The dust particles(grains) play an important role in the formation of the ordered structure in planetary plasmas, e.g., the rotating radial spokes of the Saturn's rings, as well as processing plasmas, e.g., Carbon fullerene. In order to reveal the dynamics of dust grains, we developed a new particle simulation code, which has the following characteristics. 1) The dust particles, which are initially neutral in charge, are getting negatively or positively charged due to the background plasma as the time goes by. 2) The charging process is governed by Poisson probability process, namely, a series of the random number should be used. 3) The number of dust grains increases due to the supplement of neutral grains into the system. On the other hand, the number of grains decreases when the coagulation occurs between negatively and positively charged grains.

For the parallel processing and HPF programming of the dust particle code, there are some essential points to be overcome which originate from the above characteristics. First, since each grain has a different charge, mass, Debye length and so on, the PIC (particle in cell) procedures used in the regular particle simulation codes cannot be applied. The electrostatic potential similar to Yukawa potential should be summed up over all grains which are distributed to multi-cpu system, and the indirect summation was employed. Second, a huge number of a series of the uniform random number (of the order of 2 to the power 50) is necessary for charging process. An algorithm for the random number in parallel processing is introduced. Third, the load balance among processors should be taken into account according to the increase or decrease of grains in each processor. This balance is realized by the cyclic distribution. In spite of these difficulties, since the vector processing of high-performance can not be expected, the (massively) parallel processing would be inevitable. In the presentation, we introduce the detailed HPF programing and the processing performance in comparison with the result of SX-5. The particle simulation codes in the open system, in which the number of plasma particles are increasing and/or decreasing, is being applied to various important phenomena, and hence, the application presented here will lead a new way for a large-scale particle simulation in open systems.