• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Linux SMP Client using only 80%

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

muhon_zen

Member
Joined
Apr 24, 2003
Location
Brazil
Dear Folders!

Hello! I'm right now folding in my Dell Studio XPS 13. It folds good in Linux, using the SMP client. However, I stopped for a couple of days, then I didn't reach the deadline. Since it was late and I didn't know how to make the client download the new project, I kinda deleted the log files, the fah_cores and another bunch of files till it worked. It used to download 1800 point projects, but now it downloaded a 4000 point project.

Well, the problem is that now it seems to use only 80% of both cores, instead of the usual 100%. I'm measuring both in System Monitor and using gkrellm.

BTW, I'm using Ubuntu 9.04, x64 version.

Can anyone help me with this issue please?

EDIT: Removed the Screenshot.
 
Last edited:

ChasR

Senior Member
Joined
Apr 12, 2004
Location
Atlanta
Instead of huge screen shots, I suggest you copy and paste the log, wrapped in code tags particularly at the last client start before the issue occurred. As it is, you've missed the important stuff, like the start up arguments and the project number. There aren't any 1800 or 4000 point projects so it's a bit difficult to tell what happened, but I suspect you deleted client.cfg and either skipped config or forgot the -smp flag. It would be very helpful to see the last part of the log with the last restart and wouldn't hurt a thing if you deleted the screenshot while you were at it.

Another strong possibility is you downloaded an a1 core WU, which uses less resources than a2 core WUs. In that case, nothing is wrong at all.
 

Bobnova

Senior Member
Joined
May 10, 2009
I got an A1 SMP, used core 1.78 or somesuch, used ~85-90% of the cpu and was 200ppd short of an a2 2.10 core unit.
 
OP
muhon_zen

muhon_zen

Member
Joined
Apr 24, 2003
Location
Brazil
Sorry about the huge screenshot, It was really late and I thought it would actually generate a thumbnail. Anyways, I'll add the log tonight. I kinda remember doing the config, since it is still giving points to my user. HFM shows SMP-Gromacs so I guess it is working in SMP mode.

I'll keep you guys posted later. Thanks for replying.
 
OP
muhon_zen

muhon_zen

Member
Joined
Apr 24, 2003
Location
Brazil
Here's the log I promised :)

Code:
--- Opening Log file [October 30 22:29:37 UTC] 


# Linux SMP Console Edition ###################################################
###############################################################################

                       [email protected] Client Version 6.24beta

                          http://folding.stanford.edu

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

Launch directory: /home/muhon/folding
Executable: ./fah6
Arguments: -smp -verbosity 9 -advmethods 

[22:29:37] - Ask before connecting: No
[22:29:37] - User name: Muhon (Team 32)
[22:29:37] - User ID: 3B9C83CC37A1F811
[22:29:37] - Machine ID: 1
[22:29:37] 
[22:29:37] Loaded queue successfully.
[22:29:37] - Autosending finished units... [October 30 22:29:37 UTC]
[22:29:37] Trying to send all finished work units
[22:29:37] + No unsent completed units remaining.
[22:29:37] - Autosend completed
[22:29:37] 
[22:29:37] + Processing work unit
[22:29:37] Work type a1 not eligible for variable processors
[22:29:37] Core required: FahCore_a1.exe
[22:29:37] Core found.
[22:29:37] Working on queue slot 05 [October 30 22:29:37 UTC]
[22:29:37] + Working ...
[22:29:37] - Calling './mpiexec -np 4 -host 127.0.0.1 ./FahCore_a1.exe -dir work/ -suffix 05 -checkpoint 15 -verbose -lifeline 5381 -version 624'

[22:29:38] 
[22:29:38] *------------------------------*
[22:29:38] [email protected] Gromacs SMP Core
[22:29:38] Version 1.74 (November 27, 2006)
[22:29:38] 
[22:29:38] Preparing to commence simulation
[22:29:38] - Ensuring status. Please wait.
[22:29:54] - Looking at optimizations...
[22:29:54] - Working with standard loops on this execution.
[22:29:55] Examination of work files indicates 8 consecutive improper terminations of core.
[22:29:55] - Expanded 743795 -> 3855925 (decompressed 518.4 percent)
[22:29:55] 
[22:29:55] Project: 5101 (Run 0, Clone 33, Gen 98)
[22:29:55] 
[22:29:55] Entering M.D.
[22:30:01] Protein: CMF in water
[22:30:01] Writing local files
[22:30:01] Completed 1531230 out of 2000000 steps  (76 percent)
[22:30:01] Extra SSE boost OK.
[22:30:01] 
[22:30:01] Completed 1531230 out of 2000000 steps  (76 percent)
[22:30:01] Extra SSE boost OK.
[22:42:08] Writing local files
[22:42:08] Completed 1540000 out of 2000000 steps  (77 percent)
[22:57:08] Timered checkpoint triggered.
[23:10:53] Writing local files
[23:10:53] Completed 1560000 out of 2000000 steps  (78 percent)
[23:26:53] Timered checkpoint triggered.
[23:37:50] Writing local files
[23:37:50] Completed 1580000 out of 2000000 steps  (79 percent)
[23:53:50] Timered checkpoint triggered.
[00:04:38] Writing local files
[00:04:38] Completed 1600000 out of 2000000 steps  (80 percent)
[00:20:37] Timered checkpoint triggered.
 

ChasR

Senior Member
Joined
Apr 12, 2004
Location
Atlanta
You're running normally. p5101 is an a1 core WU. A1 WUs don't use all the cpu cycles which is one of the reasons they are slower than a2 Wus. From time to time Linux rigs pick up a1 Wus. All is well.
 
OP
muhon_zen

muhon_zen

Member
Joined
Apr 24, 2003
Location
Brazil
Thanks for answering. Yeah, I finished the WU and now it picked up the a2 core and worked perfectly. :D