|Message: Re: MT in hpcc problem||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 G4 experts!
I still have a problem with MT simulation on computer cluster. I'll to try explain my problem once again .
As I wrote in my question my software running in multithread (MT) mode on conventional laptop looks OK. I can run my simulation in parallel in all available computer’s cores. Number of available cores can be determinate by user
G4int nThreads = 10; for instance
or is determined by code
G4int nThreads = G4Threading::G4GetNumberOfCores();.
The speed up of simulation (if number of cores not too high) equal the number of cores. On laptop (which have 4 cores) all looks as expected.
High Performance Compute Cluster (http://eng.auburn.edu/admin/ens/hpcc/hpcc-description.html), which I use, consists from 4 nodes with 128 core on each. The PBS-script is used for running the job. The number of nodes and the number of cores are ordered in script. When PBS-script ordered, for instance, n cores and code “nThreads = G4Threading::G4GetNumberOfCores();” the running goes on 128 threads, but on one core only (!) and load of this core 4000% or more. When PBS-script ordered n cores and code “nThreads = n;” the running goes on n threads, but on one core only (!) and load of this core is (~100 x n) %.
Note, if my job runs on 128 cores, other jobs on hpcc are braked.
Due to this reason I have quote for running 10 jobs in 2 thread mode now.
My question is “how can I fixed this problem and run multithread job with 1 thread on each core of hpcc”.
Thanks in advance
|Inline Depth:||Outline Depth:||Add message:|