Tampere University of Technology

TUTCRIS Research Portal

Executing dataflow actors as kahn processes

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Standard

Executing dataflow actors as kahn processes. / Tretter, Andreas; Boutellier, Jani; Guthrie, James; Schor, Lars; Thiele, Lothar.

2015 Proceedings of the International Conference on Embedded Software, EMSOFT 2015. Institute of Electrical and Electronics Engineers Inc., 2015. p. 105-114 7318265.

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Harvard

Tretter, A, Boutellier, J, Guthrie, J, Schor, L & Thiele, L 2015, Executing dataflow actors as kahn processes. in 2015 Proceedings of the International Conference on Embedded Software, EMSOFT 2015., 7318265, Institute of Electrical and Electronics Engineers Inc., pp. 105-114, 13th International Conference on Embedded Software, EMSOFT 2015, Amsterdam, Netherlands, 4/10/15. https://doi.org/10.1109/EMSOFT.2015.7318265

APA

Tretter, A., Boutellier, J., Guthrie, J., Schor, L., & Thiele, L. (2015). Executing dataflow actors as kahn processes. In 2015 Proceedings of the International Conference on Embedded Software, EMSOFT 2015 (pp. 105-114). [7318265] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/EMSOFT.2015.7318265

Vancouver

Tretter A, Boutellier J, Guthrie J, Schor L, Thiele L. Executing dataflow actors as kahn processes. In 2015 Proceedings of the International Conference on Embedded Software, EMSOFT 2015. Institute of Electrical and Electronics Engineers Inc. 2015. p. 105-114. 7318265 https://doi.org/10.1109/EMSOFT.2015.7318265

Author

Tretter, Andreas ; Boutellier, Jani ; Guthrie, James ; Schor, Lars ; Thiele, Lothar. / Executing dataflow actors as kahn processes. 2015 Proceedings of the International Conference on Embedded Software, EMSOFT 2015. Institute of Electrical and Electronics Engineers Inc., 2015. pp. 105-114

Bibtex - Download

@inproceedings{7921e932d362415d84355800ac64e91c,
title = "Executing dataflow actors as kahn processes",
abstract = "Programming models which specify an application as a network of independent computational elements have emerged as a promising paradigm for programming streaming applications. The antagonism between expressivity and analysability has led to a number of different such programming models, which provide different degrees of freedom to the programmer. One example are Kahn process networks (KPNs), which, due to certain restrictions in communication, can guarantee determinacy (their results are independent of timing by construction). On the other hand, certain dataflow models, such as the CAL Actor Language, allow non-determinacy and thus higher expressivity, however at the price of static analysability and thus a potentially less efficient implementation. In many cases, however, non-determinacy is not required (or even not desired), and relying on KPN for the implementation seems advantageous. In this paper, we propose an algorithm for classifying dataflow actors (i.e. computational elements) as KPN compatible or potentially not. For KPN compatible dataflow actors, we propose an automatic KPN translation method based on this algorithm. In experiments, we show that more than 75{\%} of all mature actors of a standard multimedia benchmark suite can be classified as KPN compatible and that their execution time can be reduced by up to 1.97x using our proposed translation technique. Finally, in a manual classification effort, we validate these results and list different classes of KPN incompatibility.",
keywords = "classification, Dataflow programming, Kahn process networks",
author = "Andreas Tretter and Jani Boutellier and James Guthrie and Lars Schor and Lothar Thiele",
year = "2015",
month = "11",
day = "4",
doi = "10.1109/EMSOFT.2015.7318265",
language = "English",
pages = "105--114",
booktitle = "2015 Proceedings of the International Conference on Embedded Software, EMSOFT 2015",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

RIS (suitable for import to EndNote) - Download

TY - GEN

T1 - Executing dataflow actors as kahn processes

AU - Tretter, Andreas

AU - Boutellier, Jani

AU - Guthrie, James

AU - Schor, Lars

AU - Thiele, Lothar

PY - 2015/11/4

Y1 - 2015/11/4

N2 - Programming models which specify an application as a network of independent computational elements have emerged as a promising paradigm for programming streaming applications. The antagonism between expressivity and analysability has led to a number of different such programming models, which provide different degrees of freedom to the programmer. One example are Kahn process networks (KPNs), which, due to certain restrictions in communication, can guarantee determinacy (their results are independent of timing by construction). On the other hand, certain dataflow models, such as the CAL Actor Language, allow non-determinacy and thus higher expressivity, however at the price of static analysability and thus a potentially less efficient implementation. In many cases, however, non-determinacy is not required (or even not desired), and relying on KPN for the implementation seems advantageous. In this paper, we propose an algorithm for classifying dataflow actors (i.e. computational elements) as KPN compatible or potentially not. For KPN compatible dataflow actors, we propose an automatic KPN translation method based on this algorithm. In experiments, we show that more than 75% of all mature actors of a standard multimedia benchmark suite can be classified as KPN compatible and that their execution time can be reduced by up to 1.97x using our proposed translation technique. Finally, in a manual classification effort, we validate these results and list different classes of KPN incompatibility.

AB - Programming models which specify an application as a network of independent computational elements have emerged as a promising paradigm for programming streaming applications. The antagonism between expressivity and analysability has led to a number of different such programming models, which provide different degrees of freedom to the programmer. One example are Kahn process networks (KPNs), which, due to certain restrictions in communication, can guarantee determinacy (their results are independent of timing by construction). On the other hand, certain dataflow models, such as the CAL Actor Language, allow non-determinacy and thus higher expressivity, however at the price of static analysability and thus a potentially less efficient implementation. In many cases, however, non-determinacy is not required (or even not desired), and relying on KPN for the implementation seems advantageous. In this paper, we propose an algorithm for classifying dataflow actors (i.e. computational elements) as KPN compatible or potentially not. For KPN compatible dataflow actors, we propose an automatic KPN translation method based on this algorithm. In experiments, we show that more than 75% of all mature actors of a standard multimedia benchmark suite can be classified as KPN compatible and that their execution time can be reduced by up to 1.97x using our proposed translation technique. Finally, in a manual classification effort, we validate these results and list different classes of KPN incompatibility.

KW - classification

KW - Dataflow programming

KW - Kahn process networks

UR - http://www.scopus.com/inward/record.url?scp=84962325193&partnerID=8YFLogxK

U2 - 10.1109/EMSOFT.2015.7318265

DO - 10.1109/EMSOFT.2015.7318265

M3 - Conference contribution

SP - 105

EP - 114

BT - 2015 Proceedings of the International Conference on Embedded Software, EMSOFT 2015

PB - Institute of Electrical and Electronics Engineers Inc.

ER -