Inconclusive Work Units Running AP Ver 6

Message boards : Number crunching : Inconclusive Work Units Running AP Ver 6
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 10 · Next

AuthorMessage
Profile cliff
Avatar

Send message
Joined: 16 Dec 07
Posts: 625
Credit: 3,590,440
RAC: 0
United Kingdom
Message 1220266 - Posted: 19 Apr 2012, 12:57:59 UTC - in response to Message 1220248.  

ARGH - make that 'too high unroll'

Unless it has to be not too high and not too low, in which case we're in trouble.


Roll the dice:-)


Cheers,
Cliff,
Been there, Done that, Still no damm T shirt!
ID: 1220266 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34258
Credit: 79,922,639
RAC: 80
Germany
Message 1220274 - Posted: 19 Apr 2012, 13:15:34 UTC - in response to Message 1220266.  

ARGH - make that 'too high unroll'

Unless it has to be not too high and not too low, in which case we're in trouble.


Roll the dice:-)


Cheers,


Those values provided in the installer should be absolutely safe.



With each crime and every kindness we birth our future.
ID: 1220274 · Report as offensive
David S
Volunteer tester
Avatar

Send message
Joined: 4 Oct 99
Posts: 18352
Credit: 27,761,924
RAC: 12
United States
Message 1220277 - Posted: 19 Apr 2012, 13:25:01 UTC

Here's an inconclusive that doesn't directly involve me (I was the timeout), but it caught my eye because I thought it seemed like a long time for a timeout to stay in my stats pages.

My original wingperson put in 23,000 seconds of CPU time, but the stderr_txt is blank.

I don't know if that's the kind of thing you guys are looking for,

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

ID: 1220277 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14650
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1220278 - Posted: 19 Apr 2012, 13:28:19 UTC - in response to Message 1220274.  

Those values provided in the installer should be absolutely safe.

The majority of the WUs I've downloaded so far (9 out of 13) relate to NVidia OpenCL r560, which is not distributed through the installer. Any command line values used will either have been guessed by the user on manual installation, or copied from Raistmer's release thread.

The other four tasks downloaded have been CPU inconclusives - two each for r555 and r557

For clarification - I've downloaded every task individually reported by a user, or reported in 'small' batches: so that distribution represents the user reporting pattern, not my selection. I'll be working on the 'bulk' reports by X-Files and hbomber this afternoon.
ID: 1220278 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14650
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1220281 - Posted: 19 Apr 2012, 13:31:50 UTC - in response to Message 1220277.  

Here's an inconclusive ...

Thanks, but that one's a multibeam VLAR - I've got my hands rather full with AP v6 just at the moment. I'll ask James to make the thread title a bit clearer. Sorry about that.
ID: 1220281 · Report as offensive
Profile perryjay
Volunteer tester
Avatar

Send message
Joined: 20 Aug 02
Posts: 3377
Credit: 20,676,751
RAC: 0
United States
Message 1220294 - Posted: 19 Apr 2012, 14:20:08 UTC - in response to Message 1220281.  

One of my inconclusives just validated against two stock http://setiathome.berkeley.edu/workunit.php?wuid=970100390


PROUD MEMBER OF Team Starfire World BOINC
ID: 1220294 · Report as offensive
Horacio

Send message
Joined: 14 Jan 00
Posts: 536
Credit: 75,967,266
RAC: 0
Argentina
Message 1220317 - Posted: 19 Apr 2012, 15:28:59 UTC

This WU (ap_26no11ae_B2_P0_00210_20120315_31464.wu) crunched on my GPU 560Ti, already reported here as inconclusive, finally ended as invalid.

Until now all the others remain as inconclusives.


ID: 1220317 · Report as offensive
LadyL
Volunteer tester
Avatar

Send message
Joined: 14 Sep 11
Posts: 1679
Credit: 5,230,097
RAC: 0
Message 1220331 - Posted: 19 Apr 2012, 16:02:33 UTC - in response to Message 1220258.  
Last modified: 19 Apr 2012, 16:31:56 UTC

I know they are same, the units. But I had to note it, to prevent any further questions or misunderstanding.
I can run same unit on GPU with different unroll, next on CPU(presuming it will be valid), and compare them eventually. Thus I can avoid that long wait until they get validated by server, producing junk and wasting energy.

The problem is I don't know how to compare the results.
If there is a relatively easy way - comparison of output file e.g, I'll find the proper unroll value for my card.


You can easily run these values on your card.

<cmdline>-instances_per_device 2 -unroll 10 -ffa_block 6144 -ffa_block_fetch 2048</cmdline>

unroll 12 might be possible but i suggest to try 10 first.


I may be mistaken, but I think I saw unroll 4 in hbombers units!

yes unroll 4. all the inconclusives come up with 30 rep pulses, which according to Raistmer is the sign of a silent ap failure

@ hbomber try decreasing unroll to 2.

