We have a legacy database with some (older) columns using \"SQL_Latin1_General_CP1_CI_AS\" and more recent changes have used \"Latin1_General_CI_AS\".
This is a pain
There is more info on this MSDN forum:
http://social.msdn.microsoft.com/Forums/en-US/sqlgetstarted/thread/196b4586-1338-434d-ba8c-49fa3c9bdeeb/
Which states:
You should see little difference if the collation is SQL_Latin1_General_CP1_CI_AS or Latin1_General_CI_AS, but both have instances where they are faster or slower than the other.
Latin1_General_CI_AS :- Latin1-General, case-insensitive, accent- sensitive, kanatype-insensitive, width-insensitive
SQL_Latin1_General_CP1_CI_AS:- Latin1-General, case-insensitive, accent-sensitive, kanatype-insensitive, width-insensitive for Unicode Data, SQL Server Sort Order 52 on Code Page 1252 for non-Unicode Data
Therefore in my opinion you shouldn't see a difference, especially if your data is only a-z0-9