Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver

Saleen Bhattarai; Edwin Peters; Sean O'Byrne; David Petty

doi:10.1109/FCCM62733.2025.00073

Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver

Saleen Bhattarai, Edwin Peters, Sean O'Byrne, David Petty

School of Engineering

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

Abstract

We present an FPGA-based DSMC solver using a novel spatially-parallel collision evaluation algorithm, eliminating race conditions common in conventional DSMC methods. Our solver maintains high accuracy and achieves speedups of up to 5.5× over CPU and 1.5× over GPU implementations in low particle counts, particularly benefiting microscale flow scenarios requiring numerous repeated simulations.

Original language	English
Title of host publication	Proceedings - 2025 IEEE 33rd Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025
Place of Publication	Fayetteville, AR, USA
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	294-294
ISBN (Electronic)	979-8-3315-0281-2
ISBN (Print)	979-8-3315-0282-9
DOIs	https://doi.org/10.1109/FCCM62733.2025.00073
Publication status	Published - May 2025
Event	33rd IEEE Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025 - Fayetteville, United States Duration: 4 May 2025 → 7 May 2025 https://ieeexplore.ieee.org/xpl/conhome/11008901/proceeding

Publication series

Name	Proceedings of the Annual IEEE Symposium on Field-Programmable Custom Computing Machines
Publisher	IEEE Computer Society
ISSN (Print)	2576-2613
ISSN (Electronic)	2576-2621

Conference

Conference	33rd IEEE Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025
Country/Territory	United States
City	Fayetteville
Period	4/05/25 → 7/05/25
Internet address	https://ieeexplore.ieee.org/xpl/conhome/11008901/proceeding

Access to Document

10.1109/FCCM62733.2025.00073

Cite this

Bhattarai, S., Peters, E., O'Byrne, S., & Petty, D. (2025). Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver. In Proceedings - 2025 IEEE 33rd Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025 (pp. 294-294). (Proceedings of the Annual IEEE Symposium on Field-Programmable Custom Computing Machines). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/FCCM62733.2025.00073

Bhattarai, Saleen ; Peters, Edwin ; O'Byrne, Sean et al. / Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver. Proceedings - 2025 IEEE 33rd Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025. Fayetteville, AR, USA : Institute of Electrical and Electronics Engineers Inc., 2025. pp. 294-294 (Proceedings of the Annual IEEE Symposium on Field-Programmable Custom Computing Machines).

@inproceedings{f3069d3c468b4c3f91bd321e12c6a6a2,

title = "Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver",

abstract = "We present an FPGA-based DSMC solver using a novel spatially-parallel collision evaluation algorithm, eliminating race conditions common in conventional DSMC methods. Our solver maintains high accuracy and achieves speedups of up to 5.5× over CPU and 1.5× over GPU implementations in low particle counts, particularly benefiting microscale flow scenarios requiring numerous repeated simulations.",

keywords = "collision kernel, direct simulation monte carlo, dsmc, fpga, spatial parallelism",

author = "Saleen Bhattarai and Edwin Peters and Sean O'Byrne and David Petty",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 33rd IEEE Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025 ; Conference date: 04-05-2025 Through 07-05-2025",

year = "2025",

month = may,

doi = "10.1109/FCCM62733.2025.00073",

language = "English",

isbn = "979-8-3315-0282-9",

series = "Proceedings of the Annual IEEE Symposium on Field-Programmable Custom Computing Machines",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "294--294",

booktitle = "Proceedings - 2025 IEEE 33rd Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025",

address = "United States",

url = "https://ieeexplore.ieee.org/xpl/conhome/11008901/proceeding",

}

Bhattarai, S, Peters, E, O'Byrne, S & Petty, D 2025, Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver. in Proceedings - 2025 IEEE 33rd Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025. Proceedings of the Annual IEEE Symposium on Field-Programmable Custom Computing Machines, Institute of Electrical and Electronics Engineers Inc., Fayetteville, AR, USA, pp. 294-294, 33rd IEEE Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025, Fayetteville, Arkansas, United States, 4/05/25. https://doi.org/10.1109/FCCM62733.2025.00073

Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver. / Bhattarai, Saleen; Peters, Edwin; O'Byrne, Sean et al.
Proceedings - 2025 IEEE 33rd Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025. Fayetteville, AR, USA: Institute of Electrical and Electronics Engineers Inc., 2025. p. 294-294 (Proceedings of the Annual IEEE Symposium on Field-Programmable Custom Computing Machines).

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

TY - GEN

T1 - Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver

AU - Bhattarai, Saleen

AU - Peters, Edwin

AU - O'Byrne, Sean

AU - Petty, David

PY - 2025/5

Y1 - 2025/5

N2 - We present an FPGA-based DSMC solver using a novel spatially-parallel collision evaluation algorithm, eliminating race conditions common in conventional DSMC methods. Our solver maintains high accuracy and achieves speedups of up to 5.5× over CPU and 1.5× over GPU implementations in low particle counts, particularly benefiting microscale flow scenarios requiring numerous repeated simulations.

AB - We present an FPGA-based DSMC solver using a novel spatially-parallel collision evaluation algorithm, eliminating race conditions common in conventional DSMC methods. Our solver maintains high accuracy and achieves speedups of up to 5.5× over CPU and 1.5× over GPU implementations in low particle counts, particularly benefiting microscale flow scenarios requiring numerous repeated simulations.

KW - collision kernel

KW - direct simulation monte carlo

KW - dsmc

KW - fpga

KW - spatial parallelism

UR - https://www.scopus.com/pages/publications/105007998201

U2 - 10.1109/FCCM62733.2025.00073

DO - 10.1109/FCCM62733.2025.00073

M3 - Conference Paper

AN - SCOPUS:105007998201

SN - 979-8-3315-0282-9

T3 - Proceedings of the Annual IEEE Symposium on Field-Programmable Custom Computing Machines

SP - 294

EP - 294

BT - Proceedings - 2025 IEEE 33rd Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025

PB - Institute of Electrical and Electronics Engineers Inc.

CY - Fayetteville, AR, USA

T2 - 33rd IEEE Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025

Y2 - 4 May 2025 through 7 May 2025

ER -

Bhattarai S, Peters E, O'Byrne S, Petty D. Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver. In Proceedings - 2025 IEEE 33rd Annual International Symposium on Field-Programmable Custom Computing Machines, FCCM 2025. Fayetteville, AR, USA: Institute of Electrical and Electronics Engineers Inc. 2025. p. 294-294. (Proceedings of the Annual IEEE Symposium on Field-Programmable Custom Computing Machines). doi: 10.1109/FCCM62733.2025.00073

Performance Modeling and Comparisons of an FPGA-based Direct Simulation Monte Carlo Solver

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this