Collations are sort orders. Not all alphabets order their letters in the same way, so depending on where you may live, or who you do business with, you may end up seeing this a lot more.
Collations can give you a small performance boost, so long as you are always conscious of the collation. The fastest collation is binary order. After that is case sensitive, then the default case insensitive. In case sensitive or binary collations, you have to address each object exactly. So two tables TABLE1 and Table1 are two distinct tables in a case sensitive or binary sorted system, whereas these two names could not exist side by side in a case insensitive system.
Is the performance benefit better than the hair pulling you will go through trying to remember table names in a case sensitive system? Up to you.