Linux CUDA 'Special' App finally available, featuring Low CPU use

Message boards : Number crunching : Linux CUDA 'Special' App finally available, featuring Low CPU use
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 37 · 38 · 39 · 40 · 41 · 42 · 43 . . . 83 · 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 1875374 - Posted: 27 Jun 2017, 12:04:07 UTC - in response to Message 1875364.  

...All those people testing these Apps at Beta and no one picked this up? Nevermind.


This far into the noise floor, there is no shame, only things to learn. Nobody's been here before.
"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: 1875374 · Report as offensive     Reply Quote
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 4261
Credit: 261,260,505
RAC: 363,258
United States
Message 1875398 - Posted: 27 Jun 2017, 23:00:25 UTC - in response to Message 1875364.  

And the results are...
My OSX CPU Agrees with the CPU SETI@home v8 v8.06 (alt) windows_x86_64 here, https://setiweb.ssl.berkeley.edu/beta/workunit.php?wuid=9796907
Pretty close match, the important parts are;
8.06 (alt) windows
Gaussian: peak=2.961493, mean=0.5020308, ChiSq=1.415568, time=17.62, d_freq=1420573507.32, score=0.3145776, null_hyp=2.268677, chirp=-98.677, fft_len=16k
Best gaussian: peak=3.659641, mean=0.5301717, ChiSq=1.25771, time=67.95, d_freq=1420577155.89, score=0.794832, null_hyp=2.202944, chirp=-65.055, fft_len=16k
SSE4.1xjf OS X r 3344
Gaussian: peak=2.961488, mean=0.5020312, ChiSq=1.415569, time=17.62, d_freq=1420573507.32, score=0.3145256, null_hyp=2.268674, chirp=-98.677, fft_len=16k
Best gaussian: peak=3.659646, mean=0.5301709, ChiSq=1.257715, time=67.95, d_freq=1420577155.89, score=0.7949486, null_hyp=2.202953, chirp=-65.055, fft_len=16k
I decided to see if I could get my new Windows install working with the Benchmark. I haven't run it in some time...ever since WinNSA came out. I didn't have any trouble running the Benchmark after chasing down all the files, and installing all those updates. I ran the same task above on the machine with two 1050s;

MBbench210.cmd
======================================
1 testWU(s) found
(04oc08ab.8413.17250.11.45.59.wu)
1 science app(s) found
(MB8_win_x86_SSE3_OpenCL_NV_SoG_r3584.exe -sbs 256 -spike_fft_thresh 2048 -oclfft_tune_gr 256 -oclfft_tune_wg 256 -period_iterations_num 10)
Gaussian: peak=2.961486, mean=0.5020315, ChiSq=1.415564, time=17.62, d_freq=1420573507.32, score=0.3144556, null_hyp=2.268668, chirp=-98.677, fft_len=16k
Best gaussian: peak=2.961486, mean=0.5020315, ChiSq=1.415564, time=17.62, d_freq=1420573507.32, score=0.3144556, null_hyp=2.268668, chirp=-98.677, fft_len=16k

I also ran 09no16aa.18442.2116.6.33.31.wu;
Gaussian: peak=5.736884, mean=0.6174232, ChiSq=1.393288, time=69.63, d_freq=1420305107.67, score=0.6122654, null_hyp=2.236684, chirp=-62.462, fft_len=16k
Gaussian: peak=5.793325, mean=0.5903414, ChiSq=1.403672, time=36.07, d_freq=1420298442.12, score=1.986567, null_hyp=2.315829, chirp=-71.95, fft_len=16k
Gaussian: peak=5.745589, mean=0.583622, ChiSq=1.414207, time=59.56, d_freq=1420299457.13, score=2.05784, null_hyp=2.326043, chirp=90.136, fft_len=16k
Best gaussian: peak=5.745589, mean=0.583622, ChiSq=1.414207, time=59.56, d_freq=1420299457.13, score=2.05784, null_hyp=2.326043, chirp=90.136, fft_len=16k

