RECFMM: recursive parallelization of the adaptive fast multipole method for Coulomb and screened Coulomb interactions. We present RECFMM, a program representation and implementation of a recursive scheme for parallelizing the adaptive fast multipole method (FMM) on shared-memory computers. It achieves remarkable high performance while maintaining mathematical clarity and flexibility. The parallelization scheme signifies the recursion feature that is intrinsic to the FMM but was not well exploited. The program modules of RECFMM constitute a map between numerical computation components and advanced architecture mechanisms. The mathematical structure is preserved and exploited, not obscured nor compromised, by parallel rendition of the recursion scheme. Modern software systems -- CILK in particular, which provides graph-theoretic optimal scheduling in adaptation to the dynamics in parallel execution -- are employed. RECFMM supports multiple algorithm variants that mark the major advances with low-frequency interaction kernels, and includes the asymmetrical version where the source particle ensemble is not necessarily the same as the target particle ensemble. We demonstrate parallel performance with Coulomb and screened Coulomb interactions.
Keywords for this software
References in zbMATH (referenced in 6 articles )
Showing results 1 to 6 of 6.
- Saunders, William Robert; Grant, James; Müller, Eike Hermann; Thompson, Ian: Fast electrostatic solvers for kinetic Monte Carlo simulations (2020)
- Tian, Yan; Xiang, Shuhuang; Liu, Guidong: Fast computation of the spectral differentiation by the fast multipole method (2019)
- Debuhr, J.; Zhang, B.; Sterling, T.: Revision of DASHMM: dynamic adaptive system for hierarchical multipole methods (2018)
- Fang, Fuhui; Huang, Jingfang; Huber, Gary; McCammon, J. Andrew; Zhang, Bo: Hierarchical orthogonal matrix generation and matrix-vector multiplications in rigid body simulations (2018)
- Liu, Guidong; Xiang, Shuhuang: Fast multipole methods for approximating a function from sampling values (2017)
- Zhang, Bo; Huang, Jingfang; Pitsianis, Nikos P.; Sun, Xiaobai: RECFMM: recursive parallelization of the adaptive fast multipole method for Coulomb and screened Coulomb interactions (2016)