---
_id: '63890'
abstract:
- lang: eng
text: The computation of highly contracted electron repulsion integrals (ERIs) is
essential to achieve quantum accuracy in atomistic simulations based on quantum
mechanics. Its growing computational demands make energy efficiency a critical
concern. Recent studies demonstrate FPGAs’ superior performance and energy efficiency
for computing primitive ERIs, but the computation of highly contracted ERIs introduces
significant algorithmic complexity and new design challenges for FPGA acceleration.In
this work, we present SORCERI, the first streaming overlay acceleration for highly
contracted ERI computations on FPGAs. SORCERI introduces a novel streaming Rys
computing unit to calculate roots and weights of Rys polynomials on-chip, and
a streaming contraction unit for the contraction of primitive ERIs. This shifts
the design bottleneck from limited CPU-FPGA communication bandwidth to available
FPGA computation resources. To address practical deployment challenges for a large
number of quartet classes, we design three streaming overlays, together with an
efficient memory transpose optimization, to cover the 21 most commonly used quartet
classes in realistic atomistic simulations. To address the new computation constraints,
we use flexible calculation stages with a free-running streaming architecture
to achieve high DSP utilization and good timing closure.Experiments demonstrate
that SORCERI achieves an average 5.96x, 1.99x, and 1.16x better performance per
watt than libint on a 64-core AMD EPYC 7713 CPU, libintx on an Nvidia A40 GPU,
and SERI, the prior best-performing FPGA design for primitive ERIs. Furthermore,
SORCERI reaches a peak throughput of 44.11 GERIS (109 ERIs per second) that is
1.52x, 1.13x, and 1.93x greater than libint, libintx and SERI, respectively. SORCERI
will be released soon at https://github.com/SFU-HiAccel/SORCERI.
author:
- first_name: Philip
full_name: Stachura, Philip
last_name: Stachura
- first_name: Xin
full_name: Wu, Xin
id: '77439'
last_name: Wu
- first_name: Christian
full_name: Plessl, Christian
id: '16153'
last_name: Plessl
orcid: 0000-0001-5728-9982
- first_name: Zhenman
full_name: Fang, Zhenman
last_name: Fang
citation:
ama: 'Stachura P, Wu X, Plessl C, Fang Z. SORCERI: Streaming Overlay Acceleration
for Highly Contracted Electron Repulsion Integral Computations in Quantum Chemistry.
In: Proceedings of the 2026 ACM/SIGDA International Symposium on Field Programmable
Gate Arrays (FPGA ’26). Association for Computing Machinery; 2026:224-234.
doi:10.1145/3748173.3779198'
apa: 'Stachura, P., Wu, X., Plessl, C., & Fang, Z. (2026). SORCERI: Streaming
Overlay Acceleration for Highly Contracted Electron Repulsion Integral Computations
in Quantum Chemistry. Proceedings of the 2026 ACM/SIGDA International Symposium
on Field Programmable Gate Arrays (FPGA ’26), 224–234. https://doi.org/10.1145/3748173.3779198'
bibtex: '@inproceedings{Stachura_Wu_Plessl_Fang_2026, place={New York, NY, USA},
title={SORCERI: Streaming Overlay Acceleration for Highly Contracted Electron
Repulsion Integral Computations in Quantum Chemistry}, DOI={10.1145/3748173.3779198},
booktitle={Proceedings of the 2026 ACM/SIGDA International Symposium on Field
Programmable Gate Arrays (FPGA ’26)}, publisher={Association for Computing Machinery},
author={Stachura, Philip and Wu, Xin and Plessl, Christian and Fang, Zhenman},
year={2026}, pages={224–234} }'
chicago: 'Stachura, Philip, Xin Wu, Christian Plessl, and Zhenman Fang. “SORCERI:
Streaming Overlay Acceleration for Highly Contracted Electron Repulsion Integral
Computations in Quantum Chemistry.” In Proceedings of the 2026 ACM/SIGDA International
Symposium on Field Programmable Gate Arrays (FPGA ’26), 224–34. New York,
NY, USA: Association for Computing Machinery, 2026. https://doi.org/10.1145/3748173.3779198.'
ieee: 'P. Stachura, X. Wu, C. Plessl, and Z. Fang, “SORCERI: Streaming Overlay Acceleration
for Highly Contracted Electron Repulsion Integral Computations in Quantum Chemistry,”
in Proceedings of the 2026 ACM/SIGDA International Symposium on Field Programmable
Gate Arrays (FPGA ’26), 2026, pp. 224–234, doi: 10.1145/3748173.3779198.'
