Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer

Zihao Wang; Yu Chen; Jingrong Zhang; Lun Li; Xiaohua Wan; Zhiyong Liu; Fei Sun; Fa Zhang

doi:10.1007/978-3-319-59575-7_23

Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer

Zihao Wang, Yu Chen, Jingrong Zhang, Lun Li, Xiaohua Wan, Zhiyong Liu^*, Fei Sun, Fa Zhang

^*Corresponding author for this work

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

2 Citations (Scopus)

Abstract

Electron tomography (ET) is an important method for studying three-dimensional cell ultrastructure. Combining with a sub-volume averaging approach, ET provides new possibilities for investigating in situ macromolecular complexes in sub-nanometer resolution. Because of the limited sampling angles, ET reconstruction usually suffers from the ‘missing wedge’ problem. With a validation procedure, Iterative Compressed-sensing Optimized NUFFT reconstruction (ICON) demonstrates its power in the restoration of validated missing information for low SNR biological ET dataset. However, the huge computational demand has become a bottleneck for the application of ICON. In this work, we developed the strategies of parallelization for NUFFT and ICON, and then implemented them on a Xeon Phi 31SP coprocessor to generate the parallel program ICON-MIC. We also proposed a hybrid task allocation strategy and extended ICON-MIC on multiple Xeon Phi cards on Tianhe-2 supercomputer to generate program ICON-MULT-MIC. With high accuracy, ICON-MIC has a significant acceleration compared to the CPU version, up to 13.3x, and ICON-MULT-MIC has good weak and strong scalability efficiency on Tianhe-2 supercomputer.

Original language	English
Title of host publication	Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings
Editors	Zhipeng Cai, Ovidiu Daescu, Min Li
Publisher	Springer Verlag
Pages	258-269
Number of pages	12
ISBN (Print)	9783319595740
DOIs	https://doi.org/10.1007/978-3-319-59575-7_23
Publication status	Published - 2017
Externally published	Yes
Event	13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017 - Honolulu, United States Duration: 29 May 2017 → 2 Jun 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10330 LNBI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017
Country/Territory	United States
City	Honolulu
Period	29/05/17 → 2/06/17

Keywords

Electron tomography
Hybrid task allocation strategy
ICON
MIC acceleration
Parallel NUFFT
Tianhe-2 supercomputer

Access to Document

10.1007/978-3-319-59575-7_23

Cite this

Wang, Z., Chen, Y., Zhang, J., Li, L., Wan, X., Liu, Z., Sun, F., & Zhang, F. (2017). Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer. In Z. Cai, O. Daescu, & M. Li (Eds.), Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings (pp. 258-269). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10330 LNBI). Springer Verlag. https://doi.org/10.1007/978-3-319-59575-7_23

Wang, Zihao ; Chen, Yu ; Zhang, Jingrong et al. / Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer. Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings. editor / Zhipeng Cai ; Ovidiu Daescu ; Min Li. Springer Verlag, 2017. pp. 258-269 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{4dbc2c44f46845bea00a4468a6b28842,

title = "Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer",

abstract = "Electron tomography (ET) is an important method for studying three-dimensional cell ultrastructure. Combining with a sub-volume averaging approach, ET provides new possibilities for investigating in situ macromolecular complexes in sub-nanometer resolution. Because of the limited sampling angles, ET reconstruction usually suffers from the {\textquoteleft}missing wedge{\textquoteright} problem. With a validation procedure, Iterative Compressed-sensing Optimized NUFFT reconstruction (ICON) demonstrates its power in the restoration of validated missing information for low SNR biological ET dataset. However, the huge computational demand has become a bottleneck for the application of ICON. In this work, we developed the strategies of parallelization for NUFFT and ICON, and then implemented them on a Xeon Phi 31SP coprocessor to generate the parallel program ICON-MIC. We also proposed a hybrid task allocation strategy and extended ICON-MIC on multiple Xeon Phi cards on Tianhe-2 supercomputer to generate program ICON-MULT-MIC. With high accuracy, ICON-MIC has a significant acceleration compared to the CPU version, up to 13.3x, and ICON-MULT-MIC has good weak and strong scalability efficiency on Tianhe-2 supercomputer.",

keywords = "Electron tomography, Hybrid task allocation strategy, ICON, MIC acceleration, Parallel NUFFT, Tianhe-2 supercomputer",

author = "Zihao Wang and Yu Chen and Jingrong Zhang and Lun Li and Xiaohua Wan and Zhiyong Liu and Fei Sun and Fa Zhang",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017 ; Conference date: 29-05-2017 Through 02-06-2017",

year = "2017",

doi = "10.1007/978-3-319-59575-7_23",

language = "English",

isbn = "9783319595740",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "258--269",

editor = "Zhipeng Cai and Ovidiu Daescu and Min Li",

booktitle = "Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings",

address = "Germany",

}

Wang, Z, Chen, Y, Zhang, J, Li, L, Wan, X, Liu, Z, Sun, F & Zhang, F 2017, Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer. in Z Cai, O Daescu & M Li (eds), Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10330 LNBI, Springer Verlag, pp. 258-269, 13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017, Honolulu, United States, 29/05/17. https://doi.org/10.1007/978-3-319-59575-7_23

Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer. / Wang, Zihao; Chen, Yu; Zhang, Jingrong et al.
Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings. ed. / Zhipeng Cai; Ovidiu Daescu; Min Li. Springer Verlag, 2017. p. 258-269 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10330 LNBI).

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

TY - GEN

T1 - Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer

AU - Wang, Zihao

AU - Chen, Yu

AU - Zhang, Jingrong

AU - Li, Lun

AU - Wan, Xiaohua

AU - Liu, Zhiyong

AU - Sun, Fei

AU - Zhang, Fa

N1 - Publisher Copyright: © Springer International Publishing AG 2017.

PY - 2017

Y1 - 2017

N2 - Electron tomography (ET) is an important method for studying three-dimensional cell ultrastructure. Combining with a sub-volume averaging approach, ET provides new possibilities for investigating in situ macromolecular complexes in sub-nanometer resolution. Because of the limited sampling angles, ET reconstruction usually suffers from the ‘missing wedge’ problem. With a validation procedure, Iterative Compressed-sensing Optimized NUFFT reconstruction (ICON) demonstrates its power in the restoration of validated missing information for low SNR biological ET dataset. However, the huge computational demand has become a bottleneck for the application of ICON. In this work, we developed the strategies of parallelization for NUFFT and ICON, and then implemented them on a Xeon Phi 31SP coprocessor to generate the parallel program ICON-MIC. We also proposed a hybrid task allocation strategy and extended ICON-MIC on multiple Xeon Phi cards on Tianhe-2 supercomputer to generate program ICON-MULT-MIC. With high accuracy, ICON-MIC has a significant acceleration compared to the CPU version, up to 13.3x, and ICON-MULT-MIC has good weak and strong scalability efficiency on Tianhe-2 supercomputer.

AB - Electron tomography (ET) is an important method for studying three-dimensional cell ultrastructure. Combining with a sub-volume averaging approach, ET provides new possibilities for investigating in situ macromolecular complexes in sub-nanometer resolution. Because of the limited sampling angles, ET reconstruction usually suffers from the ‘missing wedge’ problem. With a validation procedure, Iterative Compressed-sensing Optimized NUFFT reconstruction (ICON) demonstrates its power in the restoration of validated missing information for low SNR biological ET dataset. However, the huge computational demand has become a bottleneck for the application of ICON. In this work, we developed the strategies of parallelization for NUFFT and ICON, and then implemented them on a Xeon Phi 31SP coprocessor to generate the parallel program ICON-MIC. We also proposed a hybrid task allocation strategy and extended ICON-MIC on multiple Xeon Phi cards on Tianhe-2 supercomputer to generate program ICON-MULT-MIC. With high accuracy, ICON-MIC has a significant acceleration compared to the CPU version, up to 13.3x, and ICON-MULT-MIC has good weak and strong scalability efficiency on Tianhe-2 supercomputer.

KW - Electron tomography

KW - Hybrid task allocation strategy

KW - ICON

KW - MIC acceleration

KW - Parallel NUFFT

KW - Tianhe-2 supercomputer

UR - http://www.scopus.com/inward/record.url?scp=85020715751&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-59575-7_23

DO - 10.1007/978-3-319-59575-7_23

M3 - Conference contribution

AN - SCOPUS:85020715751

SN - 9783319595740

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 258

EP - 269

BT - Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings

A2 - Cai, Zhipeng

A2 - Daescu, Ovidiu

A2 - Li, Min

PB - Springer Verlag

T2 - 13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017

Y2 - 29 May 2017 through 2 June 2017

ER -

Wang Z, Chen Y, Zhang J, Li L, Wan X, Liu Z et al. Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer. In Cai Z, Daescu O, Li M, editors, Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings. Springer Verlag. 2017. p. 258-269. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-59575-7_23

Accelerating electron tomography reconstruction algorithm ICON using the intel Xeon Phi coprocessor on Tianhe-2 supercomputer

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this