Confused on points

MissilemanMissileman Orlando, Florida Icrontian
edited May 2004 in Folding@Home
I read the things about the WU and how they compute points and all has been as expected until tonight. I had a WU complete on one of my remote systems and it scored 4 points. Any idea how that happened. I didn't think they went that low.


You guys were right. It's addicting. I check the stats like 20 times a day now.

Maybe I should stop while I still can :rolleyes:

Comments

  • MissilemanMissileman Orlando, Florida Icrontian
    edited May 2004
    I think I found it. The Barton system said it terminated one early on a Gro WU and sent it back infinished. Log doesn't make a whole lot of sense to me since this machine is as stable as a rock. Has done a couple of Tinker WU's (1 for 236 points) and now it says it is unstable. Could it be just a bad fold or is it a bad machine? It's already almost done with another Gro WU.

    I was gonna change the power supply out in that box tomorrow. The 3.3 is down to 3.0 in it. I guess I'll just watch it for a while. Log excerpt is below :


    [20:28:09]
    [20:28:09] *
    *
    [20:28:09] Folding@home Gromacs Core
    [20:28:09] Version 1.65 (May 6, 2004)
    [20:28:09]
    [20:28:09] Preparing to commence simulation
    [20:28:09] - Assembly optimizations manually forced on.
    [20:28:09] - Not checking prior termination.
    [20:28:10] - Expanded 352991 -> 1761809 (decompressed 499.1 percent)
    [20:28:10]
    [20:28:10] Project: 521 (Run 76, Clone 39, Gen 3)
    [20:28:10]
    [20:28:10] Assembly optimizations on if available.
    [20:28:10] Entering M.D.
    [20:28:30] (Starting from checkpoint)
    [20:28:30] Protein: p521_BBA5_pf
    [20:28:30]
    [20:28:30] Writing local files
    [20:28:31] Extra SSE boost OK.
    [20:28:31] Writing local files
    [20:28:31] Completed 0 out of 500000 steps (0)
    [20:34:20] Writing local files
    [20:34:20] Completed 5000 out of 500000 steps (1)
    [20:40:07] Writing local files
    [20:40:07] Completed 10000 out of 500000 steps (2)
    [20:45:54] Writing local files
    [20:45:54] Completed 15000 out of 500000 steps (3)
    [20:51:44] Writing local files
    [20:51:44] Completed 20000 out of 500000 steps (4)
    [20:57:34] Writing local files
    [20:57:34] Completed 25000 out of 500000 steps (5)
    [21:03:24] Writing local files
    [21:03:24] Completed 30000 out of 500000 steps (6)
    [21:09:15] Writing local files
    [21:09:15] Completed 35000 out of 500000 steps (7)
    [21:15:06] Writing local files
    [21:15:06] Completed 40000 out of 500000 steps (8)
    [21:20:58] Writing local files
    [21:20:58] Completed 45000 out of 500000 steps (9)
    [21:23:45] Quit 101 - Fatal error:
    [21:23:45] Step 47359, time 94.718 (ps) LINCS WARNING
    [21:23:45] relative constraint deviation after LINCS:
    [21:23:45] max 125120.875000 (between atoms 196 and 198) rms 9082.249023
    [21:23:45]
    [21:23:45] Simulation instability has been encountered. The run has entered a
    [21:23:45] state from which no further progress can be made.
    [21:23:45] If you often see other project units terminating early like this
    [21:23:45] too, you may wish to check the stability of your computer (issues
    [21:23:45] such as high temperature, overclocking, etc.).
    [21:23:45] Going to send back what have done.
    [21:23:45] logfile size: 34628
    [21:23:45] - Writing 35316 bytes of core data to disk...
    [21:23:45] ... Done.
    [21:23:45]
    [21:23:45] Folding@home Core Shutdown: EARLY_UNIT_END
    [21:23:48] CoreStatus = 72 (114)
    [21:23:48] Sending work to server


    [21:23:48] + Attempting to send results
    [21:23:48] - Reading file work/wuresults_03.dat from core
    [21:23:48] (Read 35316 bytes from disk)
    [21:23:50] - Uploaded at ~17 kB/s
    [21:23:50] - Averaged speed for that direction ~29 kB/s
    [21:23:50] + Results successfully sent
    [21:23:50] Thank you for your contribution to Folding@home.
  • LeonardoLeonardo Wake up and smell the glaciers Eagle River, Alaska Icrontian
    edited May 2004
    Four points? You are correct - there are no work units worth that little. Check the log in the folding client folder on that remote machine when you get a chance. It's possible that the unit was only partially completed, or that the results were only partially readable to the Stanford server. The system does allow for partial points awarded for incomplete/improperly completed work units.
  • mmonninmmonnin Centreville, VA
    edited May 2004
    You got partial credit for the work you did.

    Is it OCed? Folding will make an otherwise stable system, unstable. At least folding wont finish right but everything else runs and works perfectly.
  • qparadoxqparadox Vancouver, BC
    edited May 2004
    The gromacs are also the only units that currently use SSE optimization so if you had all tinkers before it might have been the SSE or maybe a hot day that pushed you over the edge. 90%+ chance this is an overclocking or heat problem.
  • mmonninmmonnin Centreville, VA
    edited May 2004
    Some WUs just are bad and its not your fault at all. It does look like it was an OCing error tho.
  • csimoncsimon Acadiana Icrontian
    edited May 2004
    you may have gotten only 4 points for a partially finished WU ....partial points for partial WU.

    What were the load temps when that WU went bellyup onya missileman?

    My cpu and sys get about 3-4 degrees warmer after noon which is just about enough to cause issues.
Sign In or Register to comment.