Project: 6501 (Run 14, Clone 261, Gen 88) EUE after 43%

[GreyWhiskers]

An EUE on this classic 123 point Uniprocessor WU:
[12:03:31] Gromacs cannot continue further.
[12:03:31] Going to send back what have done.

Looking back at my WU history in HFM, I don't see any other EUEs on two different clients over the 12 P6501 WUs since July 2010.

Successful upload of partial results. Partial credit was awarded - 54 points out of 123.

Code: Select all

--- Opening Log file [January 12 13:17:00 UTC] 


# Windows CPU Systray Edition #################################################
###############################################################################

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: C:\Documents and Settings\Owner\Application Data\Folding@home-x86
Arguments: -verbosity 9 -advmethods 
...
[13:17:00] - Ask before connecting: No
[13:17:01] - User name: GreyWhiskers (Team 0)
[13:17:01] - User ID: 2F1E69777E4B5686


02:40:54] - Preparing to get new work unit...
[02:40:54] + Attempting to get work packet
[02:40:54] - Will indicate memory of 2047 MB
[02:40:54] - Connecting to assignment server
[02:40:54] Connecting to http://assign.stanford.edu:8080/
[02:40:55] Posted data.
[02:40:55] Initial: 40AB; - Successful: assigned to (171.64.65.62).
[02:40:55] + News From Folding@Home: Welcome to Folding@Home
[02:40:55] Loaded queue successfully.
[02:40:55] Connecting to http://171.64.65.62:8080/
[02:40:55] Posted data.
[02:40:55] Initial: 0000; - Receiving payload (expected size: 751053)
[02:40:56] - Downloaded at ~733 kB/s
[02:40:56] - Averaged speed for that direction ~716 kB/s
[02:40:56] + Received work.
[02:40:56] Trying to send all finished work units
[02:40:56] + No unsent completed units remaining.
[02:40:56] + Closed connections
[02:40:56] 

[02:40:56] + Processing work unit
[02:40:56] Core required: FahCore_78.exe
[02:40:56] Core found.
[02:40:56] Working on queue slot 05 [January 18 02:40:56 UTC]
[02:40:56] + Working ...
[02:40:56] - Calling '.\FahCore_78.exe -dir work/ -suffix 05 -nocpulock -checkpoint 15 -verbose -lifeline 5088 -version 623'

[02:40:57] 
[02:40:57] *------------------------------*
[02:40:57] Folding@Home Gromacs Core
[02:40:57] Version 1.90 (March 8, 2006)
[02:40:57] 
[02:40:57] Preparing to commence simulation
[02:40:57] - Looking at optimizations...
[02:40:57] - Created dyn
[02:40:57] - Files status OK
[02:41:05] - Expanded 750541 -> 3748137 (decompressed 499.3 percent)
[02:41:05] - Starting from initial work packet
[02:41:05] 
[02:41:05] Project: 6501 (Run 14, Clone 261, Gen 88)
[02:41:05] 
[02:41:07] Assembly optimizations on if available.
[02:41:07] Entering M.D.
[02:41:14] Protein: UBIQUITIN MODEL3750 in water
[02:41:14] 
[02:41:15] Writing local files
[02:41:19] Extra SSE boost OK.
[02:41:20] Writing local files
[02:41:20] Completed 0 out of 250000 steps  (0%)
....
[09:37:54] Completed 100000 out of 250000 steps  (40%)
[09:52:55] Timered checkpoint triggered.
[10:07:55] Timered checkpoint triggered.
[10:18:56] Writing local files
[10:18:57] Completed 102500 out of 250000 steps  (41%)
[10:33:57] Timered checkpoint triggered.
[10:48:57] Timered checkpoint triggered.
[10:59:46] Writing local files
[10:59:46] Completed 105000 out of 250000 steps  (42%)
[11:14:46] Timered checkpoint triggered.
[11:29:49] Timered checkpoint triggered.
[11:41:45] Writing local files
[11:41:46] Completed 107500 out of 250000 steps  (43%)
[11:56:46] Timered checkpoint triggered.
[12:03:31] Gromacs cannot continue further.
[12:03:31] Going to send back what have done.
[12:03:31] logfile size: 28212
[12:03:31] - Writing 28748 bytes of core data to disk...
[12:03:31] Done: 28236 -> 5458 (compressed to 19.3 percent)
[12:03:31]   ... Done.
[12:03:32] 
[12:03:32] Folding@home Core Shutdown: EARLY_UNIT_END
[12:03:35] CoreStatus = 72 (114)
[12:03:35] Sending work to server
[12:03:35] Project: 6501 (Run 14, Clone 261, Gen 88)
[12:03:35] - Read packet limit of 540015616... Set to 524286976.


[12:03:35] + Attempting to send results [January 19 12:03:35 UTC]
[12:03:35] - Reading file work/wuresults_05.dat from core
[12:03:35]   (Read 5970 bytes from disk)
[12:03:35] Connecting to http://171.64.65.62:8080/
[12:03:35] Posted data.
[12:03:36] Initial: 0000; - Uploaded at ~6 kB/s
[12:03:36] - Averaged speed for that direction ~211 kB/s
[12:03:36] + Results successfully sent
[12:03:36] Thank you for your contribution to Folding@Home.

[sortofageek]

Three folders had EUs with Project: 6501 (Run 14, Clone 261, Gen 88), but one folder completed it successfully. All I can determine about the successful folder is an Athlon processor and Windows OS.

[GreyWhiskers]

Thanks.

My old rig is a single core HP Pentium 4/HT 3.2 GHz with an HIS AGP ATI 4670 GPU card. Don't know what the difference would be with an Athlon. This project is a classic project running on a Pent 4 HT with SSE enabled.

This may be a little off topic, but I actually run two classic uniprocessor clients in addition to the GPU client to soak up all the processing cycles avail. When I'm not otherwise using the rig, the GPU gets about 1350 PPD, and the two CPU clients get about 90-95 ppd between them. Running one classic CPU client was netting me about 75-80 ppd, so I'm finding a few spare cycles to squeeze out for the second client. I still finish the assigned classic CPU WUs with a 91-92% perf factor for each of the two classic clients. BTW, it's obvious that even with the GPU environment variables set per instructions, the GPU client takes about 25-30% of a thread. When the GPU was down for a while awaiting the Stanford server to come back up to assign a new work unit, the two CPU clients "leaped" up to ~120 ppd between them.

[bruce]

Running two classic clients in a single core machine is actively discouraged by Vijay because it inhibits the science. It does give a few more PPD, so some people do it anyway. This has lead to many very long discussions, most of which get locked because nobody listens to any point of view other than their own. (...so don't start another discussion.)