Many streaming applications composed of multiple tasks self-adapt their tasks' execution at runtime as response to the processed data. This type of application promises a better solution to context switches at the cost of a non-deterministic task scheduling. Partial reconfiguration is a unique feature of FPGAs that not only offers a higher resource reuse but also performance improvements when properly applied. In this paper, a probabilistic approach is used to estimate the acceleration of streaming applications with unknown task schedule thanks to the application of partial reconfiguration. This novel approach provides insights in the feasible acceleration when regions of the FPGA are partially reconfigured in order to exploit the available resources by processing multiple tasks in parallel. Moreover, the impact of how different strategies or heuristics affect to the final performance is included in this analysis. As a result, not only an estimation of the achievable acceleration is obtained, but also a guide at the design stage when searching for the highest performance.
Da Silva Gomez, B, Braeken, A & Touhafi, A 2019, Probabilistic Performance Modelling when using Partial Reconfiguration to Accelerate Streaming Applications with Non-Deterministic Task Scheduling. in P Diniz, R Woods, C Hochberger, A Koch & B Nelson (eds), Applied Reconfigurable Computing - 15th International Symposium, ARC 2019, Proceedings., 1, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11444 LNCS, pp. 81-95, International Symposium on Applied Reconfigurable Computing, Darmstadt, Germany, 9/04/19. https://doi.org/10.1007/978-3-030-17227-5_7
Da Silva Gomez, B., Braeken, A., & Touhafi, A. (2019). Probabilistic Performance Modelling when using Partial Reconfiguration to Accelerate Streaming Applications with Non-Deterministic Task Scheduling. In P. Diniz, R. Woods, C. Hochberger, A. Koch, & B. Nelson (Eds.), Applied Reconfigurable Computing - 15th International Symposium, ARC 2019, Proceedings (pp. 81-95). Article 1 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11444 LNCS). https://doi.org/10.1007/978-3-030-17227-5_7
@inproceedings{aee46803950843078ba94dec3c5a4ea5,
title = "Probabilistic Performance Modelling when using Partial Reconfiguration to Accelerate Streaming Applications with Non-Deterministic Task Scheduling",
abstract = "Many streaming applications composed of multiple tasks self-adapt their tasks' execution at runtime as response to the processed data. This type of application promises a better solution to context switches at the cost of a non-deterministic task scheduling. Partial reconfiguration is a unique feature of FPGAs that not only offers a higher resource reuse but also performance improvements when properly applied. In this paper, a probabilistic approach is used to estimate the acceleration of streaming applications with unknown task schedule thanks to the application of partial reconfiguration. This novel approach provides insights in the feasible acceleration when regions of the FPGA are partially reconfigured in order to exploit the available resources by processing multiple tasks in parallel. Moreover, the impact of how different strategies or heuristics affect to the final performance is included in this analysis. As a result, not only an estimation of the achievable acceleration is obtained, but also a guide at the design stage when searching for the highest performance.",
keywords = "Partial Reconfiguration, FPGA, Probabilistic Performance Model, Performance Estimation, