Theres not much to go on here, but for start I would read
https://websmp109.sap-ag.de/~sapidb/011000358700000558182013E/HANA_ExecutionOfModels.pdf
Theres few tricks to enable filter push down.
Also if attribute view takes too long maybe You use refferential join, and if reff join is used outside of context of analytic view it is actually inner join hence it is slower.
But for sure I would look in partitioning table as big as this one, and then go from there.
BTW, 2 billion rows is size limitation for column store tables per patition and recomedations are:
Partition size Table rows < 500 milion do not partition unless:
1.Join tables are partitioned
2. Table Growth is expected
>500mil. Split into 300 mil per partition
Of course tables that gets joined together have to match data by partitions to avoid data transfer between nodes.