Running Xyce

Running Xyce from the command line is straightforward. The scripts xmpirun and runxyce created during installation (see Section 1.2 of the User's Guide) set up the runtime environment and execute Xyce. (Microsoft Windows users should use the runxyce.bat batch file.) Depending on whether you are using a version compiled with MPI support or a serial version, there are two ways, respectively, to begin running Xyce:

Running serial Xyce:

$ runxyce [options] <netlist filename>

Running Xyce in parallel:

$ xmpirun -np <# procs> [options] <netlist filename>

where [options] are the command line arguments for Xyce. For example, to log output to a file named sample.log and use faster direct matrix access type:

$ runxyce -l sample.log -dma on <netlist filename>

The next example runs parallel Xyce on four processors and places the results into a
comma separated value file named results.csv:

$ xmpirun -np 4 -delim COMMA -o results.csv <netlist filename>

These examples assume that <netlist filename> is either in the current working direc-tory or includes the path (full or relative) to the netlist file. See Appendix B of the User's Guide for command line details. Help is also accessible with the -h option.

For MPI runs, [options] may also include command line arguments to mpirun. Consult the documentation installed with MPI on your platform for more details and MPI options. The “-np <# procs>” denotes the number of processors to use for the simulation for the particular computer one is using. NOTE: It is critical that the number of processors used is less than the number of devices and voltage nodes in the netlist. The appropriate script used to run Xyce for each supported platform is listed in the table below.

While Xyce is running, the progress of the simulation is output to the command line window or optionally to a log file if so specified on the command line (see command line options in Appendix B in the User Guide).

Last Modified: August 6, 2004