Hi Arpit,
Although an aggregate can be useful, I would advice to first check if other aspects of InfoCube modeling are OK. I think of the following:
- Check the Dimensions: not too small but certainly not too big (< 10% of fact table);
- Check if Partitioning on a Time Characteristic is possible (e.g. 0FISCPER).
Especially when using Partitioning you have to compress the InfoCube regularly. Do not wait too long with that (only the last requests can be uncompressed).
Hopefully these actions will increase the performance. If not, then try using an aggregate. You can find more information and modeling tips in SAP Help and on SCN, e.g. SAP BW Query Performance Tuning with Aggregates.
Best regards,
Sander