Comment by devicenull

Why do people always benchmark on virtual machines running on someone else's server, and expect meaningful results?