In recent years, performance of mobile computer devices such as smartphones and tablet devices has been improved, and mobile computer devices are attracted attention as new computational resources for high performance computation. We are developing a cluster computer system that consists of mobile computer devices running Android OS. In the preceding researches including ours, super-user authority is required to install the system files for parallel processing environment onto the restricted area of file system. Hence, we cannot build the cluster computer system using mobile computer devices without super-user authority. Furthermore, we clarify the performance of our proposed system as compared to our previous system, by using benchmark programs for MPI parallel processing. The results of performance comparison show that our proposed system has lower performance than our previous system on only one node. However, as the number of nodes increased, the difference in performance tends to decrease.
CITATION STYLE
Nissato, M., Sugiyama, H., Ootsu, K., Ohkawa, T., & Yokota, T. (2020). Realization and Preliminary Evaluation of MPI Runtime Environment on Android Cluster. In Advances in Intelligent Systems and Computing (Vol. 926, pp. 407–418). Springer Verlag. https://doi.org/10.1007/978-3-030-15032-7_35
Mendeley helps you to discover research relevant for your work.