V. V. Babkova’s research while affiliated with Institute for System Programming, Russian Academy of Sciences and other places

The aim of the paper is to develop a scalable parallel program for numerical solution of the equation set that simulates processes and origin conditions of intensive atmospheric vortices (IAV) in 3D compressible atmosphere based on the theory of mesovortice turbulence by Nikolaevskii. The original system of nonlinear differential equations and its boundary and initial conditions are considered in the paper. The difference scheme with required precision is selected. The parameters’ restrictions are also formulated. The structure of parallel software that performs computations on a high-performance cluster is considered. The problems of the software’s optimization in order to improve its scalability are discussed. The results of numerical calculations are considered in summary.

A technology of the iterative development of parallel programs and the corresponding development tools on the basis of the ParJava environment are considered. A benefit of the ParJava environment is that the the most part of the work can be done on a development computer using a model of the parallel program to be developed. This considerably reduces the the overheads in terms of time and resources used. The proposed technology is illustrated using the development of a parallel program for simulating intense atmospheric vortices as an example.

The interaction of turbulent motion of air leads to the energy transition over the hierarchy of scales. At a specific ratio of the energy of mesoeddies and macroeddies ( E mes / E mac ) [2], an avalanche increase of intense atmospheric vortices (IAV) is observed. However, when the energy of mesoeddies is exhausted, a slow reverse process of energy transfer to the mesoscale starts. The possibility of the IAV development due to the initial energy of mesoeddies was shown for the first time in our paper [2], in which we presented the results of numerical modeling of an eddy with axial symmetry in a 1D incompressible nonstratified atmosphere. Mechanisms of the appearance of the initial field of mesoeddies can be different: the formation of coherent structures (scale 50‐500 m) in the surface atmospheric layer [4]; eddies in turbulent floating jets generated by J both natural and anthropogenic factors [5]; anomalies of monthly mean temperature [6]; and, of course, breaking of global vortices, for example, tropical cyclones [7]. In this paper, we realize a numerical model of tornado development in a 3D compressible dry adiabatic atmosphere from a cloud of mesoeddies. The enormous volume of calculations for obtaining a numerical solution to problems of such a kind requires implementation of the program code using highly efficient computational systems. Cluster system is one of the most popular types of such systems. The numerical model was implemented in the Parjava parallel program environment for cluster systems.

The goal of the paper is development of a scalable parallel program calculating the numerical solution of the system of equations modeling the processes and origin conditions of intensive atmospheric vortices (IAV) in 3D compressible atmosphere according to the theory of mesovortice turbulence by Nikolaevskiy. Original system of non-linear equations, and its initial and boundary conditions are discussed. The structure of a parallel program for high performance cluster is developed. The problems concerning to optimization of the program in order to increase its scalability are studied. In summary the results of numerical computations are discussed.

This paper is devoted to presentation of ParJava environment being developed in ISP RAS. It provides to application programmer a set of tools supporting design of parallel programs for high performance clusters using Java extended by a standard MPI library. ParJava allows to make the most part of development using instrumental computer. It is possible because of using the original model of a parallel Java program. The model is designed in such a way that its interpretation allows to predict program execution time (performance), or its other dynamic attributes (profiles, traces, etc.) for given cluster. Accuracy of prediction may be improved using special benchmarks to obtain cluster's parameters and measuring the execution time of basic blocks using single node of the cluster. Using of the interpreter in real applications development (a program of tornado modeling and a program of modeling) demonstrated that prediction error does not exceed 10%. Use of an instrumental computer instead of the cluster together with new generation language Java appreciably increases the productivity of development. In closing current research activities devoted to methods and tools allowing to discover program parallelism, automatic parallel code generation and automatic tuning of code to given hardware are discussed.