mla: 'Stachura, Philip, et al. “SORCERI: Streaming Overlay Acceleration for Highly
Contracted Electron Repulsion Integral Computations in Quantum Chemistry.” Proceedings
of the 2026 ACM/SIGDA International Symposium on Field Programmable Gate Arrays
(FPGA ’26), Association for Computing Machinery, 2026, pp. 224–34, doi:10.1145/3748173.3779198.'
short: 'P. Stachura, X. Wu, C. Plessl, Z. Fang, in: Proceedings of the 2026 ACM/SIGDA
International Symposium on Field Programmable Gate Arrays (FPGA ’26), Association
for Computing Machinery, New York, NY, USA, 2026, pp. 224–234.'
date_created: 2026-02-06T06:43:22Z
date_updated: 2026-02-09T09:16:32Z
department:
- _id: '27'
- _id: '518'
doi: 10.1145/3748173.3779198
keyword:
- electron repulsion integrals
- quantum chemistry
- atomistic simulation
- overlay architecture
- fpga acceleration
language:
- iso: eng
main_file_link:
- url: https://dl.acm.org/doi/10.1145/3748173.3779198
page: 224-234
place: New York, NY, USA
project:
- _id: '52'
name: Computing Resources Provided by the Paderborn Center for Parallel Computing
publication: Proceedings of the 2026 ACM/SIGDA International Symposium on Field Programmable
Gate Arrays (FPGA '26)
publication_identifier:
isbn:
- '9798400720796'
publication_status: published
publisher: Association for Computing Machinery
status: public
title: 'SORCERI: Streaming Overlay Acceleration for Highly Contracted Electron Repulsion
Integral Computations in Quantum Chemistry'
type: conference
user_id: '77439'
year: '2026'
...
---
_id: '64071'
abstract:
- lang: eng
text: Stimulated by the renewed interest and recent developments in semi-empirical
quantum chemical (SQC) methods for noncovalent interactions, we examine the properties
of liquid water at ambient conditions by means of molecular dynamics (MD) simulations,
both with the conventional NDDO-type (neglect of diatomic differential overlap)
methods, e.g. AM1 and PM6, and with DFTB-type (density-functional tight-binding)
methods, e.g. DFTB2 and GFN-xTB. Besides the original parameter sets, some specifically
reparametrized SQC methods (denoted as AM1-W, PM6-fm, and DFTB2-iBi) targeting
various smaller water systems ranging from molecular clusters to bulk are considered
as well. The quality of these different SQC methods for describing liquid water
properties at ambient conditions are assessed by comparison to well-established
experimental data and also to BLYP-D3 density functional theory-based ab initio
MD simulations. Our analyses reveal that static and dynamics properties of bulk
water are poorly described by all considered SQC methods with the original parameters,
regardless of the underlying theoretical models, with most of the methods suffering
from too weak hydrogen bonds and hence predicting a far too fluid water with highly
distorted hydrogen bond kinetics. On the other hand, the reparametrized force-matchcd
PM6-fm method is shown to be able to quantitatively reproduce the static and dynamic
features of liquid water, and thus can be used as a computationally efficient
alternative to electronic structure-based MD simulations for liquid water that
requires extended length and time scales. DFTB2-iBi predicts a slightly overstructured
water with reduced fluidity, whereas AM1-W gives an amorphous ice-like structure
for water at ambient conditions.
author:
- first_name: Xin
full_name: Wu, Xin
id: '77439'
last_name: Wu
- first_name: Hossam
full_name: Elgabarty, Hossam
id: '60250'
last_name: Elgabarty
orcid: 0000-0002-4945-1481
- first_name: Vahideh
full_name: Alizadeh, Vahideh
last_name: Alizadeh
- first_name: Andres
full_name: Henao Aristizabal, Andres
id: '67235'
last_name: Henao Aristizabal
- first_name: Frederik
full_name: Zysk, Frederik
id: '14757'
last_name: Zysk
- first_name: Christian
full_name: Plessl, Christian
id: '16153'
last_name: Plessl
orcid: 0000-0001-5728-9982
- first_name: Sebastian
full_name: Ehlert, Sebastian
last_name: Ehlert
- first_name: Jürg
full_name: Hutter, Jürg
last_name: Hutter
- first_name: Thomas D.
