hello madhusudhan,
1) if you load data into infocube the data will be stored in F fact table. Since F fact table is partitioned based on Request ID , the data is summerized automatically during data load .
2) Next if the data is loaded again the data will not be summrized in the cube, But when u report on the cube, the data in the report is summerized ( This is done by OLAP processor).
3) if you wanna summerize all the data , you need to compress the cube (compression transfers the data from F fact table to E fact table (on E fact Table you can manually partion using 0calmonth, 0fiscvar) ) and it deletes all the request's and summerizes the data and saves it E fact table.
4) If you report on cube after compression, the data will be summerized data due to compression (Here no olap process is used).
5) Hence Burden on OLAP processor is reduced using compression (In ordr to summerize the data)
Thanks,
Ramesh.