It definitely uses a reported signal as Best.
ID: 1875398 · Report as offensive     Reply Quote
Profile Jeff Buck Project Donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1875403 - Posted: 27 Jun 2017, 23:26:36 UTC - in response to Message 1875398.  
Last modified: 27 Jun 2017, 23:28:54 UTC

I also ran 09no16aa.18442.2116.6.33.31.wu;
Gaussian: peak=5.736884, mean=0.6174232, ChiSq=1.393288, time=69.63, d_freq=1420305107.67, score=0.6122654, null_hyp=2.236684, chirp=-62.462, fft_len=16k
Gaussian: peak=5.793325, mean=0.5903414, ChiSq=1.403672, time=36.07, d_freq=1420298442.12, score=1.986567, null_hyp=2.315829, chirp=-71.95, fft_len=16k
Gaussian: peak=5.745589, mean=0.583622, ChiSq=1.414207, time=59.56, d_freq=1420299457.13, score=2.05784, null_hyp=2.326043, chirp=90.136, fft_len=16k
Best gaussian: peak=5.745589, mean=0.583622, ChiSq=1.414207, time=59.56, d_freq=1420299457.13, score=2.05784, null_hyp=2.326043, chirp=90.136, fft_len=16k

It definitely uses a reported signal as Best.
And I ran that one overnight last night, as well.
setiathome_8.00_windows_intelx86.exe -verb -nog / 09no16aa.18442.2116.6.33.31.wu:
The 3 reported gaussians were:
<peak_power>5.7368845939636</peak_power>
<peak_power>5.7933268547058</peak_power>
<peak_power>5.7455887794495</peak_power>

... whereas the Best Gaussian was:
<peak_power>6.3855581283569</peak_power>

How there can be a "best" signal that isn't worth reporting (when there are apparently 3 inferior signals that are) is beyond me, but that's apparently the standard. :^)
ID: 1875403 · Report as offensive     Reply Quote
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1875421 - Posted: 28 Jun 2017, 1:02:47 UTC - in response to Message 1875403.  

How there can be a "best" signal that isn't worth reporting (when there are apparently 3 inferior signals that are) is beyond me, but that's apparently the standard. :^)


For best there is a check, added ~2011, of the CHiSq fit ('i.e. 'Gaussian-ness') , in addition to the score used for reporting. My cursory reading suggests the Best may be reportable, maybe not, though yet to do a full line by line analysis. The suspected variation is in the multiple different implementations in that logic in the different branches, though that doesn't rule out other bugs or cumulative error
"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: 1875421 · Report as offensive     Reply Quote
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 4261
Credit: 261,260,505
RAC: 363,258
United States
Message 1875428 - Posted: 28 Jun 2017, 1:32:28 UTC - in response to Message 1875421.  
Last modified: 28 Jun 2017, 2:30:18 UTC

Previous tests indicate only the Apps compiled using the SoG path have this *feature* of using reported as Best. All the other Apps that don't use the SoG compile path, that I've seen, give the correct Best Gaussian. That can be seen in the Results I posted at Crunchers Anonymous, the previous posts in this thread, and the recent test against the ATI Apps here, http://setiathome.berkeley.edu/workunit.php?wuid=2586601005 In that test My CPU agreed with zi3v, SETI@home v8 v8.22 (opencl_ati5_SoG_cat132) windows_intelx86 Failed, while SETI@home v8 v8.22 (opencl_ati_nocal) windows_intelx86 must have agreed with zi3v as zi3v was give the *canonical result - 5833121549*. At least that time zi3v wasn't robbed of canonical, I suspect My Inconclusives would be Lower if the SoGs were reporting the Correct Best Gaussian.

Fortunately, there are Non-SoG builds readily available for Most Apps, https://setiathome.berkeley.edu/apps.php except maybe Windows nVidia...the most used. There is a Mac Non-SoG ATI App at Beta that's been there as long as the 8.20 SoG build. The Mac nVidia SoG App never existed as I couldn't get it to work correctly using the SoG path.