full_name: Kühne, Thomas D.
id: '49079'
last_name: Kühne
citation:
ama: Wu X, Elgabarty H, Alizadeh V, et al. Benchmarking semi-empirical quantum chemical
methods on liquid water. Published online 2025.
apa: Wu, X., Elgabarty, H., Alizadeh, V., Henao Aristizabal, A., Zysk, F., Plessl,
C., Ehlert, S., Hutter, J., & Kühne, T. D. (2025). Benchmarking semi-empirical
quantum chemical methods on liquid water.
bibtex: '@article{Wu_Elgabarty_Alizadeh_Henao Aristizabal_Zysk_Plessl_Ehlert_Hutter_Kühne_2025,
title={Benchmarking semi-empirical quantum chemical methods on liquid water},
author={Wu, Xin and Elgabarty, Hossam and Alizadeh, Vahideh and Henao Aristizabal,
Andres and Zysk, Frederik and Plessl, Christian and Ehlert, Sebastian and Hutter,
Jürg and Kühne, Thomas D.}, year={2025} }'
chicago: Wu, Xin, Hossam Elgabarty, Vahideh Alizadeh, Andres Henao Aristizabal,
Frederik Zysk, Christian Plessl, Sebastian Ehlert, Jürg Hutter, and Thomas D.
Kühne. “Benchmarking Semi-Empirical Quantum Chemical Methods on Liquid Water,”
2025.
ieee: X. Wu et al., “Benchmarking semi-empirical quantum chemical methods
on liquid water.” 2025.
mla: Wu, Xin, et al. Benchmarking Semi-Empirical Quantum Chemical Methods on
Liquid Water. 2025.
short: X. Wu, H. Elgabarty, V. Alizadeh, A. Henao Aristizabal, F. Zysk, C. Plessl,
S. Ehlert, J. Hutter, T.D. Kühne, (2025).
date_created: 2026-02-09T09:03:41Z
date_updated: 2026-02-09T09:17:07Z
department:
- _id: '27'
- _id: '2'
language:
- iso: eng
main_file_link:
- url: https://arxiv.org/abs/2503.11867
project:
- _id: '52'
name: Computing Resources Provided by the Paderborn Center for Parallel Computing
status: public
title: Benchmarking semi-empirical quantum chemical methods on liquid water
type: preprint
user_id: '77439'
year: '2025'
...
---
_id: '62981'
abstract:
- lang: eng
text: "Otus is a high-performance computing cluster that was launched in 2025 and
is operated by the Paderborn Center for Parallel Computing (PC2) at Paderborn
University in Germany. The system is part of the National High Performance Computing
(NHR) initiative. Otus complements the previous supercomputer Noctua 2, offering
approximately twice the computing power while retaining the three node types that
were characteristic of Noctua 2: 1) CPU compute nodes with different memory capacities,
2) high-end GPU nodes, and 3) HPC-grade FPGA nodes. On the Top500 list, which
ranks the 500 most powerful supercomputers in the world, Otus is in position 164
with the CPU partition and in position 255 with the GPU partition (June 2025).
On the Green500 list, ranking the 500 most energy-efficient supercomputers in
the world, Otus is in position 5 with the GPU partition (June 2025).\r\n\r\n\r\nThis
article provides a comprehensive overview of the system in terms of its hardware,
software, system integration, and its overall integration into the data center
building to ensure energy-efficient operation. The article aims to provide unique
insights for scientists using the system and for other centers operating HPC clusters.