@ Mike yes higher unroll should be possible but obviously isn't.
I'm not the Pope. I don't speak Ex Cathedra!
ID: 1220331 · Report as offensive
LadyL
Volunteer tester
Avatar

Send message
Joined: 14 Sep 11
Posts: 1679
Credit: 5,230,097
RAC: 0
Message 1220338 - Posted: 19 Apr 2012, 16:36:48 UTC - in response to Message 1220254.  
Last modified: 19 Apr 2012, 16:39:47 UTC

The problem is I don't know how to compare the results.
If there is a relatively easy way - comparison of output file e.g, I'll find the proper unroll value for my card.


AP bench 210. Still haven't had a change to update that to 2.11.

Comes with a very Basic readme. the config file allows to run the same WU with the same app and different settings.

edit if you already have the result files, it's probably enough to run the mock validator
I'm not the Pope. I don't speak Ex Cathedra!
ID: 1220338 · Report as offensive
Profile arkayn
Volunteer tester
Avatar

Send message
Joined: 14 May 99
Posts: 4438
Credit: 55,006,323
RAC: 0
United States
Message 1220340 - Posted: 19 Apr 2012, 16:37:08 UTC

Currently I have had no invalids on my GTX560, I am running at stock speeds with 290.36 and these settings.
-instances_per_device 1 -unroll 10 -ffa_block 6144 -ffa_block_fetch 1536 -sbs 256

I also only run AP alongside 1MB on the card by going with 0.51 for AP and 0.49 for MB.

This computer.
http://setiathome.berkeley.edu/results.php?hostid=5338281

ID: 1220340 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34258
Credit: 79,922,639
RAC: 80
Germany
Message 1220344 - Posted: 19 Apr 2012, 16:48:04 UTC - in response to Message 1220340.  

Currently I have had no invalids on my GTX560, I am running at stock speeds with 290.36 and these settings.
-instances_per_device 1 -unroll 10 -ffa_block 6144 -ffa_block_fetch 1536 -sbs 256

I also only run AP alongside 1MB on the card by going with 0.51 for AP and 0.49 for MB.

This computer.
http://setiathome.berkeley.edu/results.php?hostid=5338281


Thanks Arkayn.

Proves my imagination.
Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.



With each crime and every kindness we birth our future.
ID: 1220344 · Report as offensive
LadyL
Volunteer tester
Avatar

Send message
Joined: 14 Sep 11
Posts: 1679
Credit: 5,230,097
RAC: 0
Message 1220348 - Posted: 19 Apr 2012, 16:57:07 UTC - in response to Message 1220344.  

Currently I have had no invalids on my GTX560, I am running at stock speeds with 290.36 and these settings.
-instances_per_device 1 -unroll 10 -ffa_block 6144 -ffa_block_fetch 1536 -sbs 256

I also only run AP alongside 1MB on the card by going with 0.51 for AP and 0.49 for MB.

This computer.
http://setiathome.berkeley.edu/results.php?hostid=5338281


Thanks Arkayn.

Proves my imagination.
Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.


That's a 560 not a 560Ti
I'm not the Pope. I don't speak Ex Cathedra!
ID: 1220348 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34258
Credit: 79,922,639
RAC: 80
Germany
Message 1220349 - Posted: 19 Apr 2012, 16:58:07 UTC - in response to Message 1220348.  
Last modified: 19 Apr 2012, 16:59:25 UTC

Currently I have had no invalids on my GTX560, I am running at stock speeds with 290.36 and these settings.
-instances_per_device 1 -unroll 10 -ffa_block 6144 -ffa_block_fetch 1536 -sbs 256

I also only run AP alongside 1MB on the card by going with 0.51 for AP and 0.49 for MB.

This computer.
http://setiathome.berkeley.edu/results.php?hostid=5338281


Thanks Arkayn.

Proves my imagination.
Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.


That's a 560 not a 560Ti


Exactely.
I was more thinking about Horacio.
He has 560ti.


With each crime and every kindness we birth our future.
ID: 1220349 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14650
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1220350 - Posted: 19 Apr 2012, 17:00:34 UTC - in response to Message 1220344.  

Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.

Of the seven hosts I'm tracking so far for inconclusives, only one is running a 560Ti

I've just downloaded 49 inconclusive datafiles from X-Files' GTX 580 with unroll=8 - what do you suggest is wrong with that one?
ID: 1220350 · Report as offensive
LadyL
Volunteer tester
Avatar

Send message
Joined: 14 Sep 11
Posts: 1679
Credit: 5,230,097
RAC: 0
Message 1220352 - Posted: 19 Apr 2012, 17:07:54 UTC - in response to Message 1220350.  

Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.

