ACCELERATION OF NEAR FIELD COMPUTATION IN MLFMA ON A SINGLE GPU BY GENERATING REDUNDANCY IN DATA

Improving efficiency of multi-level fast multi-pole algorithm (MLFMA) on distributed and parallel systemshas been vastly studied, specially for GPUs. Unlike the far-field computation, acceleration of near-fieldcomputation in MLFMA algorithm on GPUs was of less concern in the literature, however there are somesolutions that exploited special specifications of GPU’s memory. This article proposes data replication forP۲P operator and uses analytical performance models to determine its optimality criteria. By modellingthe speedup, we found that making threads independence by creating redundancy in the data makes thealgorithm for lower dense problems nearly ۱۳ times faster than non-redundant mode.

کلیدواژه ها:

Multilevel Fast Multi-Pole Algorithm ، Graphics Processors ، Performance Evaluation