A New Windows CPU App?
Windows/x86 running on an AMD x86_64 or Intel EM64T CPU 8.07 (alt) 28 Jun 2017, 1:13:32 UTC 0 GigaFLOPS
https://setiweb.ssl.berkeley.edu/beta/apps.php
OK boys, fire up those Multi-core CPUs again. This time try using the Stock CPU App.
ID: 1875428 · Report as offensive     Reply Quote
Profile Jeff Buck Project Donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1875430 - Posted: 28 Jun 2017, 1:44:56 UTC - in response to Message 1875421.  

How there can be a "best" signal that isn't worth reporting (when there are apparently 3 inferior signals that are) is beyond me, but that's apparently the standard. :^)


For best there is a check, added ~2011, of the CHiSq fit ('i.e. 'Gaussian-ness') , in addition to the score used for reporting. My cursory reading suggests the Best may be reportable, maybe not, though yet to do a full line by line analysis. The suspected variation is in the multiple different implementations in that logic in the different branches, though that doesn't rule out other bugs or cumulative error
I mean, I can understand situations where the "best" signal, of any type, still wouldn't be good enough to "report" as worthy of further investigation. However, it seems to me that if one or more signals do achieve that reportable threshold, that the "best" signal should be one of those. If it's not, it just seems really screwy to me. Out of sync, I guess. Perhaps the dictionary the scientists use has a different definition of "best" than the one most of us common folk use. ;^)
ID: 1875430 · Report as offensive     Reply Quote
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1875435 - Posted: 28 Jun 2017, 3:14:47 UTC - in response to Message 1875430.  

How there can be a "best" signal that isn't worth reporting (when there are apparently 3 inferior signals that are) is beyond me, but that's apparently the standard. :^)


For best there is a check, added ~2011, of the CHiSq fit ('i.e. 'Gaussian-ness') , in addition to the score used for reporting. My cursory reading suggests the Best may be reportable, maybe not, though yet to do a full line by line analysis. The suspected variation is in the multiple different implementations in that logic in the different branches, though that doesn't rule out other bugs or cumulative error
I mean, I can understand situations where the "best" signal, of any type, still wouldn't be good enough to "report" as worthy of further investigation. However, it seems to me that if one or more signals do achieve that reportable threshold, that the "best" signal should be one of those. If it's not, it just seems really screwy to me. Out of sync, I guess. Perhaps the dictionary the scientists use has a different definition of "best" than the one most of us common folk use. ;^)


Certainly something worthy of bringing up with Eric IMO. He may well examine the stock CPU code and say 'That's not what was intended', or say 'that's correct'. In terms of purpose, the 'best' is used for Screensaver display, so it would entirely make sense to me if the intent is to choose the most 'Gaussian-ey' looking signal to display, whether reportable or not (i.e . marketing). Naturally I can also see the point of view that if the score wasn't good enough to rep[ort, then why store it at all ? Unfortunately the CHiSq and null hypotheses aggravate a part of my brain that burned out on statistics long ago (as I was too good at it and fried that area of my brain), therefore I don't have definitive answers on what's meant to happen in this particular case.
"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: 1875435 · Report as offensive     Reply Quote
Profile Jeff Buck Project Donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1875436 - Posted: 28 Jun 2017, 4:04:12 UTC - in response to Message 1875435.  

Certainly something worthy of bringing up with Eric IMO. He may well examine the stock CPU code and say 'That's not what was intended', or say 'that's correct'.
Yeah, that's definitely the key determination needed before any "fixing" gets done to whichever is not the "correct" path. Alternatively, I suppose, he could alter the validator so that it ignores Best Gaussian differences if all other signals match.

In terms of purpose, the 'best' is used for Screensaver display, ....(i.e . marketing).
Oh, goody. Perhaps they could also use that feature to hawk some limited edition SETI@home toasters to help fund the project! ;-P
ID: 1875436 · Report as offensive     Reply Quote
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1875441 - Posted: 28 Jun 2017, 4:32:54 UTC - in response to Message 1875436.  

Still waiting on my Seti Toaster :(
"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: 1875441 · Report as offensive     Reply Quote
Profile Zalster Project Donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 4674
Credit: 294,721,646
RAC: 432,500
United States
Message 1875442 - Posted: 28 Jun 2017, 4:36:46 UTC - in response to Message 1875441.  

Still waiting on my Seti Toaster :(


Did someone say toaster???

ID: 1875442 · Report as offensive     Reply Quote
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1875443 - Posted: 28 Jun 2017, 4:43:59 UTC

Tweeted Eric:
@SETIEric If a 'best Gaussian' looks more 'Guassianey' than the reportables, why may it not necessarily be reportable ?

"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: 1875443 · Report as offensive     Reply Quote
Profile Jeff Buck Project Donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1875445 - Posted: 28 Jun 2017, 4:50:20 UTC - in response to Message 1875442.  
Last modified: 28 Jun 2017, 4:51:03 UTC

They gots to actually be flying, dontcha know!
ID: 1875445 · Report as offensive     Reply Quote
Profile Zalster Project Donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 4674
Credit: 294,721,646
RAC: 432,500
United States
Message 1875446 - Posted: 28 Jun 2017, 4:51:52 UTC - in response to Message 1875445.  

lol
ID: 1875446 · Report as offensive     Reply Quote
Profile Jeff Buck Project Donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1875448 - Posted: 28 Jun 2017, 5:10:33 UTC - in response to Message 1875443.  

Tweeted Eric:
@SETIEric If a 'best Gaussian' looks more 'Guassianey' than the reportables, why may it not necessarily be reportable ?
Heh, I suppose it's just my old age, but I tend to have a hard time keeping a straight face when I read that somebody "Tweeted" something. It always seems about as frivolous as, oh I dunno, flying toasters perhaps! ;^D
ID: 1875448 · Report as offensive     Reply Quote
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1875449 - Posted: 28 Jun 2017, 5:16:11 UTC - in response to Message 1875448.  
Last modified: 28 Jun 2017, 5:17:37 UTC

Tweeted Eric:
@SETIEric If a 'best Gaussian' looks more 'Guassianey' than the reportables, why may it not necessarily be reportable ?
Heh, I suppose it's just my old age, but I tend to have a hard time keeping a straight face when I read that somebody "Tweeted" something. It always seems about as frivolous as, oh I dunno, flying toasters perhaps! ;^D


Oh you'd be surprised [as I was]. The immediacy bypasses all sorts of tradition and other impediments. Eliminates the old 'Chinese Whispers' (aka Fake news)
"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: 1875449 · Report as offensive     Reply Quote
Profile Jeff Buck Project Donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1875451 - Posted: 28 Jun 2017, 5:25:38 UTC - in response to Message 1875449.  

Oh you'd be surprised [as I was]. The immediacy bypasses all sorts of tradition and other impediments. Eliminates the old 'Chinese Whispers' (aka Fake news)
TraDITION! Oh great, first Flying Toasters and now Fiddler on the Roof flashbacks. I think it's past my bedtime.
ID: 1875451 · Report as offensive     Reply Quote
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1875452 - Posted: 28 Jun 2017, 5:33:22 UTC - in response to Message 1875451.  

Oh you'd be surprised [as I was]. The immediacy bypasses all sorts of tradition and other impediments. Eliminates the old 'Chinese Whispers' (aka Fake news)
TraDITION! Oh great, first Flying Toasters and now Fiddler on the Roof flashbacks. I think it's past my bedtime.


