cases in the future. Answerring on another question: here is signal report statement from OpenCL PulseFind kernel: (tmp_max>cur_thresh) If true - signal reported (for re-check on CPU). So, little different from score analysis. On CPU signal report statement for best update looks like: float snr_tmp=(float)((tmp_max-avg)*sqrt((float)num_adds)*ravg); float thresh=(float)((cur_thresh-avg)*sqrt((float)num_adds)*ravg); if(snr_tmp/thresh>best_pulse->score){ and report pulse itself: if (tmp_max>cpy_thresh){.... So, score doesn't directly used for determination if Pulse will be reported or not (but score directly used for best Pulse update selection) SETI apps news We're not gonna fight them. We're gonna transcend them. |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
EDIT: Hmmm, I guess I should have refreshed this page before I hit the Reply button. I didn't see your additional edit. So, perhaps if score isn't used directly, it throws my whole theory out the window. Might still be interesting to see what the raw, unrounded score was on that questionable Pulse. EDIT2: What is the difference between "cur_thresh" and "cpy_thresh", and how does either relate to the "<thresh>" value that shows up in the result file? |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
I looked at a couple of other Inconclusives this afternoon that are similar to the Test Case. In both of these, the Pulse reported by the Special App but apparently "missed" by the other app had a score of exactly '1'. (The Linux CPU Stderr doesn't show the signals, but I ran both of these offline with the Windows CPU app and confirmed that it, too, "missed" that specific Pulse.) Workunit 2761018471 (blc03_2bit_guppi_57903_46950_HIP9480_0002.17257.0.24.47.19.vlar) Task 6197193280 (S=1, A=1, P=10, T=2, G=0, BS=24.23479, BG=0) x41p_zi3v, Cuda 8.00 special Task 6197193281 (S=1, A=1, P=9, T=2, G=0, BS=?, BG=?) v8.00 x86_64-pc-linux-gnu Pulse: peak=4.227002, time=45.99, period=9.783, d_freq=2523267853, score=1, chirp=-45.009, fft_len=4k Workunit 2761092944 (blc25_2bit_guppi_57895_52725_HIP171_0043.28725.409.24.47.77.vlar) Task 6197347694 (S=0, A=0, P=15, T=2, G=0, BS=23.96479, BG=0) v8.08 (alt) windows_x86_64 Task 6197347695 (S=0, A=0, P=16, T=2, G=0, BS=23.96478, BG=0) x41p_zi3v, Cuda 8.00 special Pulse: peak=4.237732, time=45.99, period=10.44, d_freq=2052248482.07, score=1, chirp=-35.004, fft_len=4k This last WU also had another Pulse with "score=1" that was reported by both apps. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Such borderline effects could come also from >= usage instead of > . Regarding thresholds- there are few of them with different computation complexity degree. So first candidate checked with easiest one and if it passed the the test next one applied. thresh_scale = t_funct(di,num_adds,di+tabofst); dis_thresh = avg*(thresh_scale*pulse_display_thresh + (1 - pulse_display_thresh)); float lowest_thresh = t_funct(baildi, num_adds << (ndivs - 1), baildi + bailtab); float bail_scale = best_pulse->score; if (bail_scale > 1.0f) bail_scale = 1.0f; else if (bail_scale < pulse_display_thresh) bail_scale = pulse_display_thresh; float bail_thresh = avg*(lowest_thresh*bail_scale + (1 - bail_scale)); if (tmp_max<bail_thresh) { continue; } if (tmp_max>dis_thresh) { cur_thresh = thresh_scale * avg; cpy_thresh = cur_thresh + maxd; float snr_tmp=(float)((tmp_max-avg)*sqrt((float)num_adds)*ravg); float thresh=(float)((cur_thresh-avg)*sqrt((float)num_adds)*ravg); if(snr_tmp/thresh>best_pulse->score){ ReportPulseEvent(tmp_max*ravg,avg,cperiod, TOffset+(int)(PulsePotLen/2),FOffset,snr_tmp,thresh, div, max_scale, 0); if(verbose>=2 && verbose<6)fprintf(stderr,"B:\tthreshold %.7g; unscaled peak power: %.7g exceeds threshold for %.4g%%\n", cur_thresh,tmp_max,(tmp_max-cur_thresh)/cur_thresh*100.f); } if (tmp_max>cpy_thresh) { maxp = cperiod; maxd = tmp_max - cur_thresh; max = tmp_max; snr = snr_tmp;//(tmp_max-avg)*(float)sqrt((float)num_adds_2)*ravg; fthresh = thresh;//(cur_thresh-avg)*(float)sqrt((float)num_adds_2)*ravg; mmax_scale = max_scale; memcpy(FoldedPOT, div+stoffset, di*sizeof(float)); max_cur_thresh=cur_thresh;//R: to print it if needed } if (maxp!