Project: 2677 (Run 22, Clone 43, Gen 27) -- only one core

[valleton]

Ultra-low PPD triggered my attention. As I understand I'm not the only one with this type of new(?) anomalies: On Linux SMP only one of the processes is doing something, others are using 0% of the CPU.

Normal frame time ~10mins

Code: Select all

[07:50:42] - Calling './mpiexec -np 4 -host 127.0.0.1 ./FahCore_a2.exe -dir work/ -suffix 02 -checkpoint 15 -forceasm -verbose -lifeline 23753 -version 624'

[07:50:42] 
[07:50:42] *------------------------------*
[07:50:42] Folding@Home Gromacs SMP Core
[07:50:42] Version 2.08 (Mon May 18 14:47:42 PDT 2009)
[07:50:42] 
[07:50:42] Preparing to commence simulation
[07:50:42] - Ensuring status. Please wait.
[07:50:52] - Assembly optimizations manually forced on.
[07:50:52] - Not checking prior termination.
[07:50:54] - Expanded 1504335 -> 24019405 (decompressed 1596.6 percent)
[07:50:54] Called DecompressByteArray: compressed_data_size=1504335 data_size=24019405, decompressed_data_size=24019405 diff=0
[07:50:54] - Digital signature verified
[07:50:54] 
[07:50:54] Project: 2677 (Run 22, Clone 43, Gen 27)
[07:50:54] 
[07:50:54] Assembly optimizations on if available.
[07:50:54] Entering M.D.
[07:51:14] Completed 0 out of 250000 steps  (0%)
[08:44:37] Completed 2500 out of 250000 steps  (1%)
[09:37:56] Completed 5000 out of 250000 steps  (2%)
[10:31:15] Completed 7500 out of 250000 steps  (3%)
[11:24:35] Completed 10000 out of 250000 steps  (4%)
[11:26:39] - Autosending finished units... [August 14 11:26:39 UTC]
[11:26:39] Trying to send all finished work units
[11:26:39] + No unsent completed units remaining.
[11:26:39] - Autosend completed
[12:17:54] Completed 12500 out of 250000 steps  (5%)
[13:11:14] Completed 15000 out of 250000 steps  (6%)
[14:04:32] Completed 17500 out of 250000 steps  (7%)
[14:57:52] Completed 20000 out of 250000 steps  (8%)
[15:51:12] Completed 22500 out of 250000 steps  (9%)
[16:44:31] Completed 25000 out of 250000 steps  (10%)
[17:26:39] - Autosending finished units... [August 14 17:26:39 UTC]
[17:26:39] Trying to send all finished work units
[17:26:39] + No unsent completed units remaining.
[17:26:39] - Autosend completed
[17:37:50] Completed 27500 out of 250000 steps  (11%)
[18:31:08] Completed 30000 out of 250000 steps  (12%)
[19:24:28] Completed 32500 out of 250000 steps  (13%)
[20:17:45] Completed 35000 out of 250000 steps  (14%)
[21:11:05] Completed 37500 out of 250000 steps  (15%)
[22:04:24] Completed 40000 out of 250000 steps  (16%)
[22:58:10] Completed 42500 out of 250000 steps  (17%)
[23:26:39] - Autosending finished units... [August 14 23:26:39 UTC]
[23:26:39] Trying to send all finished work units
[23:26:39] + No unsent completed units remaining.
[23:26:39] - Autosend completed
[23:52:03] Completed 45000 out of 250000 steps  (18%)

Edited title to add "only one core" -- susato

[MtM]

Yup this has been reported and they are aware of it. There isn't a single thread but a few scatterd around, I think they might got enough info on the bug to fix it and will likely only reply when the fix is there but that's just my assumption.

Thanks for reporting this, even if it's a known issue!

[parkut]

compressed_data_size=1504335

another 1.5 meg download. See a pattern here?

[BrokenWolf]

Welcome to the club.