*opens beer* ... Guess my work here is done :)
"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: 1875452 · Report as offensive     Reply Quote
Profile Jeff Buck Project Donor
Volunteer tester

Send message
Joined: 11 Feb 00
Posts: 1441
Credit: 148,764,870
RAC: 0
United States
Message 1875453 - Posted: 28 Jun 2017, 5:52:36 UTC
Last modified: 28 Jun 2017, 5:55:12 UTC

Ah, but I guess I have one more post to make before cutting some ZZZs. And back on topic, too. After 14+ hours my Windows CPU (setiathome_8.00_windows_intelx86) bench of 23se08ac.6875.22968.6.33.135 just finished. TBar had indicated that he thought this WU was a bit of a problem case. He may have been right. My original post:

Workunit 2573263722 (23se08ac.6875.22968.6.33.135)
Task 5805117074 (S=3, A=0, P=1, T=3, G=0) x41p_zi3t2b, Cuda 8.00 special
Task 5805117075 (S=3, A=0, P=1, T=3, G=0) v8.22 (opencl_nvidia_SoG) windows_intelx86

Cuda 8.00 special - Best gaussian: peak=3.252388, mean=0.5397108, ChiSq=1.344394, time=14.26, d_freq=1418816790.11,
score=-1.169299, null_hyp=2.144445, chirp=-39.071, fft_len=16k
v8.22 SoG - Best gaussian: peak=3.76217, mean=0.5480909, ChiSq=1.226871, time=39.43, d_freq=1418822660.68,
score=-1.169124, null_hyp=2.078196, chirp=43.425, fft_len=16k
Well, it seems that in this case the "gold standard" agrees with SoG:
<best_gaussian>
<peak_power>3.7621715068817</peak_power>

G'night.
ID: 1875453 · Report as offensive     Reply Quote
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 1875454 - Posted: 28 Jun 2017, 6:00:10 UTC - in response to Message 1875453.  
Last modified: 28 Jun 2017, 6:01:52 UTC

Ah, but I guess I have one more post to make before cutting some ZZZs. And back on topic, too. After 14+ hours my Windows CPU (setiathome_8.00_windows_intelx86) bench of 23se08ac.6875.22968.6.33.135 just finished. TBar had indicated that he thought this WU was a bit of a problem case. He may have been right. My original post:

Workunit 2573263722 (23se08ac.6875.22968.6.33.135)
Task 5805117074 (S=3, A=0, P=1, T=3, G=0) x41p_zi3t2b, Cuda 8.00 special
Task 5805117075 (S=3, A=0, P=1, T=3, G=0) v8.22 (opencl_nvidia_SoG) windows_intelx86

Cuda 8.00 special - Best gaussian: peak=3.252388, mean=0.5397108, ChiSq=1.344394, time=14.26, d_freq=1418816790.11,
score=-1.169299, null_hyp=2.144445, chirp=-39.071, fft_len=16k
v8.22 SoG - Best gaussian: peak=3.76217, mean=0.5480909, ChiSq=1.226871, time=39.43, d_freq=1418822660.68,
score=-1.169124, null_hyp=2.078196, chirp=43.425, fft_len=16k
Well, it seems that in this case the "gold standard" agrees with SoG:
<best_gaussian>
<peak_power>3.7621715068817</peak_power>

G'night.


Exactly. Note the Higher ChiSq. Therefore the Cuda 8 special one looks more 'Gaussianey' than the 8.22 SoG one. Hence my Tweet/Query to Eric.
"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: 1875454 · Report as offensive     Reply Quote
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6005
Credit: 83,790,133
RAC: 27,962
Russia
Message 1875460 - Posted: 28 Jun 2017, 7:13:40 UTC - in response to Message 1875364.  

All those people testing these Apps at Beta and no one picked this up? Nevermind.

And you was just one of them, if I recall correctly :D

The right question to ask how non-reportable could be better than reportable one.?....

And in this test reportable comes later in processing chain so SoG operated in "no reportable so far" path again.
SETI apps news
We're not gonna fight them. We're gonna transcend them.
ID: 1875460 · Report as offensive     Reply Quote
Previous · 1 . . . 37 · 38 · 39 · 40 · 41 · 42 · 43 . . . 83 · Next

Message boards : Number crunching : Linux CUDA 'Special' App finally available, featuring Low CPU use


 
©2018 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.