The capability of a cloud computing system to automatically provision and release resources as needed to handle varying workloads, maintaining performance and minimizing costs.