Message boards :
Number crunching :
Monitoring inconclusive GBT validations and harvesting data for testing
Message board moderation
Previous · 1 . . . 25 · 26 · 27 · 28 · 29 · 30 · 31 . . . 36 · Next
Author | Message |
---|---|
Kiska Send message Joined: 31 Mar 12 Posts: 302 Credit: 3,067,762 RAC: 0 |
Someone should been able to grab this 1, Workunit 2301590234, as it's still in progress. I have also saved this file to Google Drive https://drive.google.com/file/d/0B0-3oeXJF8g0Y3lVYnJ0ZU9WS0U/view?usp=sharing Unfortunately I did not get the other file |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
I've also got a new quadruple in my own list this evening. Since the 5th host that got this WU ran it as stock Windows CPU, it shouldn't come as a surprise that its results matched the 2nd host and the WU is now validated. However, only 4 of the 5 hosts got credit. The Mac result was declared Invalid, which seems to be fairly common for that host. Now, since a new quadruple Inconclusive for a late stage overflow seems to be almost a daily occurrence, he's my new one for this evening, 3 SoG r3528 machines and one stock Windows app, all with different counts. The 5th host will again be a stock Windows app, so it should match with the 3rd host. Workunit 2301105316 (08ja09ab.7925.16428.8.35.187) Task 5233243565 (S=16, A=7, P=6, T=1, G=0) v8.19 (opencl_nvidia_SoG) windows_intelx86 Task 5233243566 (S=17, A=7, P=5, T=1, G=0) v8.19 (opencl_nvidia_SoG) windows_intelx86 Task 5234773670 (S=21, A=7, P=2, T=0, G=0) v8.00 windows_intelx86 Task 5236410759 (S=20, A=7, P=3, T=0, G=0) SSE3xj Win32 Build 3528 |
Kiska Send message Joined: 31 Mar 12 Posts: 302 Credit: 3,067,762 RAC: 0 |
Grabbed the unit in question Saved to Google Drive: https://drive.google.com/file/d/0B0-3oeXJF8g0ejF0VEtRV19XTDA/view?usp=sharing |
petri33 Send message Joined: 6 Jun 02 Posts: 1668 Credit: 623,086,772 RAC: 156 |
Grabbed the unit in question Hi, I ran this WU with my computer with the linux CPU executable and my current development version ... Current WU: 08ja09ab.7925.16428.8.35.187.wu ---------------------------------------------------------------- Running default app with command :... setiathome_8.04_i686-pc-linux-gnu Elapsed Time: ....................... 2444 seconds ---------------------------------------------------------------- Running app with command : .......... axo -bs -pfb 8 -pfp 120 -unroll 20 --device 1 gCudaDevProps.multiProcessorCount = 20 Work data buffer for fft results size = 320864256 MallocHost G=33554432 T=33554432 P=16777216 (16) MallocHost tmp_PoTP=16777216 MallocHost tmp_PoTP2=16777216 MallocHost tmp_PoTT=16777216 MallocHost tmp_PoTG=12582912 MallocHost best_PoTP=16777216 MallocHost bestPoTG=12582912 Allocing tmp data buf for unroll 20 MallocHost tmp_smallPoT=524288 MallocHost PowerSpectrumSumMax=3145728 GPSF 3.035655 3 5.352018 AcIn 16779264 AcOut 33558528 Mallocing blockSums 24576 bytes Elapsed Time : ...................... 115 seconds Speed compared to default : ......... 