Setting inparams correctly for running on HPC

[EoghanTotten]

Hello,

I just wanted to check that I am using the optimal inparams when running MC Kernel on many, many processors and using parallel netcdf on High Performance Computing resources e.g MAXIMUM_ITERATIONS, STRAIN_BUFFER_SIZE, INTERMEDIATE_DUMP_TIME (what fraction of the wall time should this be set at or should it be left at 0), ALLOWED_ERROR and ALLOWED_RELATIVE_ERROR.

Any pointers/advice to save wall time/ run time?

Many thanks.