|Message: Simulation time increase much faster than direct proportional with events number||Not Logged In (login)|
Click on the Forum title, e.g. on the "Forums by Category" page, to read a sequence of postings to the Forum and its threads all in one page. If you are only interested in one thread or the thread following a specific posting, click the thread or the posting, which takes you to a smaller page, which contains only the part you are interested in and may be easier to navigate.
Messages are "chained" if there are only replies at the first level, i.e. 1/1.html, 1/1/1.html etc. In case of "chained" messages the message number is replaced by the icon and there is no indentation.
Inline: Display the subject line only or also the text of the posting(s); for the choice "All" the "Outline" choices are switched off.
|1||0||1||no text / full text of posting|
|2||1||All||text for level 1 only / text for All postings|
Outline: Choose the depth of the posting thread, successive toggle controls provide increasing detail.
|1||2||1||2 levels / 1 level (original posting)|
|2||3||2||3 levels / 2 levels|
|3||3||All||3 levels / all levels (all postings)|
Dear Geant4 experts!
Iím carrying out MC simulations of neutron stimulated gamma spectra of soil. To have spectra with suitable accuracy the number of simulated events should be more than 1e8. Iíve run these problems in multithread mode on the laptop (Panasonic CF53), geant4-10-01-patch02 with 4 threads, on vSMP HPCC - Virtual Symmetric Multiprocessing High Performance Computing Cluster, geant4-10-01-patch01, with 8 threads. In these cases the time of simulation increases direct proportional to the numbers of ordered events, the ratio of user time to wall time doesnít depend on number of events and practically equal to the number of used threads. All looks OK!
Iíve tried to run the same simulation on the Ceres HPC ARS SciNet, geant4-10-02-patch02, with different number of threads, from 8 to 40. If the number of events did not increase 1e6 than all looks OK, like in laptop and HPCC, while for number of events mare than 2e6 the time of simulation starts increase with number of events much faster than direct proportional and ratio of user time to wall time catastrophically drops down to one.
Could you say me reason of such behavior of simulation on Ceres HPC ARS SciNet, geant4-10-02-patch02, and how can I run simulation with events number close to 1e8 and more such way that time of simulation increases direct proportional with events number and ratio of user time to wall time stores constant and equal number of threads?
Thanks in advance
|Inline Depth:||Outline Depth:||Add message:|