Reading duplicate values from an index
Reading an index with duplicate entries incurs an additional cost over reading the table sequentially.
Each entry or set of entries with the same value must be located in the index. Then, for each entry in the index, a random access must be made to the table to read the associated row. However, if there are many duplicate rows per distinct index value, and the associated table is highly clustered, the added cost of joining through the index can be slight.