Integrated Workflow Management for Particle Accelerator Simulation

The use of high-performance computing systems for advanced numerical simulation often involves considerable effort on the part of scientists and engineers to prepare and organize simulation input and output data. In many cases, research productivity is limited by efforts to configure and work with a variety of disconnected software tools, and manually carry out data management, often by ad hoc methods. To reduce this logistical workload, an advanced software framework is proposed that integrates diverse software tools for simulation pre- and post-processing with a workflow management system. Project workflows are specified by task decomposition and simulation asset enumeration so that, in practice, simulation tools will be automatically configured for each specific task with the relevant simulation assets. Individual file management will be handled internally, relieving researchers from this work. The Phase I effort investigates several enabling technologies: (i) new software will be developed for integrating multiple software tools into the present-day workflow management system; (ii) data management software will be extended to track simulation assets that are distributed across multiple networks and computing systems; (iii) new software will be developed for specifying and executing computational pipelines, in which multiple simulation jobs are run in a daisy-chain pattern; (iv) the migration of current simulation codes to cloud-based platforms will also be investigated, as a way to support a potentially larger user base. The Phase II effort will build out and deploy a production system at full scale, for use by research institutions in government, academia, and industry.Commercial applications for this technology can be found in organizations that use advanced numerical computing for physics-based modeling and simulation. This includes major research and development laboratories and manufacturers, where the proposed software can significantly improve technical workforce efficiency.