Monitoring inconclusive GBT validations and harvesting data for testing

Message boards : Number crunching : Monitoring inconclusive GBT validations and harvesting data for testing
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 36 · Next

AuthorMessage
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1812117 - Posted: 24 Aug 2016, 10:23:15 UTC - in response to Message 1812116.  
Last modified: 24 Aug 2016, 10:24:14 UTC

More inconclusives whether they are GBT or Arecibo

blc5_2bit_guppi_57451_69044_HIP117559_OFF_0022.7520.416.18.27.221.vlar
Datafile

29au10ab.13767.17658.13.40.192
Datafile


running the first guppi [you supplied before] against reference Win32 stock CPU, and Cuda50, now; then will manually compare to your CPU result. Could turn out yours is ok or not, we'll see.
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1812117 · Report as offensive
Kiska
Volunteer tester

Send message
Joined: 31 Mar 12
Posts: 302
Credit: 3,067,762
RAC: 0
Australia
Message 1812118 - Posted: 24 Aug 2016, 10:26:07 UTC - in response to Message 1812117.  
Last modified: 24 Aug 2016, 10:28:23 UTC

I am also running CUDA as well, I'll do 4.2 then 5.0

Though it shouldn't have issues, I ran it stock
ID: 1812118 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1812120 - Posted: 24 Aug 2016, 10:30:09 UTC - in response to Message 1812118.  
Last modified: 24 Aug 2016, 10:31:29 UTC

I am also running CUDA as well, I'll do 4.2 then 5.0

Though it shouldn't have issues, I ran it stock


Looking at the other hosts in that first one, more than likely it could be one Petri will want, so will confirm and then email it to him. The first host was a broken Cuda, second was your CPU, third was Petri special.

(If my CPU + Cuda match it under bench, + yours manually, he'll likely find some useful hints in it)
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1812120 · Report as offensive
Kiska
Volunteer tester

Send message
Joined: 31 Mar 12
Posts: 302
Credit: 3,067,762
RAC: 0
Australia
Message 1812121 - Posted: 24 Aug 2016, 10:34:06 UTC - in response to Message 1812120.  
Last modified: 24 Aug 2016, 10:34:55 UTC

I can run petri special app, manually I just need to know where to download it

My manually ran one, according to stderr has the same counts of pulses, etc
ID: 1812121 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1812123 - Posted: 24 Aug 2016, 10:42:52 UTC - in response to Message 1812121.  

I can run petri special app, manually I just need to know where to download it

My manually ran one, according to stderr has the same counts of pulses, etc


For Linux, You'd need to ask Petri (if he's still working on it toward these validation issues he may say no). For Windows, my build's somewhat broken and needs updating, so not an option to go handing out. TBar has a Mac version though.
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1812123 · Report as offensive
Kiska
Volunteer tester

Send message
Joined: 31 Mar 12
Posts: 302
Credit: 3,067,762
RAC: 0
Australia
Message 1812127 - Posted: 24 Aug 2016, 11:05:15 UTC - in response to Message 1812123.  

Ok. well anyway I expect my result file and your cpu + cuda to be ~99% similar or even matching. Its too bad I can't get the Windows build.....
ID: 1812127 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1812135 - Posted: 24 Aug 2016, 11:18:32 UTC - in response to Message 1812127.  
Last modified: 24 Aug 2016, 11:20:40 UTC

Ok. well anyway I expect my result file and your cpu + cuda to be ~99% similar or even matching. Its too bad I can't get the Windows build.....


I think that'll be the case. Sadly takes 2 hours on crappy CPU here :D (not so long for the Cuda result)

[Edit:] Just got Petri's sources updated in svn... one small step at a time
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1812135 · Report as offensive
Kiska
Volunteer tester

Send message
Joined: 31 Mar 12
Posts: 302
Credit: 3,067,762
RAC: 0
Australia
Message 1812137 - Posted: 24 Aug 2016, 11:21:06 UTC - in response to Message 1812135.  

