Solving the performance issue in Database is the ART. If you have any specific table that has lot of records, you have to analyse the fetching the records query. Based on the query you have to create index.
In SQL Server, we have two kind of index , ie Clustered and Nonclustered.
Consider the query SELECT Name, age FROM emp WHERE EmployeeID = 500. When this query is executed, the query optimizer evaluates each available method for retrieving the data and selects the most efficient method. The method may be a table scan, or may be scanning one or more indexes if they exist.
So you have to create indexes based on you fetching.
Direction is important than speed