Advanced search

Message boards : Number crunching : CPU Load = 028, 0.31 etc ...

Post to thread Subscribe Sort
Author Message
Profile Carlesa25
Avatar
Send message
Joined: 13 Nov 10
Posts: 328
Credit: 72,619,453
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 22148 - Posted: 22 Sep 2011 | 17:31:47 UTC

Hello: For some time I see that the burden on the CPU varies for the same type of task, from one day to another, and not the same on Windows than on Linux.

For example: 0.28, 0.31 or 0.33 or CPU for the same type of task by day and by the OS. Suppose you have a reasonable explanation. Greetings.

ID: 22148 | Rating: 0 | rate: Rate + / Rate - Report as offensive    Reply Quote
Profile skgiven
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,995,359,260
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 22155 - Posted: 23 Sep 2011 | 14:43:37 UTC - in response to Message 22148.

Windows and Linux use different apps:
Linux is using the 6.14app and Windows is using the 6.15app.

Recently there has been variation in runtime on Windows for similar task types, so I would expect the same on Linux. Not all tasks are the same, and some naturally use the CPU more than others. Presently there are several different research types being run.

So, lots of variation. Fortunately you can choose to reduce the variation by selecting either long or standard task lengths:

    ACEMD for long runs (8-12 hours on fastest GPU)
    ACEMD standard

ID: 22155 | Rating: 0 | rate: Rate + / Rate - Report as offensive    Reply Quote
Post to thread

Message boards : Number crunching : CPU Load = 028, 0.31 etc ...

//