=0){ ReportPulseEvent(max*ravg,avg,maxp,TOffset+PulsePotLen/2,FOffset, snr, fthresh, FoldedPOT, mmax_scale, 1); if(verbose >=1) fprintf(stderr,"D:\tthreshold %.7g; unscaled peak power: %.7g exceeds threshold for %.4g%%\n", max_cur_thresh,max,(max-max_cur_thresh)/max_cur_thresh*100.f); } ReportPulseEvent(...,0) is for best pulse update ReportPulseEvent(...,1) is for reportable pulse write. int ReportPulseEvent(float PulsePower,float MeanPower, float period, int time_bin,int freq_bin, float snr, float thresh, float *folded_pot, float max_scale, int write_pulse) { PULSE_INFO pi; pulse pulse; // pulse info pi.score=snr/thresh; pi.p.peak_power=PulsePower-1; pi.p.mean_power=MeanPower; pi.p.fft_len=ChirpFftPairs[analysis_state.icfft].FftLen; pi.p.chirp_rate=ChirpFftPairs[analysis_state.icfft].ChirpRate; pi.p.period=static_cast<float>(period*static_cast<double>(pi.p.fft_len)/swi.subband_sample_rate); pi.p.snr = snr; pi.p.thresh = thresh; pi.p.len_prof = len_prof; pi.freq_bin=freq_bin; pi.time_bin=time_bin; pi.p.freq=cnvt_bin_hz(freq_bin, pi.p.fft_len); double t_offset=(static_cast<double>(time_bin)+0.5) *static_cast<double>(pi.p.fft_len)/ swi.subband_sample_rate; pi.p.detection_freq=calc_detection_freq(pi.p.freq,pi.p.chirp_rate,t_offset); pi.p.time=swi.time_recorded+t_offset/86400.0; time_to_ra_dec(pi.p.time, &pi.p.ra, &pi.p.decl); That's main stages of pulse checking. SETI apps news We're not gonna fight them. We're gonna transcend them. |
rob smith Send message Joined: 7 Mar 03 Posts: 22200 Credit: 416,307,556 RAC: 380 |
Such borderline effects could come also from >= usage instead of > That has certainly been the case in a number of real-time applications, and often the "error" is very small, but can be very significant. Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
rob smith Send message Joined: 7 Mar 03 Posts: 22200 Credit: 416,307,556 RAC: 380 |
Another one is type casting - such as in this line. I know it is done for speed these days, but we used to have to do it to save memory and clock cycles. Operations such as this may be (not "are") dependent upon the exact sequence of the bit processing during the cast, thus may return different answers for different processor families or compilers pi.p.period=static_cast<float>(period*static_cast<double>(pi.p.fft_len) Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
Such borderline effects could come also from >= usage instead of > .Okay, I give up trying to understand that code block. ;^) Too many unknown variables coming from elsewhere and nesting/indentations that give me a headache. And perhaps the fact that the only time I ever really tried to learn C++ was about 25 years ago with Turbo C++ 3.0 in DOS. Gave me a headache back then, too. (I still have the 772-page User Guide, though, copyrighted 1992.) Anyway, it doesn't look like my theory about reporting decisions being made using rounded versus unrounded scores holds any water. I had overlooked Richard's earlier post with the reported details for the 11th Pulse, showing: <snr>14.173917770386</snr> <thresh>14.173913955688</thresh> That should give an unrounded score of 1.000000269135118, so it wasn't a score that was less than 1 and then rounded up. Interestingly, I ran that WU with stock Cuda50 today and it also managed to report that Pulse, but with slightly different values: <snr>14.17388343811</snr> <thresh>14.173751831055</thresh> Theoretically, that would give an unrounded score of 1.000009285265932, so it certainly still seems plausible that somewhere in that code labyrinth, a very tiny rounding or type conversion operation could tweak a value just enough to drop the score below 1 (or at least drop "snr" below "thresh", so that it never even gets around to calculating a score). EDIT: If, for instance, an app calculated the "snr" value as Cuda50 did, but came up with a "thresh" value such as in Richard's example, I assume that it wouldn't get reported. |
©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.