Of the seven hosts I'm tracking so far for inconclusives, only one is running a 560Ti

I've just downloaded 49 inconclusive datafiles from X-Files' GTX 580 with unroll=8 - what do you suggest is wrong with that one?


the inconclusives I checked are all 30 rep pulses.
looks like with unroll 8 on that card it sometimes works and sometimes doesn't.

@ X-Files try and decrease unroll
I'm not the Pope. I don't speak Ex Cathedra!
ID: 1220352 · Report as offensive
Profile cliff
Avatar

Send message
Joined: 16 Dec 07
Posts: 625
Credit: 3,590,440
RAC: 0
United Kingdom
Message 1220353 - Posted: 19 Apr 2012, 17:07:56 UTC - in response to Message 1220350.  

Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.

Of the seven hosts I'm tracking so far for inconclusives, only one is running a 560Ti

I've just downloaded 49 inconclusive datafiles from X-Files' GTX 580 with unroll=8 - what do you suggest is wrong with that one?


Richard I'd like to point out my GTX560' are the Palit OC version..
And I'm getting as you know 'some' invalid and some valid..

It might be the 560TI's are running hot, mine are realtivly cool, fans at max and held that way..

Could this in anyway relate to the degree of blanking of the WU's in question?

Regards,
Cliff,
Been there, Done that, Still no damm T shirt!
ID: 1220353 · Report as offensive
Horacio

Send message
Joined: 14 Jan 00
Posts: 536
Credit: 75,967,266
RAC: 0
Argentina
Message 1220354 - Posted: 19 Apr 2012, 17:11:43 UTC - in response to Message 1220350.  
Last modified: 19 Apr 2012, 17:17:49 UTC

Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.

Of the seven hosts I'm tracking so far for inconclusives, only one is running a 560Ti

I've just downloaded 49 inconclusive datafiles from X-Files' GTX 580 with unroll=8 - what do you suggest is wrong with that one?


My 560Tis are not OC'ed. I've raised the core voltages to avoid the downclocks and that was good also to avoid the apps finding more than 30 false results.

Anyway, I've already said that one of them is not working very good, and stated that those inconclusives may not be relevant but my idea was to give Richard the oportunity to catch the WUs before they gone...

ETA: Ive noticed something here... Unroll 8 is what Im using... (10 gave me bad results with the previous version of the NV app)
ID: 1220354 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14650
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1220357 - Posted: 19 Apr 2012, 17:16:54 UTC - in response to Message 1220353.  

Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.

Of the seven hosts I'm tracking so far for inconclusives, only one is running a 560Ti

I've just downloaded 49 inconclusive datafiles from X-Files' GTX 580 with unroll=8 - what do you suggest is wrong with that one?

Richard I'd like to point out my GTX560' are the Palit OC version..
And I'm getting as you know 'some' invalid and some valid..

It might be the 560TI's are running hot, mine are realtivly cool, fans at max and held that way..

Could this in anyway relate to the degree of blanking of the WU's in question?

Regards,

The big problem with (some of the) early 560Ti wasn't just that they were overclocked, but that various other factory settings weren't properly set to support the overclock - notably, chipset voltages. And if the card was being powered by a marginal power supply, that just made things worse.

Overclocking, especially factory overclocking, is usually OK on its own - I've never had any problems with my factory overclocked Zotac AMP 9800GTX+ - but if you push it beyond the factory overclock envelope, and if you don't do it right, there can be problems.
ID: 1220357 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34258
Credit: 79,922,639
RAC: 80
Germany
Message 1220358 - Posted: 19 Apr 2012, 17:18:29 UTC - in response to Message 1220352.  

Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.

Of the seven hosts I'm tracking so far for inconclusives, only one is running a 560Ti

I've just downloaded 49 inconclusive datafiles from X-Files' GTX 580 with unroll=8 - what do you suggest is wrong with that one?


the inconclusives I checked are all 30 rep pulses.
looks like with unroll 8 on that card it sometimes works and sometimes doesn't.

@ X-Files try and decrease unroll


12 out of my last 20 APs i crunched with GPU had 30 rep pulses.
All validated in first attempt.



With each crime and every kindness we birth our future.
ID: 1220358 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34258
Credit: 79,922,639
RAC: 80
Germany
Message 1220362 - Posted: 19 Apr 2012, 17:21:15 UTC - in response to Message 1220350.  

Keep in mind most of 560ti are factory overclocked.
Remember my thread at lunatics about this behaviour.

Of the seven hosts I'm tracking so far for inconclusives, only one is running a 560Ti

I've just downloaded 49 inconclusive datafiles from X-Files' GTX 580 with unroll=8 - what do you suggest is wrong with that one?


We should at least know if those cards are overclocked or not.
Lots of users overclock specially memmory for better output.

Doesn´t mean i dissagree with Raistmer here.



With each crime and every kindness we birth our future.
ID: 1220362 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 10 · Next

Message boards : Number crunching : Inconclusive Work Units Running AP Ver 6


 
©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.