MDSIMAID: automatic parameter optimization in fast electrostatic algorithms

Michael S Crocker; Scott S Hampton; Thierry Matthey; Jesús A Izaguirre

doi:10.1002/jcc.20240

MDSIMAID: automatic parameter optimization in fast electrostatic algorithms

J Comput Chem. 2005 Jul 30;26(10):1021-31. doi: 10.1002/jcc.20240.

Authors

Michael S Crocker¹, Scott S Hampton, Thierry Matthey, Jesús A Izaguirre

Affiliation

¹ Department of Computer Science and Engineering, University of Notre Dame, Notre Dame, Indiana 46556, USA.

PMID: 15884103
DOI: 10.1002/jcc.20240

Abstract

MDSIMAID is a recommender system that optimizes parallel Particle Mesh Ewald (PME) and both sequential and parallel multigrid (MG) summation fast electrostatic solvers. MDSIMAID optimizes the running time or parallel scalability of these methods within a given error tolerance. MDSIMAID performs a run time constrained search on the parameter space of each method starting from semiempirical performance models. Recommended parameters are presented to the user. MDSIMAID's optimization of MG leads to configurations that are up to 14 times faster or 17 times more accurate than published recommendations. Optimization of PME can improve its parallel scalability, making it run twice as fast in parallel in our tests. MDSIMAID and its Python source code are accessible through a Web portal located at http://mdsimaid.cse.nd.edu.