No worries, it took 2 hrs on my E3-1230 v3.
ID: 1812137 · Report as offensive
Kiska
Volunteer tester

Send message
Joined: 31 Mar 12
Posts: 302
Credit: 3,067,762
RAC: 0
Australia
Message 1812146 - Posted: 24 Aug 2016, 11:54:06 UTC

ID: 1812146 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1812154 - Posted: 24 Aug 2016, 12:06:48 UTC - in response to Message 1812146.  

got it. Will probably pass out while CPU is still running, but will have multiple ones to compare there. Will probably have enough on that one by the morning to work out if it might be useful to petri, then package everything up to email him.
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1812154 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1812194 - Posted: 24 Aug 2016, 15:09:42 UTC
Last modified: 24 Aug 2016, 15:10:07 UTC

Well Kiska's original inconclusive against Petri's older build (on WezH's Linux Machine) confirmed a large amount we knew, and possibly holds some hints for development to sift through, so will bundle up and send to Petri.

Note that my Windows zkpr3 test build, which I consider broken, and is already outdated, probably reflects similar characteristics to the Linux one WezH's running, and precedes the last changes in alpha. Probably will get to attempting a revised test build after having a Wisdom tooth pulled later today (we'll see)

C:\Users\Jason\Downloads\[alpha]\kiska\Comparison>rescmpv4.exe cuda42_result_fro
m_kiska.res ref-setiathome_8.00_windows_intelx86.exe-kiska_guppi.vlar.wu.res
Result      : Strongly similar,  Q= 99.22%

C:\Users\Jason\Downloads\[alpha]\kiska\Comparison>rescmpv4.exe jason-result-Luna
tics_x41zj_win32_cuda50.exe-kiska_guppi.vlar.wu.res ref-setiathome_8.00_windows_
intelx86.exe-kiska_guppi.vlar.wu.res
Result      : Strongly similar,  Q= 99.22%

C:\Users\Jason\Downloads\[alpha]\kiska\Comparison>rescmpv4.exe jason-result-Luna
tics_x41zj_win32_cuda50.exe-kiska_guppi.vlar.wu.res cuda42_result_from_kiska.res

Result      : Strongly similar,  Q= 100.0%

C:\Users\Jason\Downloads\[alpha]\kiska\Comparison>rescmpv4.exe result-Lunatics_x
41zkpr3_winx64_cuda65.exe-kiska_guppi.vlar.wu.res ref-setiathome_8.00_windows_in
telx86.exe-kiska_guppi.vlar.wu.res
                ----- R1:R2 ------     ----- R2:R1 ------
                Tight  Good    Bad     Tight  Good    Bad
        Spike      0      0      0        0      0      0
     Autocorr      1      1      0        1      1      0
     Gaussian      0      0      0        0      0      0
        Pulse      3      3      0        3      3      2
      Triplet      4      4      0        4      4      0
   Best Spike      1      1      0        1      1      0
Best Autocorr      1      1      0        1      1      0
Best Gaussian      1      1      0        1      1      0
   Best Pulse      0      0      1        0      0      1
 Best Triplet      1      1      0        1      1      0
                ----   ----   ----     ----   ----   ----
                  12     12      1       12     12      3

Unmatched signal(s) in R1 at line(s) 554
Unmatched signal(s) in R2 at line(s) 383 471 606
Result      : Weakly similar.

