Improve query performance on large tables in SQL Server

Hi,

I have few tables which I am calling in view with joins.

Its taking a lot of time to get records, due to the fact that all the columns in every table are nvarchar(999) including the primary keys.

I can't change primary keys to int of any table(for some reason).

But just wanted to know apart from applying non clustered indexes to other columns, what may be the other techniques to improve the performance.