5101 run 0 clone 76 gen 105 running very slow

[dschief]

getting 33 min frame times, this rig { C2D E6400 W 2 G ram } usually sees 12 min frame times for SMP Wu's

Code: Select all

08:47:17] Project: 5101 (Run 0, Clone 76, Gen 105)
[08:47:17] 
[08:47:17] Entering M.D.
NNODES=4, MYRANK=0, HOSTNAME=Grendel
NNODES=4, MYRANK=1, HOSTNAME=Grendel
NNODES=4, MYRANK=2, HOSTNAME=Grendel
NNODES=4, MYRANK=3, HOSTNAME=Grendel
      Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
       Copyright (c) 1991-2000, University of Groningen, The Netherlands.
             Copyright (c) 2001-2004, The GROMACS development team,
            check out http://www.gromacs.org for more information.

        This inclusion of Gromacs code in the Folding@Home Core is under
        a special license (see http://folding.stanford.edu/gromacs.html)
         specially granted to Stanford by the copyright holders. If you
          are interested in using Gromacs, visit www.gromacs.org where
                you can download a free version of Gromacs under
         the terms of the GNU General Public License (GPL) as published
       by the Free Software Foundation; either version 2 of the License,
                     or (at your option) any later version.

NODEID=0 argc=15
NODEID=1 argc=15
NODEID=2 argc=15
NODEID=3 argc=15
starting mdrun 'CMF in water'
2000000 steps,   4000.0 ps.

[08:47:23] Extra SSE boost OK.
[08:47:23] r
[08:47:23] Writing local files
[08:47:23] Extra SSE boost OK.
[08:47:23] 0000 steps  (0 percent)
[09:02:23] Timered checkpoint triggered.
[09:17:23] Timered checkpoint triggered.
[09:20:25] Writing local files
[09:20:25] Completed 20000 out of 2000000 steps  (1 percent)
[09:35:24] Timered checkpoint triggered.
[09:50:24] Timered checkpoint triggered.
[09:53:27] Writing local files
[09:53:27] Completed 40000 out of 2000000 steps  (2 percent)
[09:59:16] - Autosending finished units... [August 12 09:59:16 UTC]
[09:59:16] Trying to send all finished work units
[09:59:16] + No unsent completed units remaining.
[09:59:16] - Autosend completed
[10:08:27] Timered checkpoint triggered.
[10:23:27] Timered checkpoint triggered.
[10:26:32] Writing local files
[10:26:32] Completed 60000 out of 2000000 steps  (3 percent)
[10:41:32] Timered checkpoint triggered.
[10:56:32] Timered checkpoint triggered.
[10:59:29] Writing local files
[10:59:29] Completed 80000 out of 2000000 steps  (4 percent)
[11:14:29] Timered checkpoint triggered.
[11:29:29] Timered checkpoint triggered.
[11:32:21] Writing local files
[11:32:21] Completed 100000 out of 2000000 steps  (5 percent)
[11:47:21] Timered checkpoint triggered.
[12:02:21] Timered checkpoint triggered.
[12:05:14] Writing local files
[12:05:14] Completed 120000 out of 2000000 steps  (6 percent)
[12:20:13] Timered checkpoint triggered.
[12:35:13] Timered checkpoint triggered.
[12:38:06] Writing local files
[12:38:06] Completed 140000 out of 2000000 steps  (7 percent)
[12:53:06] Timered checkpoint triggered.
[13:08:06] Timered checkpoint triggered.
[13:10:51] Writing local files
[13:10:52] Completed 160000 out of 2000000 steps  (8 percent)
[13:25:51] Timered checkpoint triggered.
[13:40:51] Timered checkpoint triggered.
[13:43:30] Writing local files
[13:43:30] Completed 180000 out of 2000000 steps  (9 percent)
[13:58:30] Timered checkpoint triggered.
[14:13:30] Timered checkpoint triggered.
[14:15:56] Writing local files
[14:15:56] Completed 200000 out of 2000000 steps  (10 percent)
[14:30:56] Timered checkpoint triggered.
[14:45:56] Timered checkpoint triggered.
[14:48:11] Writing local files
[14:48:11] Completed 220000 out of 2000000 steps  (11 percent)
[15:03:11] Timered checkpoint triggered.
[15:18:11] Timered checkpoint triggered.
[15:20:45] Writing local files
[15:20:45] Completed 240000 out of 2000000 steps  (12 percent)

are these times reasonable for this wu?

[BrokenWolf]

If you have been getting the FahCore_a2.exe WUs' then that is why this one seems so slow. This is an FahCore_a1.exe WU and the A1 core is not as efficient as the A2 core.

Most of the A2 WUs' are in the 2669-2677 range recently. I have many Linux SMP folding rigs and right now I have 5 of them folding an A1 WU. The normal PPD I would get on one of my systems with P2677 WU is 1580. It is working on a P5102 (a1) and it is getting 845 PPD. Downloaded 7 hours ago, 3d 16h 14mn until completed. This is on a dual core system so it really hurts these puppies.

Hope this helps explain the slowness.

BrokenWolf

[dschief]

My bad, it's been so long since I had a A1, it did'nt dawn on me to check

[parkut]

Yours seems reasonable. Two examples from my log files - CentOS 5.3 x64 Linux SMP

Code: Select all

[05:56:27] Project: 5101 (Run 0, Clone 98, Gen 103)
...
model name	: Intel(R) Core(TM)2 Duo CPU     E6750  @ 2.66GHz
cpu MHz		: 2664.000
cache size	: 4096 KB
Memory: 1.94 GB physical, 1.94 GB virtual
...
[07:03:18] Completed 60000 out of 2000000 steps  (3 percent)
[06:41:04] Completed 40000 out of 2000000 steps  (2 percent)
...
22 minutes, 14 seconds / frame



[07:07:03] Project: 5101 (Run 0, Clone 198, Gen 79)
...
model name	: Intel(R) Core(TM)2 Duo CPU     E6550  @ 2.33GHz
cpu MHz		: 2333.000
cache size	: 4096 KB
Memory: 1.93 GB physical, 1.94 GB virtual
...
[13:59:08] Completed 40000 out of 2000000 steps  (2 percent)
[13:33:58] Completed 20000 out of 2000000 steps  (1 percent)
...
25 minutes, 10 seconds / frame