2125 % ----------------- Comparing results Result : Strongly similar, Q= 99.93% ---------------------------------------------------------------- Done with 08ja09ab.7925.16428.8.35.187.wu ... and the rescmpv5_l says with Q100 option ... root@Linux1:~/KWSN-Bench-Linux-MBv7_v2.01.08# ./rescmpv5_l testData/ref-result.setiathome_8.04_i686-pc-linux-gnu.08ja09ab.7925.16428.8.35.187.wu.sah testData/result.axo.08ja09ab.7925.16428.8.35.187.wu.sah Q100 ------------- R1:R2 ------------ ------------- R2:R1 ------------ Exact Super Tight Good Bad Exact Super Tight Good Bad Spike 0 21 21 21 0 0 21 21 21 0 Autocorr 0 7 7 7 0 0 7 7 7 0 Gaussian 0 0 0 0 0 0 0 0 0 0 Pulse 0 2 2 2 0 0 2 2 2 0 Triplet 0 0 0 0 0 0 0 0 0 0 Best Spike 0 0 0 0 0 0 0 0 0 0 Best Autocorr 0 0 0 0 0 0 0 0 0 0 Best Gaussian 0 0 0 0 0 0 0 0 0 0 Best Pulse 0 0 0 0 0 0 0 0 0 0 Best Triplet 0 0 0 0 0 0 0 0 0 0 ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- 0 30 30 30 0 0 30 30 30 0 Result : Strongly similar, Q= 99.93% To overcome Heisenbergs: "You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
Well, I don't know if there's anything further that can be learned from these, but...another day, another quadruple Inconclusive, so I'll go ahead and post it. This one appears to be more of the "Instant" overflow variety. Workunit 2303805767 (20jl16aa.1361.42541.16.43.252) Task 5239013701 (S=28, A=2, P=0, T=0, G=0) SSE3xj Win32 Build 3528 Task 5239013702 (S=29, A=1, P=0, T=0, G=0) v8.00 windows_intelx86 Task 5241134325 (S=30, A=0, P=0, T=0, G=0) v8.00 (cuda42) windows_intelx86 Task 5241657014 (S=30, A=0, P=0, T=0, G=0) x41p_zi3i, Cuda 7.50 special Tbar, the last one appears to be one of yours. The good news, I suppose, is that it only took about 5 seconds to join the party. :^) EDIT: Looks like this WU file is gone already, even though at the time of this posting the WU page shows that it's still in progress on the 5th host. I guess the Replica DB hasn't quite caught up from the outage yet. Sorry about that. |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
An Instant Overflow. Nothing strange here. This has been a problem for as long as I remember. It wasn't unusual to see these go through a full 10 Hosts and still not validate. Only taking 5 is an improvement. The Special App definitely doesn't care for them, look through the Top Linux Hosts and you will see them Full of Instant Overflow Inconclusives. My Mac is actually on the low side. I'd try a newer build, but zi3i is the last version I have. |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13736 Credit: 208,696,464 RAC: 304 |
It is frustrating when you get a noisy WU, and you don't agree with your wingmen what goes where. My Invalid Spike count: 7 Autocorr count: 2 Pulse count: 6 Triplet count: 16 Gaussian count: 0 Them Spike count: 18 Autocorr count: 11 Pulse count: 1 Triplet count: 0 Gaussian count: 0 Spike count: 18 Autocorr count: 11 Pulse count: 1 Triplet count: 0 Gaussian count: 0 blc3_2bit_guppi_57432_27585_HIP57494_OFF_0010.31802.0.17.26.8.vlar Grant Darwin NT |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
blc3_2bit_guppi_57432_27585_HIP57494_OFF_0010.31802.0.17.26.8.vlar It's too late to catch the datafile for this one, but the two 18-11 results were from: Stock v8.03 x86_64-apple-darwin Stock v8.00 x86_64-pc-linux-gnu It's about time we swung the microscope over to them. |
Kiska Send message Joined: 31 Mar 12 Posts: 302 Credit: 3,067,762 RAC: 0 |
blc3_2bit_guppi_57432_27585_HIP57494_OFF_0010.31802.0.17.26.8.vlar Would setting up a hackintosh be a 'good' enough solution? Since I am sure we don't want to spend money on a mac for debugging |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
