Compilation

< Supernova User Guide < Compilation

Where the compilation process should be run

Users should not compile their programs on the access node. It has different software (e.g. operating system version) and hardware (including processors) than execution nodes.

To compile your software please run an interactive job in a dedicated queue prace-int with default resources (1 core, 2GB RAM):

 qsub -I -q prace-int 

or with specified resources (if the cluster is heavy loaded it may be easier to get less resources then default):

 qsub -I -q prace-int -l select=1:ncpus=1:mem=1GB

If you want to learn more about "qsub -I" please check the article about running jobs.

Available compilers

Only compilers with own modules are supported. To list versions run:

 module avail gcc
 module avail intel
 module avail pgi

< Supernova User Guide < Compilation