Optimized Windows Application

Message boards : Number crunching : Optimized Windows Application

To post messages, you must log in.

1 · 2 · 3 · Next

AuthorMessage
Profile Kajunfisher
Avatar

Send message
Joined: 16 Sep 05
Posts: 15
Credit: 10,292
RAC: 0
Message 93 - Posted: 17 Sep 2005, 16:04:07 UTC

Could you please post something on the front page (or here) when the optimized client is available.

Could you also please post a link showing what applications are being used.

Thank you :-)
ID: 93 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile David E K
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 1 Jul 05
Posts: 1018
Credit: 4,334,829
RAC: 0
Message 102 - Posted: 17 Sep 2005, 17:22:45 UTC

Will do.
ID: 102 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile PhilippeCM
Avatar

Send message
Joined: 17 Sep 05
Posts: 2
Credit: 39,560
RAC: 0
Message 265 - Posted: 20 Sep 2005, 22:55:48 UTC - in response to Message 102.  

Will do.


Could you maybe give us some info on progress?
Thanks
P

"Tamdiu discendum est, quamdiu vivas" Seneca
ID: 265 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile FZB

Send message
Joined: 17 Sep 05
Posts: 84
Credit: 4,948,999
RAC: 0
Message 267 - Posted: 21 Sep 2005, 0:34:02 UTC

david posted in some other post that he is on vacation next week and hopes to have the optimized client out before then.
--
Florian
www.domplatz1.de
ID: 267 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile David E K
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 1 Jul 05
Posts: 1018
Credit: 4,334,829
RAC: 0
Message 306 - Posted: 22 Sep 2005, 4:30:57 UTC - in response to Message 267.  

Hi all,

I just updated the windows and linux app to version 4.76, still working on the mac app. The windows app should run faster and both hopefully checkpoint the total cpu time. It checkpoints after each structure is made, so at most, the time it takes to make a structure can still be lost. 12 structures are being made for the current workunits. The percent complete does not relate to the time complete. Let me know how it goes.
ID: 306 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Snake Doctor
Avatar

Send message
Joined: 17 Sep 05
Posts: 182
Credit: 6,401,938
RAC: 0
Message 308 - Posted: 22 Sep 2005, 5:24:31 UTC - in response to Message 306.  
Last modified: 22 Sep 2005, 5:24:57 UTC

Hi all,

I just updated the windows and linux app to version 4.76, still working on the mac app. The windows app should run faster and both hopefully checkpoint the total cpu time. It checkpoints after each structure is made, so at most, the time it takes to make a structure can still be lost. 12 structures are being made for the current workunits. The percent complete does not relate to the time complete. Let me know how it goes.

On behalf of the Mac community thank you for the update. Not to put too fine a point on this but if you could make use of altivec code in the compile it ill save you a lot of time later optimizing the Mac application.

Regards and thank you for your efforts.
Phil

We Must look for intelligent life on other planets as,
it is becoming increasingly apparent we will not find any on our own.
ID: 308 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
RDC

Send message
Joined: 16 Sep 05
Posts: 43
Credit: 101,644
RAC: 0
Message 310 - Posted: 22 Sep 2005, 5:37:05 UTC - in response to Message 306.  

Hi all,

I just updated the windows and linux app to version 4.76, still working on the mac app.


Thanks for the update. Can't wait to finish this WU up that I'm crunching now to try the new application out.
ID: 310 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile David E K
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 1 Jul 05
Posts: 1018
Credit: 4,334,829
RAC: 0
Message 313 - Posted: 22 Sep 2005, 5:49:13 UTC - in response to Message 310.  

Hi all,

I just updated the windows and linux app to version 4.76, still working on the mac app.


Thanks for the update. Can't wait to finish this WU up that I'm crunching now to try the new application out.



I'm seeing errors. I hope there isn't a consistent problem. I guess I will know tomorrow.
ID: 313 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Webmaster Yoda
Avatar

Send message
Joined: 17 Sep 05
Posts: 161
Credit: 162,253
RAC: 0
Message 314 - Posted: 22 Sep 2005, 9:13:03 UTC - in response to Message 313.  
Last modified: 22 Sep 2005, 9:15:50 UTC

I'm seeing errors. I hope there isn't a consistent problem. I guess I will know tomorrow.


https://boinc.bakerlab.org/rosetta/result.php?resultid=16270 seems to have worked fine with the new app. Progress meter is still giving seemingly useless information but the work unit finished in 4 hours and 6 minutes.

The estimated time for a work unit still needs more work. BOINC went straight into panic mode when it got this work unit. And that's despite this machine having ample crunching power and being on 24/7. (3.4GHz Pentium 4, 1GB RAM)

I know it's hard to estimate these things, but 162 hours for something that ends up taking 4 isn't even close (at least, not on a linear scale)

I don't watch the progress meter all the time but I did notice that after about an hour, it was at 73.33% and a while later it was back at 1%. Time elapsed did not however go back to zero.

Things are definitely looking up.

*** Join BOINC@Australia today ***
ID: 314 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Honza

Send message
Joined: 18 Sep 05
Posts: 48
Credit: 173,517
RAC: 0
Message 315 - Posted: 22 Sep 2005, 9:57:17 UTC

Rosetta 4.76 on AMD X2 done in about 3 hours with no errors.
Got another 5 to do...
ID: 315 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Honza

Send message
Joined: 18 Sep 05
Posts: 48
Credit: 173,517
RAC: 0
Message 317 - Posted: 22 Sep 2005, 10:33:23 UTC

after 35 minutes, two Rosetta WUs are in 40%, resp 66% of computing.
Memory usage is still only 35MB?
Anybody else has been monitoring memory usage of 4.76 Rosetta WUs?
ID: 317 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Webmaster Yoda
Avatar

Send message
Joined: 17 Sep 05
Posts: 161
Credit: 162,253
RAC: 0
Message 318 - Posted: 22 Sep 2005, 10:44:26 UTC - in response to Message 317.  
Last modified: 22 Sep 2005, 10:53:50 UTC

Anybody else has been monitoring memory usage of 4.76 Rosetta WUs?