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1812194 · Report as offensive
Profile Jeff Buck Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1812254 - Posted: 24 Aug 2016, 18:59:31 UTC

Started out as a gray and dismal morning here, so I decided to see if I could use the current Inconclusives in my own local DB to programmatically generate some sort of formatted list that could possibly be useful to those of you doing the research and testing in this area. Here's a sampling of WUs from my initial stab at it:

Workunit 2213859925
Task 5047460794 (S=0, A=0, P=14, T=2, G=0) Anonymous platform (NVIDIA GPU)
Task 5047460795 (S=0, A=0, P=14, T=2, G=0) v8.12 (opencl_intel_gpu_sah) windows_intelx86

Workunit 2237343062
Task 5097864317 (S=0, A=1, P=4, T=0, G=0) Anonymous platform (NVIDIA GPU)
Task 5097864318 (S=0, A=1, P=3, T=0, G=0) v8.00 (opencl_intel_gpu_sah) x86_64-apple-darwin
Task 5102639717 (S=3, A=1, P=4, T=0, G=0) v8.12 (opencl_ati5_SoG_cat132) windows_intelx86

Workunit 2236974648
Task 5097074623 (S=4, A=0, P=8, T=1, G=0) Anonymous platform (NVIDIA GPU)
Task 5111493046 (S=4, A=0, P=10, T=1, G=0) v8.12 (opencl_intel_gpu_sah) windows_intelx86
Task 5113043247 (S=4, A=0, P=5, T=1, G=0) v8.00 (opencl_intel_gpu_sah) x86_64-apple-darwin

The full list (in html format, with WU and Task links) is available in a cloud file, if anyone wants to take a further look.

These are all just WUs that one of my own hosts is involved in, either already with an Inconclusive result, or in my queue waiting to run as a tiebreaker. I notice that there are already a few that have been resolved, just in the few hours since my DB was updated.

Let me know if this is at all useful. Meantime, the sun is now out (albeit through a high smoky haze) and it's lunchtime.
ID: 1812254 · Report as offensive
Kiska
Volunteer tester

Send message
Joined: 31 Mar 12
Posts: 302
Credit: 3,067,762
RAC: 0
Australia
Message 1812329 - Posted: 25 Aug 2016, 0:08:07 UTC - in response to Message 1812254.  
Last modified: 25 Aug 2016, 0:39:55 UTC

I guess thats a start, but however we would prefer the datafile so we could get some sample results match and see how similar they are.

Example:


Workunit 2236974648
Task 5097074623 (S=4, A=0, P=8, T=1, G=0) Anonymous platform (NVIDIA GPU)
Task 5111493046 (S=4, A=0, P=10, T=1, G=0) v8.12 (opencl_intel_gpu_sah) windows_intelx86
Task 5113043247 (S=4, A=0, P=5, T=1, G=0) v8.00 (opencl_intel_gpu_sah) x86_64-apple-darwin


Workunit

Datafile

Workunit 2237343062
Task 5097864317 (S=0, A=1, P=4, T=0, G=0) Anonymous platform (NVIDIA GPU)
Task 5097864318 (S=0, A=1, P=3, T=0, G=0) v8.00 (opencl_intel_gpu_sah) x86_64-apple-darwin
Task 5102639717 (S=3, A=1, P=4, T=0, G=0) v8.12 (opencl_ati5_SoG_cat132) windows_intelx86


Workunit

Datafile

Workunit 2213859925
Task 5047460794 (S=0, A=0, P=14, T=2, G=0) Anonymous platform (NVIDIA GPU)
Task 5047460795 (S=0, A=0, P=14, T=2, G=0) v8.12 (opencl_intel_gpu_sah) windows_intelx86


Workunit

Datafile
ID: 1812329 · Report as offensive
Profile Jeff Buck Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1812335 - Posted: 25 Aug 2016, 1:00:53 UTC - in response to Message 1812329.  

I guess thats a start, but however we would prefer the datafile so we could get some sample results match and see how similar they are.

Sure, but there are over 100 Workunits in that list and it's likely that only a very few of them might be useful for this testing. That's why I included the signal counts in the summary and embedded links for each work unit and task. That way, those who know better than I what it is specifically that they're looking for can more easily winnow down the potential testing candidates and only retrieve those particular WU files from the server.

