Java 9 High Performance

Understanding the limitations of the JMH

Even the best tools can fail to provide you a solution which is ultimate and free from the problems that are associated with setting up microbenchmarks. The same is the issue when using microbenchmarks in Java through the use of the JMH. Remember, a benchmark is an experiment, which typically finds out the time and throughput of the machine in order to find out the functional benefits of using a particular code.

The use of a harness usually eliminates most of the problems. However, the basic principle behind the use of a microbenchmark remains the same. We cannot directly measure how a particular code performs within the CPU and the Java virtual machine. However, we can measure how much time and memory ...