Creo Simulate > Additional Information > Improving Performance > Guidelines for Disk Usage and Allocating Swap Space
  
Guidelines for Disk Usage and Allocating Swap Space
One of the major factor that influences performance is disk usage. During an analysis, Creo Simulate writes all its results to disk. Also, intermediate data that is required during the analysis is temporarily stored on disk . The following guidelines should help improve performance.
Make sure you are not using any drives that are mounted across the network.
Use drives that have a generous amount of empty space on them.
Occasionally defragment your disks so that data can be written and read in large contiguous blocks.
Use fast hard drives.
Use disk striping with a redundant array of independent disks (RAID) to increase IO performance.
Use a RAM disk instead of a hard disk.
Use a solid-state drive instead of a hard disk drive.
If possible, allocate swap space on disks that are not used for the Creo Simulate working directories.
For jobs using the direct solver, total disk space used is very approximately the size of the global matrix file plus three times the size of the element matrix file.
 For jobs using the iterative solver, total disk space used is very approximately the size of the element matrix file plus the global matrix profile for the last pass that used the direct solver (that is, the pass immediately before the first iterative solver pass).
Estimating Job Memory
You can estimate the job memory of an MPA analysis by performing these steps:
1. Start running the analysis.
2. After the first two passes, check the memory usage reported in the summary file .rpt at the end of each pass.
3. Use linear extrapolation to determine the amount of memory the run will need for the pass at which you expect the analysis to converge (for example, the maximum polynomial order you set for the analysis).
 
* If you are using the iterative solver, the memory usage will jump by the solram value after the first iterative solver pass.
* For example, if you are running with –iter 2, then you should extrapolate after pass 3 ends or add solram to the estimate you extrapolated from passes 1 and 2. Your result will be the estimate of total memory use for the job.