A lightweight CaaS private cloud architecture and models for many task computing

CPU time has long been a remaining problem for large-scale scientific computing. For some less computingintensive but many-task-computing codes, it may take several weeks or even months to run. With the increase of datasets scale, computing time grows dramatically - this hence can be more and more challenging. To address this problem, most solutions based on ongoing hardware and/or software investment. To address this issue, this research is to make the following contributions to literature: a new “Computing performance first” Computing as a Service (CaaS) model to support scientific computing and contribute as a novel high-performance private cloud computing (HPPCC) method. Efficiency of this approach is demonstrated by experiments with numerical simulation on both with End nodes and without End nodes.