A high performance implementation for Molecular Dynamics simulations on a FPGA supercomputer

S. Kasap; Khaled Benkrid

doi:10.1109/AHS.2011.5963962

A high performance implementation for Molecular Dynamics simulations on a FPGA supercomputer

S. Kasap, Khaled Benkrid

Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding › peer-review

2 Citations (Scopus)

Abstract

The design and implementation of an FPGA core that parallelises all the necessary operations to compute the non-bonded interactions in a MD simulation with the purpose of accelerating the LAMMPS MD software is presented in this paper. Our MD processor core comprised of 4 identical pipelines working independently in parallel to evaluate the non-bonded potentials, forces and virials was implemented on the nodes of a FPGA-based supercomputer, namely Maxwell. Implementing our FPGA core on multiple nodes of Maxwell allowed us to produce a special-purpose parallel machine for the hard-ware acceleration of MD simulations. The timing performance figures of this machine for the pairwise LJ and short-range Coulombic (via PPPM) interaction computations in the MD simulations of the solvated Rhodopsin protein systems show performance gains over the pure software implementation by factors of up to 14 on two or more nodes of the Maxwell machine.

Original language	English
Title of host publication	Proceedings of the 2011 NASA/ESA Conference on Adaptive Hardware and Systems, AHS 2011
Publisher	IEEE
Pages	375-382
Number of pages	8
ISBN (Electronic)	978-1-4577-0599-1
ISBN (Print)	978-1-4577-0598-4
DOIs	https://doi.org/10.1109/AHS.2011.5963962
Publication status	Published - 29 Jul 2011
Externally published	Yes
Event	2011 NASA/ESA Conference on Adaptive Hardware and Systems - San Diego, United States Duration: 6 Jun 2011 → 9 Jun 2011

Conference

Conference	2011 NASA/ESA Conference on Adaptive Hardware and Systems
Country/Territory	United States
City	San Diego
Period	6/06/11 → 9/06/11

Access to Document

10.1109/AHS.2011.5963962

Cite this

Kasap, S & Benkrid, K 2011, A high performance implementation for Molecular Dynamics simulations on a FPGA supercomputer. in Proceedings of the 2011 NASA/ESA Conference on Adaptive Hardware and Systems, AHS 2011. IEEE, pp. 375-382, 2011 NASA/ESA Conference on Adaptive Hardware and Systems , San Diego, California, United States, 6/06/11. https://doi.org/10.1109/AHS.2011.5963962

@inproceedings{c867875201b7405eaac73f99e595437e,

title = "A high performance implementation for Molecular Dynamics simulations on a FPGA supercomputer",

abstract = "The design and implementation of an FPGA core that parallelises all the necessary operations to compute the non-bonded interactions in a MD simulation with the purpose of accelerating the LAMMPS MD software is presented in this paper. Our MD processor core comprised of 4 identical pipelines working independently in parallel to evaluate the non-bonded potentials, forces and virials was implemented on the nodes of a FPGA-based supercomputer, namely Maxwell. Implementing our FPGA core on multiple nodes of Maxwell allowed us to produce a special-purpose parallel machine for the hard-ware acceleration of MD simulations. The timing performance figures of this machine for the pairwise LJ and short-range Coulombic (via PPPM) interaction computations in the MD simulations of the solvated Rhodopsin protein systems show performance gains over the pure software implementation by factors of up to 14 on two or more nodes of the Maxwell machine.",

author = "S. Kasap and Khaled Benkrid",

year = "2011",

month = jul,

day = "29",

doi = "10.1109/AHS.2011.5963962",

language = "English",

isbn = " 978-1-4577-0598-4",

pages = "375--382",

booktitle = "Proceedings of the 2011 NASA/ESA Conference on Adaptive Hardware and Systems, AHS 2011",

publisher = "IEEE",

address = "United States",

note = "2011 NASA/ESA Conference on Adaptive Hardware and Systems ; Conference date: 06-06-2011 Through 09-06-2011",

}

TY - GEN

T1 - A high performance implementation for Molecular Dynamics simulations on a FPGA supercomputer

AU - Kasap, S.

AU - Benkrid, Khaled

PY - 2011/7/29

Y1 - 2011/7/29

N2 - The design and implementation of an FPGA core that parallelises all the necessary operations to compute the non-bonded interactions in a MD simulation with the purpose of accelerating the LAMMPS MD software is presented in this paper. Our MD processor core comprised of 4 identical pipelines working independently in parallel to evaluate the non-bonded potentials, forces and virials was implemented on the nodes of a FPGA-based supercomputer, namely Maxwell. Implementing our FPGA core on multiple nodes of Maxwell allowed us to produce a special-purpose parallel machine for the hard-ware acceleration of MD simulations. The timing performance figures of this machine for the pairwise LJ and short-range Coulombic (via PPPM) interaction computations in the MD simulations of the solvated Rhodopsin protein systems show performance gains over the pure software implementation by factors of up to 14 on two or more nodes of the Maxwell machine.

AB - The design and implementation of an FPGA core that parallelises all the necessary operations to compute the non-bonded interactions in a MD simulation with the purpose of accelerating the LAMMPS MD software is presented in this paper. Our MD processor core comprised of 4 identical pipelines working independently in parallel to evaluate the non-bonded potentials, forces and virials was implemented on the nodes of a FPGA-based supercomputer, namely Maxwell. Implementing our FPGA core on multiple nodes of Maxwell allowed us to produce a special-purpose parallel machine for the hard-ware acceleration of MD simulations. The timing performance figures of this machine for the pairwise LJ and short-range Coulombic (via PPPM) interaction computations in the MD simulations of the solvated Rhodopsin protein systems show performance gains over the pure software implementation by factors of up to 14 on two or more nodes of the Maxwell machine.

UR - http://www.scopus.com/inward/record.url?eid=2-s2.0-80052133904&partnerID=MN8TOARS

U2 - 10.1109/AHS.2011.5963962

DO - 10.1109/AHS.2011.5963962

M3 - Conference proceeding

SN - 978-1-4577-0598-4

SP - 375

EP - 382

BT - Proceedings of the 2011 NASA/ESA Conference on Adaptive Hardware and Systems, AHS 2011

PB - IEEE

T2 - 2011 NASA/ESA Conference on Adaptive Hardware and Systems

Y2 - 6 June 2011 through 9 June 2011

ER -

A high performance implementation for Molecular Dynamics simulations on a FPGA supercomputer

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this