Simulations and high performance computing
Computer simulations of physical systems are an indispensable tool in the research process. Simulation codes use mathematical models implemented in software to approximate physical systems and processes. They can bridge the gap between theory and experiments.
Numerical simulations can serve the following purposes:
- understanding experimental results: many experiments carried out at the CLF or elsewhere do not allow probing of all inner processes that may be of interest. Simulations that replicate the important aspects of an experiment can be used to identify the key processes at work
- designing experiments: before an experiment is performed, numerical simulations are used to validate and optimise the experimental design. This increases the confidence that a desired effect will be observed
- supporting theory: simulations of idealised systems can be used to isolate effects and provide support for theoretical predictions. This allows testing of fundamental ideas
Large scale numerical simulations require the appropriate computational infrastructure. High-performance computing (HPC) hardware and software resources available at the CLF are used to run simulations on hundreds and thousands of computer cores. Various simulation codes are developed and used that make use of the highly parallel hardware.