Research On Parallel Processing Of The Energy In Biomacromolecules By ABEEMσπ Model

Currently, Molecular dynamics simulation is the prospect of large and complex calculation of the most widely used methods of biological systems. Along with the molecular mechanics simulation of the rapid development of technology, molecular dynamics simulation technique used in many biological systems, biological macromolecules, such as water molecules, peptide compounds, polysaccharides and so on. The technology enables a variety of large and complex biological systems in a variety of physical and chemical properties of the accuracy of the calculation has improved significantly. Molecular dynamics simulation is the force field and Newton's mechanics developed by computer simulation and a new calculation. Dynamics simulation has the advantage for the movement of particles in the simulation environment to provide a correct theoretical basis and can also access the system dynamics and other statistics; simulation results of high accuracy, generally applicable to a variety of systems and the Discussion of species characteristics. In this paper, parallel to solve the energy field in the design and implementation. The main work:1 ABEEMσπforce field of energy on solving the serial code for analysis. ABEEMσπmodel to calculate the energy system, ie find all the atoms in the system,σbond, lone pairs,πbond between the sites and energy. Static energy of the original program, seeking more than one item with a serial iteration loop nested within each other to achieve, this method to solve the energy function of biological macromolecules items take a long time, so that decreased the efficiency of research questions.2 Macromolecules system for solving the energy under the disadvantages of several proposed solutions for solving the parallel processing, by comparing several different parallel program, the MPI, OpenMP and MPI OpenMP hybrid parallel method is applied to ABEEMσπforce field , in the comparative analysis of several parallel strategy, based on the successful implementation of the ABEEMσπforce field energy term in the solution of all parallel.3 OpenMP through analysis and comparison of STATIC, DYNAMIC, GUIDE, such as several scheduling policies, and application of different scheduling strategies to solve the energy of the parallel program. The experimental data showed that different scheduling policies, procedures, parallel performance is not the same. Successfully found a molecular dynamics simulation for solving the energy of the scheduling policy in parallel, shortening the different scheduling strategies for parallel program execution time.4 The introduction of the Intel-Thread-Checking tool, and the use of MPI OpenMP hybrid parallel program thread checks to eliminate the use of parallel multi-core platforms caused by data-race, avoided because data-race results caused by inaccurate and low efficiency of conflict.