Script logics performance improvements - help with parallelization

Hi experts,

I have been browsing the HTG and help documents for BPC 10.1 in NW in order to find out how to improve some of my script logic performance.

I found out that one improvement could be the use of parallelization.

I am a bit confused with which solution to use :

- using RUNLOGIC_PH (http://scn.sap.com/docs/DOC-35152) for having calculations done in parallel

- using *XDIM_PACKAGEBY in the script logic (http://help.sap.com/saphelp_bopacnw101/helpdata/en/f0/5b716f5507458eabc2f57b029a5066/content.htm?frameset=/en/59/f145f708724b9c86d8a65dd57d5e4f/frameset.htm&current_toc=/en/82/f51cf12cfc48c58975b9b5e6fba9aa/plain.htm&node_id=165)

From what I understand, both solutions will allow the script logics to run the calculation on datasets in parallel. Am I right?

What would be the benefits of one or the other solution? Can they be used together?

Also when using parallel execution (with any of the 2 solutions above), is there any interest of setting XDIM_MAXMEMBERS too ?

Finally I was wondering if the Parallelization set up described in the HTG Enable Parallelization for Datamanager packages and Consolidation (http://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/a0255ebe-8a1b-3110-de89-dd2b81b9c477?QuickLink=index&…) would have any impact on performance of script logics called through the DEFAULT_FORMULAS datamanager package.

Thanks for your advices!

Alexandre