New Protomol Core

This forum is for discussions about the Motherboards.org Folding team. What is folding? Venture on in for a look.

Moderator: The Mod Squad

New Protomol Core

Postby Pette Broad » Tue Dec 22, 2009 4:46 pm

This has been testing in Beta and has had many, many problems. For some reason the guy from Stanford in charge of these units has released them to advmethods. So if you use the -advmethods you will get one of these..

BEWARE......They have a nasty habit of getting into a loop of death so if you get a bad one here are the steps to get rid of it. First thing you will notice is that it crashes out onto the desktop in this case go into the work folder and delete everything, then remove the -advmethods flag and restart folding. Don't just try to delete using the -delete xx flag, you will get sent the same bad unit again and again.

I would advise everyone to remove the -advmethods flag ASAP especially if you have any machines that you can't monitor.

Pete
Image
Pette Broad
Black Belt 5th Degree
Black Belt 5th Degree
 
Posts: 5487
Joined: Tue Jul 10, 2001 12:01 am
Location: Flintshire, U.K

Postby MrG » Thu Dec 24, 2009 9:31 am

Thanks Pete.

Do you have a link to the work unit details i.e. the PRCG details we need to look out for?
MrG
Brown Belt
Brown Belt
 
Posts: 335
Joined: Sun Dec 14, 2008 2:38 pm

Postby Pette Broad » Thu Dec 24, 2009 5:35 pm

The only info I can give you is that these are numbered Project 10000 and 10001. Since I posted the original message a new Protomol core has been released which is supposed to fix the problem with bad units and I've had no problems recently.

These are small units worth 50 points (10000) and 84 points(10001). These are also very advanced units, capable of using sse3 and sse4. However, at the moment only Intel CPU's can take advantage of this due to the way the code is written and certain issues between AMD and Intel. Having said that, there is not a huge ppd difference as AMD CPU's can use sse2 which is the main factor.

At some stage there's talk of GPU also using a variation of Protomol, I guess we'll just have to wait and see. There are also more new cores in the pipeline, I think one of them will replace Gromacs eventually.

Pete
Image
Pette Broad
Black Belt 5th Degree
Black Belt 5th Degree
 
Posts: 5487
Joined: Tue Jul 10, 2001 12:01 am
Location: Flintshire, U.K

Postby Ripshod » Fri Dec 25, 2009 4:01 am

Folding one now. P10000 (R704,C0,G7). Thanks for the heads up, will babysit this one and report back.
CALL &BD19 (If you need to ask you don't need to know)
Image
Ripshod
Black Belt
Black Belt
 
Posts: 708
Joined: Wed Aug 13, 2003 2:57 pm
Location: Yorkshire, UK

Postby Ripshod » Sat Dec 26, 2009 1:39 am

Completed 100%, this was a 10000 worth 84.48 points.

According to Stanford:
p10000 : 544 atoms, 84.48 points, preferred deadline 3.07 days, final deadline 23.04 days. This project uses conventional simulation methods.
p10001 : 544 atoms, 50.56 points, preferred deadline 1.84 days, final deadline 13.79 days. This project uses the NML simulation method.


Maybe the 10001 may give problems?
CALL &BD19 (If you need to ask you don't need to know)
Image
Ripshod
Black Belt
Black Belt
 
Posts: 708
Joined: Wed Aug 13, 2003 2:57 pm
Location: Yorkshire, UK

Postby whiteboy » Sun Dec 27, 2009 7:22 am

Thanks for the headups Pete. Just looked on my system and saw that I had one download.

[12:59:17] *********************** Log Started 25/Dec/2009 12:59:17 ***********************
[12:59:17] ************************** ProtoMol Folding@Home Core **************************
[12:59:17] Version: 21
[12:59:17] Type: 180
[12:59:17] Core: ProtoMol
[12:59:17] Website: http://folding.stanford.edu/
[12:59:17] Copyright: (c) 2009 Stanford University
[12:59:17] Author: Joseph Coffland <joseph@cauldrondevelopment.com>
[12:59:17] Args: -dir work/ -suffix 02 -checkpoint 15 -service -lifeline 2028 -version
[12:59:17] 620
[12:59:17] ************************************ Build *************************************
[12:59:17] Date: Dec 24 2009
[12:59:17] Time: 14:36:31
[12:59:17] Revision: 1748
[12:59:17] Compiler: Intel(R) C++ MSVC 1500 mode 1110
[12:59:17] Options: /TP /nologo /EHsc /wd4297 /wd4103 /wd1786 /arch:IA32 /Ox
[12:59:17] /QaxSSE2,SSE3,SSSE3,SSE4.1,SSE4.2 /Qrestrict /MT
[12:59:17] Platform: Windows XP
[12:59:17] Bits: 32
[12:59:17] ************************************ System ************************************
[12:59:17] OS: Microsoft Windows Vista Home Premium
[12:59:17] CPU: AMD Phenom(tm) 9500 Quad-Core Processor
[12:59:17] CPU ID: AuthenticAMD Family 16 Model 2 Stepping 2
[12:59:17] CPUs: 4 Logical, 4 Physical
[12:59:17] Memory: 4.00 GB
[12:59:17] ********************************************************************************
[12:59:17] Project: 10001 (Run 117, Clone 3, Gen 10)
[12:59:17] Reading tar file par_all27_prot_lipid.inp
[12:59:17] Reading tar file scpismQuartic.inp
[12:59:17] Reading tar file ww_exteq_nowater1.pdb
[12:59:17] Reading tar file ww_exteq_nowater1.psf
[12:59:17] Reading tar file checkpt
[12:59:17] Reading tar file ww_exteq_nowater1.298.pos
[12:59:17] Reading tar file ww_exteq_nowater1.298.vel
[12:59:17] Reading tar file protomol.conf
[12:59:17] Reading tar file core.xml
[12:59:17] Digital signatures verified
[12:59:17] Completed 0 out of 200000 steps (0%)
[13:04:29] Completed 2000 out of 200000 steps (1%)
[13:09:58] Completed 4000 out of 200000 steps (2%)....
.....
[21:10:28] Completed 172000 out of 200000 steps (86%)
[21:13:56] WARNING: UnexpectedExitHandler triggered
[21:13:56] WARNING: Unexpected exit from science code
[21:13:56] Saving result file logfile_02.txt
[21:13:56] Saving result file checkpt
[21:13:56] Saving result file checkpt.crc
[21:13:56] Saving result file log.txt
[21:13:56] Saving result file protomol.conf
[21:13:56] Saving result file ww.dcd
[21:13:58] Saving result file ww_exteq_nowater1.330.pos
[21:13:58] Saving result file ww_exteq_nowater1.330.vel
[21:13:58] Folding@home Core Shutdown: UNKNOWN
[21:14:03] CoreStatus = 7B (123)
[21:14:03] Client-core communications error: ERROR 0x7b
[21:14:03] This is a sign of more serious problems, shutting down.


