The internet is full of micro benchmarks explaining the world to beginners. Everyone able to use JMH is trying to explain the world, but only a few do it right. Most benchmarks are either entirely wrong, or draw the wrong conclusions. We're going on a journey to discover and question published microbenchmarks, figure out when you need your own, and decide when you shouldn't run any.The presentation begins with a straightforward introduction to microbenchmarking, using a simple approach that anyone can grasp. You will gain insights into potential pitfalls and learn how to avoid them, ensuring the validity of benchmark results. By the end of the presentation, you will not only be equipped with the skills to conduct meaningful microbenchmarking, but also develop a critical eye for evaluating the validity of results from others.Besides the obvious microbenchmarking tools and techniques, you will also learn a lot about the inner workings of the JVM, your hardware, and how they interact with each other. At the end, you will hopefully be able to make the call whether microbenchmarking is needed at all in your daily work.
Get notified about new features and conference additions.