Performance of LAMMPS Code on Intel Quad-Core Xeon

MZ Bai and YS Dou and H Tang and SX Sun, 2009 6TH IFIP INTERNATIONAL CONFERENCE ON NETWORK AND PARALLEL COMPUTING, 156-+ (2009).

DOI: 10.1109/NPC.2009.31

Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) code was examined in an Intel Quad-Core Xeon platform for its speedup and scaling ability. The study shows that the most time-consuming task in the code is force computing and this part can be scaled linearly. However, other tasks do not have this feature. Future work will focus on speeding up of these other parts to enhance the performance of the molecular dynamics program.

Return to Publications page