Zitat:
Original erstellt von Tobias1982:
Hallo,
ich moechte eine Dynamic-Implicit-Rechnung weitestmoeglich beschleunigen, da diese dann standadisiert 200-mal ausgefuehrt werden soll. Um einen eventuellen Grund fuer die langsame Berechnung zu finden, habe ich mir in der Outlook-Item-Datei die Job-Time-Summary angeschaut mit folgenden Werten:User Time (sec) = 2148.9
System Time (sec) = 150.2
Total CPU Time (sec) = 2299.1
Wallclock Time (sec) = 5717
Weiss jemand was die einzelnen Zeitangaben bedeuten und mit welchen Massnahmen/Einstellungen ich eine Dynamic-Implicit-Rechnung beschleunigen kann?
Vielen Dank im Voraus, Gruss Tobias
Was die Zeitangaben bedeuten ist in einer Answer in der Supportdatenbank erklärt. Hier die Antwort:
Both Abaqus/Standard and Abaqus/Explicit print preprocessing job time summary information in the data (.dat) file. Abaqus/Standard prints analysis job time summary information in both the data and message (.msg) files as well. Abaqus/Explicit does not print analysis job time summary information.
The different components are defined as follows:
* USER TIME refers to the CPU time spent executing Abaqus.
* SYSTEM TIME refers to the amount of OS kernel CPU time spent by the operating system doing work on behalf of the Abaqus process.
* TOTAL CPU TIME is the sum of these two numbers.
* WALL CLOCK time refers to the actual physical time spent for the analysis process to complete. If the analysis job is running on a single CPU, and the job has exclusive access to that CPU, the difference between TOTAL CPU TIME and WALL CLOCK TIME is largely the time taken to perform all I/O requests.
If the job is run across several CPUs, the USER TIME, SYSTEM TIME, and TOTAL CPU TIME reported is the sum across all the CPUs.
Information about wallclock time is also printed to the .log file for both Abaqus/Standard and Abaqus/Explicit. Specifically, the wallclock times at which Abaqus processes (pre.x, package.x, standard.x, explicit.x) start and stop are printed. Differences in the reported wall clock times between the .log file and the .dat file may be noticed, and these are due to the time spent in checking out the licences for the various processes. These differences, which look significant when running small jobs, are marginal when running comparatively bigger jobs.
Eine Antwort auf diesen Beitrag verfassen (mit Zitat/Zitat des Beitrags) IP