36 hour work unit. normal?

[jebusv20]

Was just looking for confirmation as to whether or not the below is normal.
I am currently running a AMD PII 955 @ 3.6, this is my second work unit (the first was using the non-smp client and was ALOT faster) im just woundering if there is a possibility of a huge work unit or not. like 36 hours is a long time.
I am currently running the GPU3 on the side. Im not 100% sure of switches, because im afraid to close the client to check and look 8 hours work. I ran this over night so the system was otherwise idle during this time.

[PantherX]

Welcome to the F@H Forum jebusv20,

P6040 is among a few large SMP Projects. The Client is configured with -smp flag and I would advice that you also add -verbosity 9 which will help us during troubleshooting.

[jebusv20]

sorry, yeah. im EXTREMELY new to this, um... can I close the client and reopen it with the -verbosity 9 flag without loosing the current state of the work unit?

[jebusv20]

I think I might be running -bigadv

considering I am only running a quad, would that be bad?

[jebusv20]

Just ran it again with the -verbosity 9 and -smp switches
was not running -bigadv, I am running advanced methods.
Just looking for a confirmation that my computer isnt 'spinning its wheels'
I was having issues before the SMP client with my cpu sitting at 40%
and now a WU thats taking almost 1 1/2 days to complete?

Code: Select all

# Windows SMP Console Edition #################################################
###############################################################################

                       Folding@Home Client Version 6.34

                          http://folding.stanford.edu

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

Launch directory: C:\Users\Jebusv20\FAH
Executable: fah6
Arguments: -verbosity 9 -smp 

[21:53:07] - Ask before connecting: No
[21:53:07] - User name: jebusv20 (Team 202863)
[21:53:07] - User ID: 4203962853B5E805
[21:53:07] - Machine ID: 1
[21:53:07] 
[21:53:07] Loaded queue successfully.
[21:53:07] 
[21:53:07] - Autosending finished units... [March 15 21:53:07 UTC]
[21:53:07] Trying to send all finished work units
[21:53:07] + Processing work unit
[21:53:07] + No unsent completed units remaining.
[21:53:07] - Autosend completed
[21:53:07] Core required: FahCore_a3.exe
[21:53:07] Core found.
[21:53:07] Working on queue slot 01 [March 15 21:53:07 UTC]
[21:53:07] + Working ...
[21:53:07] - Calling '.\FahCore_a3.exe -dir work/ -nice 19 -suffix 01 -np 4 -priority 96 -checkpoint 15 -verbose -lifeline 5852 -version 634'

[21:53:07] 
[21:53:07] *------------------------------*
[21:53:07] Folding@Home Gromacs SMP Core
[21:53:07] Version 2.27 (Dec. 15, 2010)
[21:53:07] 
[21:53:07] Preparing to commence simulation
[21:53:07] - Ensuring status. Please wait.
[21:53:17] - Looking at optimizations...
[21:53:17] - Working with standard loops on this execution.
[21:53:17] - Previous termination of core was improper.
[21:53:17] - Going to use standard loops.
[21:53:17] - Files status OK
[21:53:18] - Expanded 7882938 -> 10126021 (decompressed 128.4 percent)
[21:53:18] Called DecompressByteArray: compressed_data_size=7882938 data_size=10126021, decompressed_data_size=10126021 diff=0
[21:53:18] - Digital signature verified
[21:53:18] 
[21:53:18] Project: 6040 (Run 0, Clone 108, Gen 129)
[21:53:18] 
[21:53:18] Entering M.D.
[21:53:24] Using Gromacs checkpoints
[21:53:25] Mapping NT from 4 to 4 
[21:53:26] Resuming from checkpoint
[21:53:26] Verified work/wudata_01.log
[21:53:26] Verified work/wudata_01.trr
[21:53:26] Verified work/wudata_01.xtc
[21:53:26] Verified work/wudata_01.edr
[21:53:27] Completed 67156 out of 250000 steps  (26%)

[uncle fuzzy]

What is the cpu load now? Everything looks OK in the logs. I'm not near my box that runs SMP, so I can't give you a speed check.

[jebusv20]

cpu load is at 100% and has been at 100% for the past 10 hours now.

[uncle fuzzy]

I'd say your speed is right. I'll check my box later to verify.

[Pick2]

That's about normal time for a p6040 or p6041. They are rather large. A p6040 took 1.56 Days on My AMD Quad core.

[jebusv20]

can I ask how many points it might be worth?
its not 1 is it

[topodisc]

Using the folding calculator here -> http://www.linuxforge.net/bonuscalc2.php

It's about 7K points, or 4600 PPD.

[bruce]

Using the folding calculator here -> http://www.linuxforge.net/bonuscalc2.php

It's about 7K points, or 4600 PPD.

Topodisc is assuming you're an old-timer. You said this was your second WU so I'll explain further.

SMP WUs are generally well suited for quads like yours which run 24x7. If you fold part-time, smp may or may not be right for you.

Project 6040 is worth 1395.00 points on completion. it is also eligible for a bonus, depending on whether you're using a passkey that has been qualified or not. With the bonus, it would be worth about 7000 points if it is completed at that speed. Without the bonus, you'll get the 1395 points upon completion.

If you don't have a passkey, get one immediately and start work on completing your first 10 bonus-eligible WUs. You'll start getting the bonus on the eleventh WU.

[Baowoulf]

Code: Select all

[21:53:17] - Looking at optimizations...
[21:53:17] - Working with standard loops on this execution.
[21:53:17] - Previous termination of core was improper.
[21:53:17] - Going to use standard loops.

Usually you get this message if you didn't close your client correctly. Close it with "Ctrl+C" I believe, Panther or Bruce can correct me if I'm wrong, to close it properly and restart it.

[HendricksSA]

You may see this warning whenever you close your SMP client. Ctrl-C is the right way to close it and you can ignore it when you are working on SMP work units. The optimization is built into the SMP client.

[gwildperson]

Ctrl-C is the right way to close any client, including SMP. If it's shut down normally you won't get that message Optimizations are always on, so you can ignore the message as is a leftover from an earlier time. It's appropriate to have some concern, though, because when you fail to shut down normally, the chance of encountering a corrupt checkpoint is increased. Though it's unlikely, you really don't want to have the client restart the WU from the beginning and discard the work you've done on it