OpenSeesMPI

KO | EN

gitlite — search

#python #hacktoberfest #typescript #llm #golang #go #machine-learning #nodejs #java #kubernetes #pytorch #mcp #openai #open-source

(★ 9)

Portable MPI for OpenSees

OpenSeesMPI 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# OpenSeesMPI
OpenSeesMPI uses the TclMPI package [1] to replicate the message passing functionality of OpenSeesMP [2], without requiring a parallel version of OpenSees.

## Installation and Basic Use
* Install mpiexec
* Install [Tcl](https://platform.activestate.com/ambaker/ActiveTcl-8.6.10)
* Install [OpenSees](https://github.com/ambaker1/OpenSees-Installer)
* Install [TclMPI](https://github.com/akohlmey/tclmpi)
* Install OpenSeesMPI from this repository's latest release.

Then, OpenSeesMPI can be called as shown below, where $np represents the number of parallel processes, and $inputFile represents the Tcl input script.

`OpenSeesMPI -n $np $inputfile`
  
## Advanced Use
Arguments up to the last argument or up to "--" are taken as mpiexec options.
The remaining arguments are taken as the input file and the input arguments.
If input arguments are desired, the option terminator "--" must be used.

`OpenSeesMPI <$opt1 $opt2 ...> <--> $inputFile <$arg1 $arg2 ...>`
  
## Command Documentation
Functionality is compatible with OpenSeesMP, with a few additional features.
  
* The command _getPID_ returns the integer rank of the process, using the TclMPI binding _::tclmpi::comm_rank_.
  
  `getPID` 
  
* The command _getNP_ returns the number of parallel processes, using the TclMPI binding _::tclmpi::comm_size_.
  
  `getNP` 
  
* The command _send_ either sends $data from the current process to process $pid, or broadcasts $data to all other processes (must be called from process 0), using the TclMPI bindings _::tclmpi::send_ and _::tclmpi::bcast_.
  
  `send <-pid $pid> $data`
  
* The command _recv_ either returns data sent from pid $pid (if $pid is "ANY", it will receive data from any process), or returns data broadcast from process 0, using the TclMPI bindings _::tclmpi::recv_ and _::tclmpi::bcast_. Optionally, the received data can be stored in a variable with the $varName argument. Note that the $varName argument is optional with OpenSeesMPI but mandatory in OpenSeesMP.
  
  `recv <-pid $pid> <$varName>`
  
* The command _barrier_ waits for all processes to reach the barrier, using the TclMPI binding _::tclmpi::barrier_.
  
  `barrier`
  
More advanced MPI functionality is available with the TclMPI package, which is automatically loaded when using OpenSeesMPI. TclMPI commands are documented here: https://akohlmey.github.io/tclmpi/

## Citations
1. Axel Kohlmeyer. (2021). TclMPI: Release 1.2 [Data set]. Zenodo. DOI: 10.5281/zenodo.5637677
2. Mckenna, F. (2011). OpenSees: A Framework for Earthquake Engineering Simulation. Computing in Science & Engineering, 13(4), 58–66. https://doi.org/10.1109/MCSE.2011.66

## Acknowledgements
Thanks to Dr. Axel Kohlmeyer for making TclMPI available on Windows.

## To Do:
* Provide Linux and Mac OS equivalents of OpenSeesMPI.bat

OpenSeesMPI

OpenSeesMPI 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

OpenSeesMPI

OpenSeesMPI 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README