Possibility of HPC Application on Cloud Infrastructure by Container Cluster

Today, High Performance Computing (HPC) has been used to solve various domain problems including science and engineering field or streaming processing application. All of these domains require a lot of computation resources. For HPC application, wall time and accuracy of calculation are important factors. Therefore, cloud infrastructure was hardly acceptable to the HPC application, because cloud infrastructure has performance overhead as compared with native environments. However, new hardware acceleration devices, increased demand for large scale calculations using AI applications and evolution of container technology have increased the possibility of running HPC applications on cloud infrastructure. In this study, we evaluate and compare the performance of several applications such as ; Poisson's solver, ResNet50, and Recurrent Neural Networks (RNNs) with Long Short-Term Memory models (LSTM) - to confirm the possibility of running HPC applications on cloud infrastructure. We submit that there is only 0.05% overhead for ResNet50 on cloud infrastructure. It indicates the possibility of running special purpose HPC applications such as AI training or General-Purpose computing on Graphics Processing Units (GPGPU) oriented applications on cloud infrastructure. We also observe that there is no performance overhead for cache miss rate and InfiniBand latency.