The article will be continuously updated to reflect the latest system setup and
measurements. "
author:
- first_name: Sadaf
full_name: Ehtesabi, Sadaf
id: '116116'
last_name: Ehtesabi
- first_name: Manoar
full_name: Hossain, Manoar
id: '114619'
last_name: Hossain
orcid: https://orcid.org/0000-0002-0737-7981
- first_name: Tobias
full_name: Kenter, Tobias
id: '3145'
last_name: Kenter
- first_name: Andreas
full_name: Krawinkel, Andreas
id: '15275'
last_name: Krawinkel
- first_name: Lukas
full_name: Ostermann, Lukas
id: '69976'
last_name: Ostermann
- first_name: Christian
full_name: Plessl, Christian
id: '16153'
last_name: Plessl
orcid: 0000-0001-5728-9982
- first_name: Heinrich
full_name: Riebler, Heinrich
id: '8961'
last_name: Riebler
- first_name: Stefan
full_name: Rohde, Stefan
id: '34009'
last_name: Rohde
- first_name: Robert
full_name: Schade, Robert
id: '75963'
last_name: Schade
orcid: 0000-0002-6268-5397
- first_name: Michael
full_name: Schwarz, Michael
id: '5312'
last_name: Schwarz
- first_name: Jens
full_name: Simon, Jens
id: '15273'
last_name: Simon
- first_name: Nils
full_name: Winnwa, Nils
id: '61189'
last_name: Winnwa
- first_name: Alex
full_name: Wiens, Alex
id: '23522'
last_name: Wiens
orcid: 0000-0003-1764-9773
- first_name: Xin
full_name: Wu, Xin
id: '77439'
last_name: Wu
citation:
ama: Ehtesabi S, Hossain M, Kenter T, et al. Otus Supercomputer. Vol 1. Paderborn
Center for Parallel Computing (PC2); 2025. doi:10.48550/ARXIV.2512.07401
apa: Ehtesabi, S., Hossain, M., Kenter, T., Krawinkel, A., Ostermann, L., Plessl,
C., Riebler, H., Rohde, S., Schade, R., Schwarz, M., Simon, J., Winnwa, N., Wiens,
A., & Wu, X. (2025). Otus Supercomputer (Vol. 1). Paderborn Center
for Parallel Computing (PC2). https://doi.org/10.48550/ARXIV.2512.07401
bibtex: '@book{Ehtesabi_Hossain_Kenter_Krawinkel_Ostermann_Plessl_Riebler_Rohde_Schade_Schwarz_et
al._2025, place={Paderborn}, series={PC2 Technical Report Series}, title={Otus
Supercomputer}, volume={1}, DOI={10.48550/ARXIV.2512.07401},
publisher={Paderborn Center for Parallel Computing (PC2)}, author={Ehtesabi, Sadaf
and Hossain, Manoar and Kenter, Tobias and Krawinkel, Andreas and Ostermann, Lukas
and Plessl, Christian and Riebler, Heinrich and Rohde, Stefan and Schade, Robert
and Schwarz, Michael and et al.}, year={2025}, collection={PC2 Technical Report
Series} }'
chicago: 'Ehtesabi, Sadaf, Manoar Hossain, Tobias Kenter, Andreas Krawinkel, Lukas
Ostermann, Christian Plessl, Heinrich Riebler, et al. Otus Supercomputer.
Vol. 1. PC2 Technical Report Series. Paderborn: Paderborn Center for Parallel
Computing (PC2), 2025. https://doi.org/10.48550/ARXIV.2512.07401.'
ieee: 'S. Ehtesabi et al., Otus Supercomputer, vol. 1. Paderborn:
Paderborn Center for Parallel Computing (PC2), 2025.'
mla: Ehtesabi, Sadaf, et al. Otus Supercomputer. Paderborn Center for Parallel
Computing (PC2), 2025, doi:10.48550/ARXIV.2512.07401.
short: S. Ehtesabi, M. Hossain, T. Kenter, A. Krawinkel, L. Ostermann, C. Plessl,
H. Riebler, S. Rohde, R. Schade, M. Schwarz, J. Simon, N. Winnwa, A. Wiens, X.
Wu, Otus Supercomputer, Paderborn Center for Parallel Computing (PC2), Paderborn,
2025.
date_created: 2025-12-09T09:11:04Z
date_updated: 2026-03-25T11:50:31Z
ddc:
- '004'
department:
- _id: '27'
- _id: '518'
doi: 10.48550/ARXIV.2512.07401
file:
- access_level: open_access
content_type: application/pdf
creator: deffel
date_created: 2025-12-09T09:19:12Z
date_updated: 2026-03-25T11:50:30Z
file_id: '62982'
file_name: 2512.07401v1.pdf
file_size: 4535595
relation: main_file
file_date_updated: 2026-03-25T11:50:30Z
has_accepted_license: '1'
intvolume: ' 1'
keyword:
- Otus
- Supercomputer
- FPGA
- PC2
- Paderborn Center for Parallel Computing
- Noctua 2
- HPC
language:
- iso: eng
oa: '1'
page: '33'
place: Paderborn
publication_status: published
publisher: Paderborn Center for Parallel Computing (PC2)
report_number: PC2TR-2025-1
series_title: PC2 Technical Report Series
status: public
title: Otus Supercomputer
type: report
user_id: '23522'
volume: 1
year: '2025'
...