The one thing I will attempt to do this evening is to replace those "Anonymous platform" IDs with more specific app identifiers wherever I can. Unfortunately, there's not a consistent format for that info in the Stderrs.
ID: 1812335 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1812340 - Posted: 25 Aug 2016, 1:21:32 UTC - in response to Message 1812335.  
Last modified: 25 Aug 2016, 1:23:48 UTC

Yeah choices of which task and result files would be good to dig for and poke at, will probably come down to the individual problems, and how Richard+Raistmer plan to look at them.

For Cuda mysteries I prefer lab condition runs with actual files, while that's just my approach and some others may not need that, depending. Many can be ruled out as flaky hosts/GPUs pretty easily (with stock Cuda).

In the case of Petri Special, we already have a good headstart with the one Pulses example currently under the microscope, because the result files say more than the stderr prints. Probably we'll be looking for more after some things have been nailed down with that, but extensive collection for the Cuda special builds probably won't be necessary short term.
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 1812340 · Report as offensive
Profile Jeff Buck Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1812373 - Posted: 25 Aug 2016, 3:39:01 UTC - in response to Message 1812340.  

Okay, I've tweaked the format of my listing a bit, replacing the "Anonymous platform" designations with, hopefully, more specific app descriptions. I've also identified which Work Units are guppis. The latest file is available from this link. (Unfortunately, Amazon's cloud drive apparently gets all screwed up when one tries to replace an existing file, so the link in my earlier post no longer seems to be valid.)

Examples of the slightly altered format:

Workunit 2243574106
Task 5111259127 (S=7, A=0, P=5, T=0, G=1) v8.12 (opencl_intel_gpu_sah) windows_intelx86
Task 5111259128 (S=8, A=0, P=5, T=0, G=0) x41zi (baseline v8), Cuda 5.00

Workunit 2244749285 (guppi)
Task 5113705190 (S=2, A=0, P=8, T=2, G=0) AVXxj Win64 Build 3330
Task 5113705191 (S=2, A=0, P=11, T=2, G=0) v8.12 (opencl_intel_gpu_sah) windows_intelx86

Many can be ruled out as flaky hosts/GPUs pretty easily (with stock Cuda).

That's one of the reasons I figured that a listing showing the signal counts for each task would help to more easily weed out the "off the rails" hosts, without having to download the WU files or otherwise dig any deeper.
ID: 1812373 · Report as offensive
Kiska
Volunteer tester

Send message
Joined: 31 Mar 12
Posts: 302
Credit: 3,067,762
RAC: 0
Australia
Message 1812412 - Posted: 25 Aug 2016, 6:33:26 UTC

Again an excellent time to see if we get some resends
ID: 1812412 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19012
Credit: 40,757,560
RAC: 67
United Kingdom
Message 1812475 - Posted: 25 Aug 2016, 13:47:21 UTC

Is this the type of thing you are investigating.
blc5_2bit_guppi_57403_68833_HIP11048_OFF_0003.20883.416.21.44.64.vlar

I have completed _4, but not reported yet. Network activity suspended.
ID: 1812475 · Report as offensive
Kiska
Volunteer tester

Send message
Joined: 31 Mar 12
Posts: 302
Credit: 3,067,762
RAC: 0
Australia
Message 1812481 - Posted: 25 Aug 2016, 14:07:40 UTC - in response to Message 1812475.  
Last modified: 25 Aug 2016, 14:12:49 UTC

Yes please upload both result and datafile for analysis, if possible pm me with your email so I can add you to my google drive authorised edit list, so you can upload the results and data
ID: 1812481 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19012
Credit: 40,757,560
RAC: 67
United Kingdom
Message 1812482 - Posted: 25 Aug 2016, 14:13:28 UTC - in response to Message 1812481.  
Last modified: 25 Aug 2016, 14:17:55 UTC

Yes please upload both result and datafile for analysis, if possible pm me with your email so I can add you to my google drive authorised edit list

Do you need my to copy and send you any files still on my computer.

edit] We seem to have got crossed up, PM sent
ID: 1812482 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 . . . 36 · Next

Message boards : Number crunching : Monitoring inconclusive GBT validations and harvesting data for testing


 
©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.