Java vs C Performance. Again. There seems to be a set of widely held wildly incorrect beliefs about Java and C performance. I say "widely held" because I hear them constantly - even after 20 years of pointing out the evidence. I say "incorrect beliefs" because they are shown wrong with even the most rudimentary scientific study, i.e. a simple benchmark. I say "wildly incorrect" because nearly always I see blatantly broken benchmarking, being used to drive home some wrong point.

E.g.: I've heard a person claim some web server is better than another because "printf (in C) is faster than Java's printing!"; or seen Java locking performance tested - on a single-threaded and single-core systems; or read reports using system timers that only report to 10msec accuracy used for nano-second timing. I've seen an HPC C program vary speed by 30% depending on the user name length, or by linking it with file-globbing vs some other order. I've seen both languages run an order of magnitude faster than the other, both purportedly trying to accomplish the same task, but with wildly different implementations. The hall of benchmark shame goes on and on.

This is my attempt to shed some light on the situation; we'll look at both languages' strong and weak points - and then look at all the different broken ways people gather "facts" to make claims - and then up level to looking at programming context surrounding language choices, which often have a much stronger impact on the business of programming than mere language choice! This talk is a fun romp through everybody's favorite whipping boy: the *other* (not mine!) silly languages Out There, And Why They Are Silly (and I'm Not)