---
_id: '53663'
abstract:
- lang: eng
text: 'Noctua 2 is a supercomputer operated at the Paderborn Center for Parallel
Computing (PC2) at Paderborn University in Germany. Noctua 2 was inaugurated in
2022 and is an Atos BullSequana XH2000 system. It consists mainly of three node
types: 1) CPU Compute nodes with AMD EPYC processors in different main memory
configurations, 2) GPU nodes with NVIDIA A100 GPUs, and 3) FPGA nodes with Xilinx
Alveo U280 and Intel Stratix 10 FPGA cards. While CPUs and GPUs are known off-the-shelf
components in HPC systems, the operation of a large number of FPGA cards from
different vendors and a dedicated FPGA-to-FPGA network are unique characteristics
of Noctua 2. This paper describes in detail the overall setup of Noctua 2 and
gives insights into the operation of the cluster from a hardware, software and
facility perspective.'
article_type: original
author:
- first_name: Carsten
full_name: Bauer, Carsten
id: '90082'
last_name: Bauer
- first_name: Tobias
full_name: Kenter, Tobias
id: '3145'
last_name: Kenter
- first_name: Michael
full_name: Lass, Michael
id: '24135'
last_name: Lass
orcid: 0000-0002-5708-7632
- first_name: Lukas
full_name: Mazur, Lukas
id: '90492'
last_name: Mazur
orcid: ' 0000-0001-6304-7082'
- first_name: Marius
full_name: Meyer, Marius
id: '40778'
last_name: Meyer
- first_name: Holger
full_name: Nitsche, Holger
id: '15272'
last_name: Nitsche
- first_name: Heinrich
full_name: Riebler, Heinrich
id: '8961'
last_name: Riebler
- first_name: Robert
full_name: Schade, Robert
id: '75963'
last_name: Schade
orcid: 0000-0002-6268-5397
- first_name: Michael
full_name: Schwarz, Michael
id: '5312'
last_name: Schwarz
- first_name: Nils
full_name: Winnwa, Nils
id: '61189'
last_name: Winnwa
- first_name: Alex
full_name: Wiens, Alex
id: '23522'
last_name: Wiens
orcid: 0000-0003-1764-9773
- first_name: Xin
full_name: Wu, Xin
id: '77439'
last_name: Wu
- first_name: Christian
full_name: Plessl, Christian
id: '16153'
last_name: Plessl
orcid: 0000-0001-5728-9982
- first_name: Jens
full_name: Simon, Jens
id: '15273'
last_name: Simon
citation:
ama: Bauer C, Kenter T, Lass M, et al. Noctua 2 Supercomputer. Journal of large-scale
research facilities. 2024;9. doi:10.17815/jlsrf-8-187
apa: Bauer, C., Kenter, T., Lass, M., Mazur, L., Meyer, M., Nitsche, H., Riebler,
H., Schade, R., Schwarz, M., Winnwa, N., Wiens, A., Wu, X., Plessl, C., &
Simon, J. (2024). Noctua 2 Supercomputer. Journal of Large-Scale Research Facilities,
9. https://doi.org/10.17815/jlsrf-8-187
bibtex: '@article{Bauer_Kenter_Lass_Mazur_Meyer_Nitsche_Riebler_Schade_Schwarz_Winnwa_et
al._2024, title={Noctua 2 Supercomputer}, volume={9}, DOI={10.17815/jlsrf-8-187 }, journal={Journal of large-scale research facilities},
author={Bauer, Carsten and Kenter, Tobias and Lass, Michael and Mazur, Lukas and
Meyer, Marius and Nitsche, Holger and Riebler, Heinrich and Schade, Robert and
Schwarz, Michael and Winnwa, Nils and et al.}, year={2024} }'
chicago: Bauer, Carsten, Tobias Kenter, Michael Lass, Lukas Mazur, Marius Meyer,
Holger Nitsche, Heinrich Riebler, et al. “Noctua 2 Supercomputer.” Journal
of Large-Scale Research Facilities 9 (2024). https://doi.org/10.17815/jlsrf-8-187 .
ieee: 'C. Bauer et al., “Noctua 2 Supercomputer,” Journal of large-scale
research facilities, vol. 9, 2024, doi: 10.17815/jlsrf-8-187 .'
