I don't wanna be the smartarse here but doing this on a standalone not used server is not a life time example.
What need to be done is a performance test and traffic simulation with p.e. a tool like JMeter Load Testing tool , this way you will get a glimpse of what will go on when it's live
else you are just relying on the different hardware everyone posses when running.
Anyhow for the sake of a test example I guess all the above is not relevant but taken in consideration if for example 100 users try to do the same thing on the server (also I'm guessing you should not process row 1bil rows in a live environment, except if you have a server at CERN or something :P ) and see the results there.

Also this should probably be better in codebank.