blc3_2bit_guppi_57432_27585_HIP57494_OFF_0010.31802.0.17.26.8.vlar From the list I generated for my hosts last evening: v8.03 x86_64-apple-darwin Workunit 2297635785 (19ja09ab.20199.22976.7.34.9) Task 5225887776 (S=24, A=6, P=0, T=0, G=0) v8.03 x86_64-apple-darwin Task 5225887777 (S=22, A=8, P=0, T=0, G=0) SSE3xj Win32 Build 3528 Workunit 2298286244 (blc3_2bit_guppi_57451_20612_HIP62472_0007.24494.416.18.27.181.vlar) Task 5227261751 (S=19, A=1, P=10, T=0, G=0) SSE3xj Win32 Build 3528 Task 5227261752 (S=20, A=1, P=9, T=0, G=0) SSE3xj Win32 Build 3500 Task 5229822975 (S=21, A=1, P=8, T=0, G=0) v8.03 x86_64-apple-darwin Workunit 2299858620 (blc3_2bit_guppi_57451_22465_HIP63121_OFF_0012.23004.0.18.27.235.vlar) Task 5230589212 (S=24, A=1, P=5, T=0, G=0) SSE3xj Win32 Build 3528 Task 5231099938 (S=25, A=1, P=4, T=0, G=0) v8.03 x86_64-apple-darwin Workunit 2306647376 (10jl16ab.6877.93530.12.39.148) Task 5245051755 (S=0, A=1, P=0, T=3, G=1) v8.19 (opencl_intel_gpu_sah) windows_intelx86 Task 5245051756 (S=0, A=1, P=0, T=13, G=0) v8.03 x86_64-apple-darwin v8.00 x86_64-pc-linux-gnu Workunit 2293324473 (15mr09ad.13715.17659.14.41.108) Task 5216731859 (S=8, A=0, P=0, T=0, G=3) v8.19 (opencl_intel_gpu_sah) windows_intelx86 Task 5216731860 (S=8, A=0, P=0, T=0, G=2) v8.00 x86_64-pc-linux-gnu Workunit 2298108679 (24dc09aa.2403.24198.5.32.10) Task 5226886598 (S=23, A=7, P=0, T=0, G=0) v8.00 x86_64-pc-linux-gnu Task 5226886599 (S=22, A=8, P=0, T=0, G=0) SSE3xj Win32 Build 3528 Workunit 2300494353 (19jl16aa.19093.174068.12.39.135) Task 5231941244 (S=2, A=0, P=3, T=0, G=1) v8.00 x86_64-pc-linux-gnu Task 5231941245 (S=2, A=0, P=2, T=0, G=1) v8.00 (opencl_intel_gpu_sah) x86_64-apple-darwin |
Jeff Buck Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0 |
I refrained from posting this one when it was just a quadruple, a couple days ago, but now that it's become a quintuple Inconclusive, I think it merits some publicity. Workunit 2304111745 (blc3_2bit_guppi_57432_27257_HIP57494_0009.15810.831.18.27.104.vlar) Task 5239674006 (S=20, A=0, P=10, T=0, G=0) SSE3xj Win32 Build 3528 Task 5239674007 (S=18, A=0, P=12, T=0, G=0) v8.19 (opencl_nvidia_SoG) windows_intelx86 Task 5242168321 (S=20, A=0, P=10, T=0, G=0) v8.19 (opencl_nvidia_SoG) windows_intelx86 Task 5244395277 (S=19, A=0, P=11, T=0, G=0) v8.19 (opencl_nvidia_SoG) windows_intelx86 Task 5245993619 (S=21, A=0, P=9, T=0, G=0) SSE2xj Win32 Build 3500 The sixth host looks like it will be running stock Windows CPU. |
Kiska Send message Joined: 31 Mar 12 Posts: 302 Credit: 3,067,762 RAC: 0 |
I refrained from posting this one when it was just a quadruple, a couple days ago, but now that it's become a quintuple Inconclusive, I think it merits some publicity. I have downloaded and uploaded this file to my Google Drive. I will also try to run it using to apps, that may give inconclusive: https://drive.google.com/file/d/0B0-3oeXJF8g0WW1qQTlaaFE2dzA/view?usp=sharing |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
I refrained from posting this one when it was just a quadruple, a couple days ago, but now that it's become a quintuple Inconclusive, I think it merits some publicity. The interesting thing about that WU is that the first four tasks all used exactly the same application - NV SoG r3528. Jeff has his anonymous platform deployment tuned up with a command line, but the other three are all operating in default mode. And there's not a great deal of difference in the GPUs either - GTXs 960, 980, 760, 760 Ti OEM respectively. I'm getting a hint of overclocking from the 980 (Max clock 1367Mhz feels a bit toasty, and it has a previous invalid) - but why can't the other three come up with a matched pair between them? |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