mla: Bauer, Carsten, et al. “Noctua 2 Supercomputer.” Journal of Large-Scale
Research Facilities, vol. 9, 2024, doi:10.17815/jlsrf-8-187 .
short: C. Bauer, T. Kenter, M. Lass, L. Mazur, M. Meyer, H. Nitsche, H. Riebler,
R. Schade, M. Schwarz, N. Winnwa, A. Wiens, X. Wu, C. Plessl, J. Simon, Journal
of Large-Scale Research Facilities 9 (2024).
date_created: 2024-04-26T07:39:41Z
date_updated: 2024-04-26T08:44:30Z
ddc:
- '004'
department:
- _id: '27'
- _id: '518'
doi: '10.17815/jlsrf-8-187 '
file:
- access_level: open_access
content_type: application/pdf
creator: deffel
date_created: 2024-04-26T07:30:20Z
date_updated: 2024-04-26T08:35:17Z
file_id: '53664'
file_name: Noctua2_Supercomputer.pdf
file_size: 3825480
relation: main_file
file_date_updated: 2024-04-26T08:35:17Z
has_accepted_license: '1'
intvolume: ' 9'
keyword:
- Noctua 2
- Supercomputer
- FPGA
- PC2
- Paderborn Center for Parallel Computing
language:
- iso: eng
oa: '1'
project:
- _id: '52'
name: 'PC2: Computing Resources Provided by the Paderborn Center for Parallel Computing'
publication: Journal of large-scale research facilities
publication_status: published
status: public
title: Noctua 2 Supercomputer
type: journal_article
user_id: '8961'
volume: 9
year: '2024'
...
---
_id: '56609'
abstract:
- lang: eng
text: 'The computation of electron repulsion integrals (ERIs) is a key component
for quantum chemical methods. The intensive computation and bandwidth demand for
ERI evaluation presents a significant challenge for quantum-mechanics-based atomistic
simulations with hybrid density functional theory: due to the tens of trillions
of ERI computations in each time step, practical applications are usually limited
to thousands of atoms. In this work, we propose SERI, a high-throughput streaming
accelerator for ERI computation on HBM-based FPGAs. In contrast to prior buffer-based
designs, SERI proposes a novel streaming architecture to address the on-chip buffer
limitation and the floorplanning challenge, and leverages the high-bandwidth memory
to overcome the bandwidth bottleneck in prior designs. Moreover, to meet the varying
computation, bandwidth, and floorplanning requirements between the 55 canonical
quartet classes in ERI calculation, we design an automation tool, together with
an accurate performance model, to automatically customize the architecture and
floorplanning strategy for each canonical quartet class to maximize their throughput.
Our performance evaluation on the AMD/Xilinx Alveo U280 FPGA board shows that,
SERI achieves an average speedup of 9.80 x over the previous best-performing FPGA
design, a 3.21x speedup over a 64-core AMD EPYC 7713 CPU, and a 15.64x speedup
over an Nvidia A40 GPU. It reaches a peak throughput of 23.8 GERIS ($10^9$ ERIs
per second) on one Alveo U280 FPGA. SERI will be released soon at https://github.com/SFU-HiAccel/SERI.'
author:
- first_name: Philip
full_name: Stachura, Philip
last_name: Stachura
- first_name: Guanyu
full_name: Li, Guanyu
last_name: Li
- first_name: Xin
full_name: Wu, Xin
id: '77439'
last_name: Wu
- first_name: Christian
full_name: Plessl, Christian
id: '16153'
last_name: Plessl
orcid: 0000-0001-5728-9982
- first_name: Zhenman
full_name: Fang, Zhenman
last_name: Fang
citation:
ama: 'Stachura P, Li G, Wu X, Plessl C, Fang Z. SERI: High-Throughput Streaming
Acceleration of Electron Repulsion Integral Computation in Quantum Chemistry using
HBM-based FPGAs. In: 2024 34th International Conference on Field-Programmable
Logic and Applications (FPL). IEEE; 2024:60-68. doi:10.1109/fpl64840.2024.00018'
apa: 'Stachura, P., Li, G., Wu, X., Plessl, C., & Fang, Z. (2024). SERI: High-Throughput
Streaming Acceleration of Electron Repulsion Integral Computation in Quantum Chemistry
using HBM-based FPGAs. 2024 34th International Conference on Field-Programmable
Logic and Applications (FPL), 60–68. https://doi.org/10.1109/fpl64840.2024.00018'
bibtex: '@inproceedings{Stachura_Li_Wu_Plessl_Fang_2024, title={SERI: High-Throughput
Streaming Acceleration of Electron Repulsion Integral Computation in Quantum Chemistry
using HBM-based FPGAs}, DOI={10.1109/fpl64840.2024.00018},
booktitle={2024 34th International Conference on Field-Programmable Logic and
Applications (FPL)}, publisher={IEEE}, author={Stachura, Philip and Li, Guanyu
and Wu, Xin and Plessl, Christian and Fang, Zhenman}, year={2024}, pages={60–68}
}'
chicago: 'Stachura, Philip, Guanyu Li, Xin Wu, Christian Plessl, and Zhenman Fang.
