Heterogeneous Acceleration and Optimization of the Molecular-Dynamics Simulation Software FIREBALL

doi:10.11871/jfdc.issn.2096-742X.2026.04.009

Abstract

Abstract:

[Objective] It is of great significance to address the high computational cost and limited scalability of FIREBALL for large-scale ab initio molecular dynamics (AIMD) simulations on domestic heterogeneous supercomputers. [Methods] This paper proposes a multi-level heterogeneous parallelization scheme: GPU/DCU acceleration of the Ewald summation via a Kokkos+ FLCL cross-language framework combined with a kernel-splitting strategy; parallelization of the matrix diagonalization module using a decoupled master-worker MPI architecture; and OpenMP optimization of complex modules to balance performance and cost. [Results] Performance evaluation shows speedups of up to 130×, 7×, and 6× for the core Ewald module, Hamiltonian construction, and diagonalization modules, respectively. Under a hybrid parallel configuration, the overall software performance is improved by a factor of 4. [Limitations] At present, the optimization and validation are mainly conducted for limited system sizes and specific heterogeneous platforms, and the GPU/DCU acceleration is still concentrated on the Ewald module. [Conclusions] This multi-level parallel optimization scheme successfully alleviates several performance bottlenecks of FIREBALL and provides a powerful tool for investigating key physical processes such as defect evolution and interfacial reactions in materials.

Key words: FIREBALL, heterogeneous parallel computing, GPU acceleration, high performance computing, first principles

JIN Runfeng, LIANG Wenhao, MA Yingjin, REN Pengju, JAMES P Lewis, JIN Zhong. Heterogeneous Acceleration and Optimization of the Molecular-Dynamics Simulation Software FIREBALL[J]. Frontiers of Data and Computing, 2026, 8(4): 123-135, https://cstr.cn/32002.14.jfdc.CN10-1649/TP.2026.04.009.

