---
_id: '1588'
abstract:
- lang: eng
text: The exploration of FPGAs as accelerators for scientific simulations has so
far mostly been focused on small kernels of methods working on regular data structures,
for example in the form of stencil computations for finite difference methods.
In computational sciences, often more advanced methods are employed that promise
better stability, convergence, locality and scaling. Unstructured meshes are shown
to be more effective and more accurate, compared to regular grids, in representing
computation domains of various shapes. Using unstructured meshes, the discontinuous
Galerkin method preserves the ability to perform explicit local update operations
for simulations in the time domain. In this work, we investigate FPGAs as target
platform for an implementation of the nodal discontinuous Galerkin method to find
time-domain solutions of Maxwell's equations in an unstructured mesh. When maximizing
data reuse and fitting constant coefficients into suitably partitioned on-chip
memory, high computational intensity allows us to implement and feed wide data
paths with hundreds of floating point operators. By decoupling off-chip memory
accesses from the computations, high memory bandwidth can be sustained, even for
the irregular access pattern required by parts of the application. Using the Intel/Altera
OpenCL SDK for FPGAs, we present different implementation variants for different
polynomial orders of the method. In different phases of the algorithm, either
computational or bandwidth limits of the Arria 10 platform are almost reached,
thus outperforming a highly multithreaded CPU implementation by around 2x.
author:
- first_name: Tobias
full_name: Kenter, Tobias
id: '3145'
last_name: Kenter
- first_name: Gopinath
full_name: Mahale, Gopinath
last_name: Mahale
- first_name: Samer
full_name: Alhaddad, Samer
id: '42456'
last_name: Alhaddad
- first_name: Yevgen
full_name: Grynko, Yevgen
id: '26059'
last_name: Grynko
- first_name: Christian
full_name: Schmitt, Christian
last_name: Schmitt
- first_name: Ayesha
full_name: Afzal, Ayesha
last_name: Afzal
- first_name: Frank
full_name: Hannig, Frank
last_name: Hannig
- first_name: Jens
full_name: Förstner, Jens
id: '158'
last_name: Förstner
orcid: 0000-0001-7059-9862
- first_name: Christian
full_name: Plessl, Christian
id: '16153'
last_name: Plessl
orcid: 0000-0001-5728-9982
citation:
ama: 'Kenter T, Mahale G, Alhaddad S, et al. OpenCL-based FPGA Design to Accelerate
the Nodal Discontinuous Galerkin Method for Unstructured Meshes. In: Proc.
Int. Symp. on Field-Programmable Custom Computing Machines (FCCM). IEEE; 2018.
doi:10.1109/FCCM.2018.00037'
apa: Kenter, T., Mahale, G., Alhaddad, S., Grynko, Y., Schmitt, C., Afzal, A., Hannig,
F., Förstner, J., & Plessl, C. (2018). OpenCL-based FPGA Design to Accelerate
the Nodal Discontinuous Galerkin Method for Unstructured Meshes. Proc. Int.
Symp. on Field-Programmable Custom Computing Machines (FCCM). Proc. Int. Symp.
on Field-Programmable Custom Computing Machines (FCCM). https://doi.org/10.1109/FCCM.2018.00037
bibtex: '@inproceedings{Kenter_Mahale_Alhaddad_Grynko_Schmitt_Afzal_Hannig_Förstner_Plessl_2018,
title={OpenCL-based FPGA Design to Accelerate the Nodal Discontinuous Galerkin
Method for Unstructured Meshes}, DOI={10.1109/FCCM.2018.00037},
booktitle={Proc. Int. Symp. on Field-Programmable Custom Computing Machines (FCCM)},
publisher={IEEE}, author={Kenter, Tobias and Mahale, Gopinath and Alhaddad, Samer
and Grynko, Yevgen and Schmitt, Christian and Afzal, Ayesha and Hannig, Frank
and Förstner, Jens and Plessl, Christian}, year={2018} }'
chicago: Kenter, Tobias, Gopinath Mahale, Samer Alhaddad, Yevgen Grynko, Christian
Schmitt, Ayesha Afzal, Frank Hannig, Jens Förstner, and Christian Plessl. “OpenCL-Based
FPGA Design to Accelerate the Nodal Discontinuous Galerkin Method for Unstructured
Meshes.” In Proc. Int. Symp. on Field-Programmable Custom Computing Machines
(FCCM). IEEE, 2018. https://doi.org/10.1109/FCCM.2018.00037.
ieee: 'T. Kenter et al., “OpenCL-based FPGA Design to Accelerate the Nodal
Discontinuous Galerkin Method for Unstructured Meshes,” presented at the Proc.
Int. Symp. on Field-Programmable Custom Computing Machines (FCCM), 2018, doi:
10.1109/FCCM.2018.00037.'
mla: Kenter, Tobias, et al. “OpenCL-Based FPGA Design to Accelerate the Nodal Discontinuous
Galerkin Method for Unstructured Meshes.” Proc. Int. Symp. on Field-Programmable
Custom Computing Machines (FCCM), IEEE, 2018, doi:10.1109/FCCM.2018.00037.
short: 'T. Kenter, G. Mahale, S. Alhaddad, Y. Grynko, C. Schmitt, A. Afzal, F. Hannig,
J. Förstner, C. Plessl, in: Proc. Int. Symp. on Field-Programmable Custom Computing
Machines (FCCM), IEEE, 2018.'
conference:
name: Proc. Int. Symp. on Field-Programmable Custom Computing Machines (FCCM)
date_created: 2018-03-22T10:48:01Z
date_updated: 2023-09-26T11:47:52Z
ddc:
- '000'
department:
- _id: '27'
- _id: '518'
- _id: '61'
doi: 10.1109/FCCM.2018.00037
file:
- access_level: closed
content_type: application/pdf
creator: ups
date_created: 2018-11-02T14:45:05Z
date_updated: 2018-11-02T14:45:05Z
file_id: '5282'
file_name: 08457652.pdf
file_size: 269130
relation: main_file
success: 1
file_date_updated: 2018-11-02T14:45:05Z
has_accepted_license: '1'
keyword:
- tet_topic_hpc
language:
- iso: eng
project:
- _id: '33'
grant_number: 01|H16005A
name: HighPerMeshes
- _id: '1'
grant_number: '160364472'
name: SFB 901
- _id: '4'
name: SFB 901 - Project Area C
- _id: '14'
grant_number: '160364472'
name: SFB 901 - Subproject C2
publication: Proc. Int. Symp. on Field-Programmable Custom Computing Machines (FCCM)
publisher: IEEE
quality_controlled: '1'
status: public
title: OpenCL-based FPGA Design to Accelerate the Nodal Discontinuous Galerkin Method
for Unstructured Meshes
type: conference
user_id: '15278'
year: '2018'
...