“SERI: High-Throughput Streaming Acceleration of Electron Repulsion Integral Computation
in Quantum Chemistry Using HBM-Based FPGAs.” In 2024 34th International Conference
on Field-Programmable Logic and Applications (FPL), 60–68. IEEE, 2024. https://doi.org/10.1109/fpl64840.2024.00018.'
ieee: 'P. Stachura, G. Li, X. Wu, C. Plessl, and Z. Fang, “SERI: High-Throughput
Streaming Acceleration of Electron Repulsion Integral Computation in Quantum Chemistry
using HBM-based FPGAs,” in 2024 34th International Conference on Field-Programmable
Logic and Applications (FPL), 2024, pp. 60–68, doi: 10.1109/fpl64840.2024.00018.'
mla: 'Stachura, Philip, et al. “SERI: High-Throughput Streaming Acceleration of
Electron Repulsion Integral Computation in Quantum Chemistry Using HBM-Based FPGAs.”
2024 34th International Conference on Field-Programmable Logic and Applications
(FPL), IEEE, 2024, pp. 60–68, doi:10.1109/fpl64840.2024.00018.'
short: 'P. Stachura, G. Li, X. Wu, C. Plessl, Z. Fang, in: 2024 34th International
Conference on Field-Programmable Logic and Applications (FPL), IEEE, 2024, pp.
60–68.'
date_created: 2024-10-14T08:44:44Z
date_updated: 2024-10-15T08:37:27Z
department:
- _id: '27'
- _id: '518'
doi: 10.1109/fpl64840.2024.00018
language:
- iso: eng
main_file_link:
- url: https://ieeexplore.ieee.org/document/10705609
page: 60-68
project:
- _id: '52'
name: 'PC2: Computing Resources Provided by the Paderborn Center for Parallel Computing'
publication: 2024 34th International Conference on Field-Programmable Logic and Applications
(FPL)
publication_status: published
publisher: IEEE
quality_controlled: '1'
status: public
title: 'SERI: High-Throughput Streaming Acceleration of Electron Repulsion Integral
Computation in Quantum Chemistry using HBM-based FPGAs'
type: conference
user_id: '77439'
year: '2024'
...
---
_id: '43228'
abstract:
- lang: eng
text: "The computation of electron repulsion integrals (ERIs) over Gaussian-type
orbitals (GTOs) is a challenging problem in quantum-mechanics-based atomistic
simulations. In practical simulations, several trillions of ERIs may have to be\r\ncomputed
for every time step.\r\nIn this work, we investigate FPGAs as accelerators for
the ERI computation. We use template parameters, here within the Intel oneAPI
tool flow, to create customized designs for 256 different ERI quartet classes,
based on their orbitals. To maximize data reuse, all intermediates are buffered
in FPGA on-chip memory with customized layout. The pre-calculation of intermediates
also helps to overcome data dependencies caused by multi-dimensional recurrence\r\nrelations.