Figures/Tables 13

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Table 1

Fig.6

Fig.7

Fig.8

Fig.9

Fig.10

Fig.11

Fig.12

References 33

[1]	LEWIS J P, JELÍNEK P, ORTEGA J, et al. Advances and applications in the FIREBALL ab initio tight-binding molecular-dynamics formalism[J]. Physica Status Solidi B Basic Research, 2011, 248: 1989-2007. doi: 10.1002/pssb.v248.9
[2]	SHELLMAN S D, LEWIS J P, GLAESEMANN K R, et al. Massively parallel linear-scaling algorithm in an ab initio local-orbital total-energy method[J]. Journal of Computational Physics, 2003, 188(1): 1-15. doi: 10.1016/S0021-9991(03)00069-X
[3]	CAPPELLETTI R L, COBB M, DRABOLD D A, et al. Neutron-scattering and ab initio molecular-dynamics study of vibrations in glassy GeSe2[J]. Physical Review B, 1995, 52(13): 9133-9136. pmid: 9979942
[4]	OYABU N, POU P, SUGIMOTO Y, et al. Single atomic contact adhesion and dissipation in dynamic force microscopy[J]. Physical Review Letters, 2006, 96(10): 106101. doi: 10.1103/PhysRevLett.96.106101
[5]	KLOCKE M, WOLF D E. Dissipation signals due to lateral tip oscillations in FM-AFM[J]. Beilstein Journal of Nanotechnology, 2014, 5: 2048-2057. doi: 10.3762/bjnano.5.213 pmid: 25551032
[6]	BACHELET G B, HAMANN D R, SCHLÜTER M. Pseudopotentials that work: From H to pu[J]. Physical Review B, 1982, 26(8): 4199-4228. doi: 10.1103/PhysRevB.26.4199
[7]	HAMANN D R, SCHLÜTER M, CHIANG C. Norm-conserving pseudopotentials[J]. Physical Review Letters, 1979, 43(20): 1494-1497. doi: 10.1103/PhysRevLett.43.1494
[8]	FOULKES W M C, HAYDOCK R. Tight-binding models and density-functional theory[J]. Physical Review B, 1989, 39(17): 12520-12536. pmid: 9948117
[9]	HARRIS J. Simplified method for calculating the energy of weakly interacting fragments[J]. Physical Review B, 1985, 31(4): 1770-1779. pmid: 9935980
[10]	JELÍNEK P, WANG H, LEWIS J P, et al. Multicenter approach to the exchange-correlation interactions in ab initio tight-binding methods[J]. Physical Review B, 2005, 71(23): 235101. doi: 10.1103/PhysRevB.71.235101
[11]	SANKEY O F, DEMKOV A A, LENOSKY T. Electronic structure of black sodalite[J]. Physical Review B, 1998, 57(24): 15129-15139. doi: 10.1103/PhysRevB.57.15129
[12]	SANKEY O F, NIKLEWSKI D J. Ab initio multicenter tight-binding model for molecular-dynamics simulations and other applications in covalent systems[J]. Physical Review B, 1989, 40(6): 3979-3995. pmid: 9992372
[13]	CARTER EDWARDS H, TROTT C R, SUNDERLAND D. Kokkos: Enabling manycore performance portability through polymorphic memory access patterns[J]. Journal of Parallel and Distributed Computing, 2014, 74(12): 3202-3216. doi: 10.1016/j.jpdc.2014.07.003
[14]	TROTT C, BERGER-VERGIAT L, POLIAKOFF D, et al. The kokkos EcoSystem: Comprehensive performance portability for high performance computing[J]. Computing in Science & Engineering, 2021, 23(5): 10-18.
[15]	TROTT C R, LEBRUN-GRANDIÉ D, ARNDT D, et al. Kokkos 3: Programming model extensions for the exascale era[J]. IEEE Transactions on Parallel and Distributed Systems, 2022, 33(4): 805-817. doi: 10.1109/TPDS.2021.3097283
[16]	SLIWINSKI T. Performance portable fortran code using kokkos: A case study with the YSU PBL scheme[C]// 105th Annual AMS Meeting 2025, 2025: 454405[2025-11-17].
[17]	SHENDE S S, MALONY A D. The tau parallel performance system[J]. The International Journal of High Performance Computing Applications, 2006, 20(2): 287-311. doi: 10.1177/1094342006064482
[18]	HUCK K A, SHENDE S, MALONY A D, et al. Preparing the TAU performance system for exascale and beyond[J]. International Journal of High Performance Computing Applications, 2025, 39(4): 532-552. doi: 10.1177/10943420251334456
[19]	LINDLAN K A, CUNY J, MALONY A D, et al. A tool framework for static and dynamic analysis of object-oriented software with templates[C/OL]// SC’00: Proceedings of the 2000 ACM/IEEE Conference on Supercomputing, 2000: 49[2025-11-18]. https://ieeexplore.ieee.org/document/1592762.
[20]	ALLAN B A, ARMSTRONG R, BERNHOLDT D E, et al. A Component Architecture for High-Performance Scientific Computing[J]. International Journal of High Performance Computing Applications, 2006, 20(2): 163-202.. doi: 10.1177/1094342006064488
[21]	MALONY A, SHENDE S, TREBON N, et al. Performance technology for parallel and distributed component software[J]. Concurrency and Computation: Practice and Experience, 2005, 17(2-4): 117-141. doi: 10.1002/cpe.v17:2/4
[22]	HARVEY M J, DE FABRITIIS G. An implementation of the smooth particle mesh ewald method on GPU hardware[J]. Journal of Chemical Theory and Computation, 2009, 5(9): 2371-2377. doi: 10.1021/ct900275y pmid: 26616618
[23]	ESPOSITO R, MENSITIERI G, ZHOU Y, et al. GPU accelerated hybrid particle-field molecular dynamics: Multi-node/multi-GPU implementation and large-scale benchmarks of the OCCAM code[J]. Journal of Computational Chemistry, 2025, 46(13): e70126. doi: 10.1002/jcc.v46.13
[24]	SALOMON-FERRER R, GÖTZ A W, POOLE D, et al. Routine microsecond molecular dynamics simulations with AMBER on GPUs. 2. Explicit solvent particle mesh ewald[J]. Journal of Chemical Theory and Computation, 2013, 9(9): 3878-3888. doi: 10.1021/ct400314y
[25]	PÁLL S, ZHMUROV A, BAUER P, et al. Heterogeneous parallelization and acceleration of molecular dynamics simulations in GROMACS[J]. The Journal of Chemical Physics, 2020, 153(13): 134110. doi: 10.1063/5.0018516
[26]	PHILLIPS J C, BRAUN R, WANG W, et al. Scalable molecular dynamics with NAMD[J]. Journal of Computational Chemistry, 2005, 26(16): 1781-1802. doi: 10.1002/jcc.20289 pmid: 16222654
[27]	ACUN B, HARDY D J, KALE L V, et al. Scalable molecular dynamics with NAMD on the summit system[J]. IBM Journal of Research and Development, 2018, 62(6): 4:1-4:9. doi: 10.1147/jrd.2018.2888986 pmid: 32154805
[28]	PÁLL S, HESS B. A flexible algorithm for calculating pair interactions on SIMD architectures[J]. Computer Physics Communications, 2013, 184(12): 2641-2650. doi: 10.1016/j.cpc.2013.06.003
[29]	KÜHNE T D, IANNUZZI M, DEL BEN M, et al. CP2K: An electronic structure and molecular dynamics software package-Quickstep:Efficient and accurate electronic structure calculations[J/OL]. The Journal of Chemical Physics, 2020, 152(19)[2024-02-14]. https://pubs.aip.org/aip/jcp/article/152/19/194103/199081/CP2K-An-electronic-structure-and-molecular.
[30]	HACENE M, ANCIAUX-SEDRAKIAN A, ROZANSKA X, et al. Accelerating VASP electronic structure calculations using graphic processing units[J]. Journal of Computational Chemistry, 2012, 33(32): 2581-2589. doi: 10.1002/jcc.23096 pmid: 22903247
[31]	KŮS P, MAREK A, KÖCHER S S, et al. Optimizations of the eigensolvers in the ELPA library[J]. Parallel Computing, 2019, 85: 167-177. doi: 10.1016/j.parco.2019.04.003
[32]	MAREK A, BLUM V, JOHANNI R, et al. The ELPA library: Scalable parallel eigenvalue solutions for electronic structure theory and computational science[J]. Journal of Physics: Condensed Matter, 2014, 26(21): 213201. doi: 10.1088/0953-8984/26/21/213201
[33]	CHOI J, DEMMEL J, DHILLON I, et al. A portable linear algebra library for distributed memory computers—design issues and performance[J]. Computer Physics Communications, 1996, 97(1): 1-15.

组件	版本
DTK	23.10
GCC	7.3.1
MKL	2021.1.1
Boost	1.73.0
ELPA	2024.05
HPC-X MPI	2.11.0