blc3_2bit_guppi_57432_27585_HIP57494_OFF_0010.31802.0.17.26.8.vlar By itself, that wouldn't help us much. Judging by the deployment dates (7 Jan 2016 and 30 Dec 2015 respectively), both those apps were developed in the labs at Berkeley: it's really a question of persuading whoever has 'ownership' of those two builds (I suspect our Eric K did the codebase and probably the Linux build: Charlie Fenton probably did the Mac build from Eric's code) to revisit their past and review their work. Unless somebody here has the skill to replicate their work under laboratory conditions (i.e. using the same OS version, build tools and settings), and work out where they went wrong, I don't think that simply running more tests is going to move us forward. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
I described why same SoG revision will give different set of signals on overflow even on similar (not speaking about different) hardware earlier. When/if SoG transformation will be completed this effect will be eliminated. 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 |
I described why same SoG revision will give different set of signals on overflow even on similar (not speaking about different) hardware earlier. When/if SoG transformation will be completed this effect will be eliminated. Ah, yes - I think I have a vague memory of that. Wasn't it something like 'this is a SSoG build' - "Some Signals on GPU"? Any ideas on the timeline for a fully-finished build? |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
"After SoG TwinChirp" - that's all I could say about timeline. SETI apps news We're not gonna fight them. We're gonna transcend them. |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
"After SoG TwinChirp" - that's all I could say about timeline. Any idea on an updated OSX build? I decided to run the latest System updates and that caused me to have to install new nVidia graphics drivers. So, while I had the GTS 250 in the machine I booted to Mavericks and compiled the Apps there. Same results as with the Apps compiled in El Capitan, the Intel build is about normal whereas the NV & NV-SoG builds took much longer to finish and crashed frequently with the BLC tasks. I still haven't heard a word on how the Intel Build works on a real iGPU. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
"After SoG TwinChirp" - that's all I could say about timeline. All I could do is to make NV path equal to iGPU one for OS X. W/o target hardware hard to do any serious debugging. SETI apps news We're not gonna fight them. We're gonna transcend them. |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
All I could do is to make NV path equal to iGPU one for OS X. Well, it looks as though the Intel Path is the only one that works on my machine. Even the App using just "Build features: SETI8 Non-graphics OpenCL OCL_CHIRP3 ASYNC_SPIKE FFTW SSSE3 64bit" crashes on both types of tasks. The BLC6 got most of the way through and the reference_work_unit_r3215 crashed about half way with the Same Error. Triplet: peak=8.657347, time=80.39, period=0.5439, d_freq=1418924863.12, chirp=0.91687, fft_len=128 Google says; Result: ruby(28425,0x7fff70bc9ca0) malloc: *** error for object 0x100400000: pointer being freed was not allocated *** set a breakpoint in malloc_error_break to debug https://bugs.ruby-lang.org/issues/4389 This issue was solved with changeset r30840 The only one that doesn't crash is the Intel build. |
©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.