The involved loop structures are partially or even fully unrolled for high throughput
of FPGA kernels. Furthermore, a lossy compression algorithm utilizing arbitrary
bitwidth integers is integrated in the FPGA kernels. To our\r\nbest knowledge,
this is the first work on ERI computation on FPGAs that supports more than just
the single most basic quartet class. Also, the integration of ERI computation
and compression it a novelty that is not even covered by CPU or GPU libraries
so far.\r\nOur evaluation shows that using 16-bit integer for the ERI compression,
the fastest FPGA kernels exceed the performance of 10 GERIS ($10 \\times 10^9$
ERIs per second) on one Intel Stratix 10 GX 2800 FPGA, with maximum absolute errors
around $10^{-7}$ - $10^{-5}$ Hartree. The measured throughput can be accurately
explained by a performance model. The FPGA kernels deployed on 2 FPGAs outperform
similar computations using the widely used libint reference on a two-socket server
with 40 Xeon Gold 6148 CPU cores of the same process technology by factors up
to 6.0x and on a new two-socket server with 128 EPYC 7713 CPU cores by up to 1.9x."
author:
- first_name: Xin
full_name: Wu, Xin
id: '77439'
last_name: Wu
- first_name: Tobias
full_name: Kenter, Tobias
id: '3145'
last_name: Kenter
- first_name: Robert
full_name: Schade, Robert
id: '75963'
last_name: Schade
orcid: 0000-0002-6268-539
- first_name: Thomas
full_name: Kühne, Thomas
id: '49079'
last_name: Kühne
- first_name: Christian
full_name: Plessl, Christian
id: '16153'
last_name: Plessl
orcid: 0000-0001-5728-9982
citation:
ama: 'Wu X, Kenter T, Schade R, Kühne T, Plessl C. Computing and Compressing Electron
Repulsion Integrals on FPGAs. In: 2023 IEEE 31st Annual International Symposium
on Field-Programmable Custom Computing Machines (FCCM). ; 2023:162-173. doi:10.1109/FCCM57271.2023.00026'
apa: Wu, X., Kenter, T., Schade, R., Kühne, T., & Plessl, C. (2023). Computing
and Compressing Electron Repulsion Integrals on FPGAs. 2023 IEEE 31st Annual
International Symposium on Field-Programmable Custom Computing Machines (FCCM),
162–173. https://doi.org/10.1109/FCCM57271.2023.00026
bibtex: '@inproceedings{Wu_Kenter_Schade_Kühne_Plessl_2023, title={Computing and
Compressing Electron Repulsion Integrals on FPGAs}, DOI={10.1109/FCCM57271.2023.00026},
booktitle={2023 IEEE 31st Annual International Symposium on Field-Programmable
Custom Computing Machines (FCCM)}, author={Wu, Xin and Kenter, Tobias and Schade,
Robert and Kühne, Thomas and Plessl, Christian}, year={2023}, pages={162–173}
}'
chicago: Wu, Xin, Tobias Kenter, Robert Schade, Thomas Kühne, and Christian Plessl.
“Computing and Compressing Electron Repulsion Integrals on FPGAs.” In 2023
IEEE 31st Annual International Symposium on Field-Programmable Custom Computing
Machines (FCCM), 162–73, 2023. https://doi.org/10.1109/FCCM57271.2023.00026.
ieee: 'X. Wu, T. Kenter, R. Schade, T. Kühne, and C. Plessl, “Computing and Compressing
Electron Repulsion Integrals on FPGAs,” in 2023 IEEE 31st Annual International
Symposium on Field-Programmable Custom Computing Machines (FCCM), 2023, pp.
162–173, doi: 10.1109/FCCM57271.2023.00026.'
mla: Wu, Xin, et al. “Computing and Compressing Electron Repulsion Integrals on
FPGAs.” 2023 IEEE 31st Annual International Symposium on Field-Programmable
Custom Computing Machines (FCCM), 2023, pp. 162–73, doi:10.1109/FCCM57271.2023.00026.
short: 'X. Wu, T. Kenter, R. Schade, T. Kühne, C. Plessl, in: 2023 IEEE 31st Annual
International Symposium on Field-Programmable Custom Computing Machines (FCCM),
2023, pp. 162–173.'
date_created: 2023-03-30T11:15:40Z
date_updated: 2023-08-02T15:05:42Z
department:
- _id: '27'
- _id: '518'
doi: 10.1109/FCCM57271.2023.00026
external_id:
arxiv:
- '2303.13632'
language:
- iso: eng
main_file_link:
- url: https://ieeexplore.ieee.org/document/10171537
page: 162-173
project:
- _id: '52'
name: 'PC2: Computing Resources Provided by the Paderborn Center for Parallel Computing'
publication: 2023 IEEE 31st Annual International Symposium on Field-Programmable Custom
Computing Machines (FCCM)
quality_controlled: '1'
status: public
title: Computing and Compressing Electron Repulsion Integrals on FPGAs
type: conference
user_id: '75963'
year: '2023'
...