How can researchers make their software more efficient and reduce the environmental impact of computational workloads?
Our latest guide: Green Computing: Introduction to Profiling and Optimisation, written by Robert Chisholm, introduces the fundamentals of code profiling and optimisation, helping researchers identify performance bottlenecks and make their software more efficient.
Profiling allows developers to collect detailed information about how their code executes, identifying which functions or lines consume the most time and computational resources. The guide explains how to interpret profiling results using tools such as flame graphs, icicle graphs and performance summary tables.
It also explores practical optimisation techniques, from eliminating redundant calculations and unnecessary operations to choosing more efficient data structures and library functions.
Whether you are new to profiling or looking to improve existing research software, the guide offers practical advice on identifying opportunities for optimisation, validating changes and improving code performance.