looks like 1001 hung up on mine, and since I don't monitor my system everyday I failed to catch this when this download 2 days ago. time to wipe it out and start over I guess.
Amd Phenom 9500 Quad on a ECS a770m mobo
4 gigs ram
ATI Radeon x1950pro w/1 gig ram video card
2 250 gig samsung sata HDs
2 lite on sata dvd burners
all in a raidmax blue neon case running Vista64
total cost under $600 built march 2008
whiteboy
Anti-Static Strap
Anti-Static Strap
 
Posts: 437
Joined: Wed Jan 15, 2003 8:36 pm
Location: Lehigh Acres, Fl

Postby Pette Broad » Sun Dec 27, 2009 6:20 pm

Yeah, you need to get rid of it. The new core looks more promising (although as I write this there are no protomol units available). You may need to delete the old core because the guy looking after this project is a protomol expert but a novice on how to setup the server correctly.

Pete
Image
Pette Broad
Black Belt 5th Degree
Black Belt 5th Degree
 
Posts: 5487
Joined: Tue Jul 10, 2001 12:01 am
Location: Flintshire, U.K

Postby Ripshod » Mon Dec 28, 2009 8:10 am

Just received 10001 (R502,C2,G9).
Will let you know if this one fails (if I don't post the unit completed).
CALL &BD19 (If you need to ask you don't need to know)
Image
Ripshod
Black Belt
Black Belt
 
Posts: 708
Joined: Wed Aug 13, 2003 2:57 pm
Location: Yorkshire, UK

Postby Ripshod » Wed Dec 30, 2009 5:26 am

I don't understand what happened with this one but, apparently, the unit was finished, without error, at 52%
[09:04:30] Completed 84000 out of 200000 steps (42%)
[09:10:10] Completed 86000 out of 200000 steps (43%)
[09:16:32] Completed 88000 out of 200000 steps (44%)
[09:22:23] Completed 90000 out of 200000 steps (45%)
[09:28:35] Completed 92000 out of 200000 steps (46%)
[09:34:59] Completed 94000 out of 200000 steps (47%)
[09:41:14] Completed 96000 out of 200000 steps (48%)
[09:48:56] Completed 98000 out of 200000 steps (49%)
[09:55:00] Completed 100000 out of 200000 steps (50%)
[10:00:07] Completed 102000 out of 200000 steps (51%)
[10:06:00] Completed 104000 out of 200000 steps (52%)
[10:08:23] Saving result file logfile_02.txt
[10:08:23] Saving result file checkpt
[10:08:23] Saving result file checkpt.crc
[10:08:23] Saving result file log.txt
[10:08:24] Saving result file protomol.conf
[10:08:24] Saving result file ww.dcd
[10:08:24] Saving result file ww_exteq_nowater1.621.pos
[10:08:24] Saving result file ww_exteq_nowater1.621.vel
[10:08:25] Folding@home Core Shutdown: FINISHED_UNIT
[10:08:28] CoreStatus = 64 (100)
[10:08:28] Sending work to server
[10:08:28] Project: 10001 (Run 451, Clone 2, Gen 13)


[10:08:28] + Attempting to send results [December 30 10:08:28 UTC]
[10:09:05] + Results successfully sent
[10:09:05] Thank you for your contribution to Folding@Home.
[10:09:05] + Number of Units Completed: 2

[10:09:12] - Preparing to get new work unit...


Far from being a loop of death, more an early bath.
CALL &BD19 (If you need to ask you don't need to know)
Image
Ripshod
Black Belt
Black Belt
 
Posts: 708
Joined: Wed Aug 13, 2003 2:57 pm
Location: Yorkshire, UK

Postby Karlsweldt » Wed Dec 30, 2009 12:50 pm

Have had a few work units terminate early. Some Beta-testing before full release is done, after lab trials.. to ensure consistent results while in the field.
F@H.. to solve mankind's maladies.. in our lifetimes!
Karlsweldt
Mobo-fu Master
Mobo-fu Master
 
Posts: 19408
Joined: Wed Nov 12, 2003 11:57 am
Location: 07438

Next

Return to Motherboards.org Folding Team

Who is online

Users browsing this forum: No registered users and 1 guest