Dynamic data redistribution enhances data locality and im-proves algorithm performance for numerous scientific problems on dis-tributed memory multi-computers systems. Previous results focus on reducing index computational cost, schedule computational cos
