Optimization of POM Based on Parallel Supercomputing Grid Cloud Platform

As the ocean model continues to improve, the resolution of the model continues to increase. However, this also puts higher and higher requirements on the computing power of the running platform, which makes the existing POM mode operating platform face more and more pressure. In this paper, the POM model based on MPI parallel optimization is transplanted to the high-performance scientific computing and system simulation platform based on Intel Xeon. Through extensive testing of the POM model, we analyzed the problems in the model and developed an optimization strategy to improve the POM model from the function level. The experimental results show that our optimization strategy can effectively reduce the execution time of POM mode, and the parallel acceleration ratio reached 111.26%, which makes the POM mode well improved.