I'm glad you asked out to test this yourself - that is a good sign :)
I usually compare the execution plans (if there's no difference then it doesn't matter which you use - for that statement). And I also compare output of the statement with these SETs:
set statistics io on --physical and logical reads set statistics time on --execution time by statement in milliseconds
Let us know what you find
Be One with the Optimizer TG
usually use thing liks showplan xml, and the execution plans but was wondering if there were any further ways of doing. Can be a bit tricky when dealing with nearly a billion rows! also cautious of the fact the first execution plan may be cached and the following partially using that and five a slightly false performance boost.