Not constantly, but the one WU I processed did get up to 100MB (maybe more)

Running one at the moment, using 35.5MB (20% progress, 25 minutes, Athlon XP 3000+)

*** Join BOINC@Australia today ***
ID: 318 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile KWSN Sir Clark

Send message
Joined: 18 Sep 05
Posts: 46
Credit: 387,432
RAC: 0
Message 319 - Posted: 22 Sep 2005, 10:49:46 UTC - in response to Message 318.  

Anybody else has been monitoring memory usage of 4.76 Rosetta WUs?


Not constantly, but the one WU I processed did get up to 100kB (maybe more)

Running one at the moment, using 35.5MB (20% progress, 25 minutes, Athlon XP 3000+)

Mine's running at 99MB memory wise....for some reason, it started at 1.5hrs to completion and now has dropped back to 1% with 120+ hours but I'm not worried given the est. times are way off
ID: 319 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Honza

Send message
Joined: 18 Sep 05
Posts: 48
Credit: 173,517
RAC: 0
Message 320 - Posted: 22 Sep 2005, 10:55:08 UTC

See the results there
https://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=434

All with error.
I *think* I know how it happend: with no memory left, ALL Rosetta WUs are immediately erroing-out. I opened large files in Photoshop and run of out memory and ALL Rosetta WUs terminated - both those partially completed and those that haven't started (I assume they get to run but erroer out imediatelly).
ID: 320 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Webmaster Yoda
Avatar

Send message
Joined: 17 Sep 05
Posts: 161
Credit: 162,253
RAC: 0
Message 321 - Posted: 22 Sep 2005, 11:25:34 UTC - in response to Message 319.  
Last modified: 22 Sep 2005, 11:27:01 UTC

Mine's running at 99MB memory wise....for some reason, it started at 1.5hrs to completion and now has dropped back to 1% with 120+ hours but I'm not worried given the est. times are way off


Mine just (at about the 58 minute mark) dropped back to 1% progress and memory usage is fluctuating wildly between 105 and 140 MB. Time elapsed is correct however (Boinc time matches CPU time on the work unit in Windows Task Manager)

David, is this drop to 1% and wildly fluctuating memory usage expected behaviour (or is that a dumb question)?

*** Join BOINC@Australia today ***
ID: 321 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile KWSN Sir Clark

Send message
Joined: 18 Sep 05
Posts: 46
Credit: 387,432
RAC: 0
Message 322 - Posted: 22 Sep 2005, 11:55:52 UTC - in response to Message 321.  

Mine's running at 99MB memory wise....for some reason, it started at 1.5hrs to completion and now has dropped back to 1% with 120+ hours but I'm not worried given the est. times are way off


Mine just (at about the 58 minute mark) dropped back to 1% progress and memory usage is fluctuating wildly between 105 and 140 MB. Time elapsed is correct however (Boinc time matches CPU time on the work unit in Windows Task Manager)

David, is this drop to 1% and wildly fluctuating memory usage expected behaviour (or is that a dumb question)?


He mentioned above that each WU is examining 12 structures....perhaps it drops back down to 0 after each one. It would make sense.
ID: 322 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile David E K
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 1 Jul 05
Posts: 1018
Credit: 4,334,829
RAC: 0
Message 325 - Posted: 22 Sep 2005, 14:53:52 UTC

The app should not drop back to 1%. I have to fix that. The fluctuating memory is normal.
ID: 325 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Webmaster Yoda
Avatar

Send message
Joined: 17 Sep 05
Posts: 161
Credit: 162,253
RAC: 0
Message 326 - Posted: 22 Sep 2005, 15:43:38 UTC - in response to Message 325.  

The app should not drop back to 1%. I have to fix that. The fluctuating memory is normal.


Thanks David, I thought as much. I haven't monitored it closely (and haven't crunched many work units yet) but I have seen it twice and both times memory use jumped sharply after that (from ~35MB to 100MB+). Perhaps that gives you a clue about the stage in the process where it happens.

I just saw it again. I think it was showing 75% just before it went back to 1%.
*** Join BOINC@Australia today ***
ID: 326 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile David E K
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 1 Jul 05
Posts: 1018
Credit: 4,334,829
RAC: 0
Message 327 - Posted: 22 Sep 2005, 15:48:02 UTC - in response to Message 326.  

The app should not drop back to 1%. I have to fix that. The fluctuating memory is normal.


Thanks David, I thought as much. I haven't monitored it closely (and haven't crunched many work units yet) but I have seen it twice and both times memory use jumped sharply after that (from ~35MB to 100MB+). Perhaps that gives you a clue about the stage in the process where it happens.

I just saw it again. I think it was showing 75% just before it went back to 1%.



The sudden increase to ~100MB and up is due to the way this current workunit is setup. It generates 10 low res models takes the best 2 scoring ones and then does hi res on them. The hi res takes more memory because it considers all atoms in the protein.
ID: 327 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Divide Overflow

Send message
Joined: 17 Sep 05
Posts: 82
Credit: 921,382
RAC: 0
Message 328 - Posted: 22 Sep 2005, 15:56:57 UTC - in response to Message 313.  


ID: 328 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
1 · 2 · 3 · Next

Message boards : Number crunching : Optimized Windows Application



©2024 University of Washington
https://www.bakerlab.org