Abstract: Cloud computing is gaining significant interest for modeling and simulation (M&S). The concept of using computing power as a utility is very attractive to users, who can access state-of-the-art hardware and software without capital investment. Cloud computing makes M&S affordable, especially for small and medium enterprises (SMEs) that otherwise do not have the resources and expertise to build and use such tools. However, if not managed carefully, use of cloud resources can lead to waste (unused resources), which results in high cost. To manage and optimize multiple cloud resources is not a trivial task.
We have developed a cloud platform for easy deployment of simulation software and applications. The work was developed within the CloudSME (Cloud-based simulation platform for manufacturing and engineering) project funded by the EC FP7 programme. This work is now being extended with services auto-scaling and application orchestration functionalities under the COLA (cloud orchestration at the level of application) project supported by the EC H2020 programme.
This talk will present recent experiences and results and discuss how end-users of both agent-based simulation and discrete-event simulation can access and benefit from auto-scaling cloud-based M&S. In particular, we will discuss how these might impact the future development of scientific applications.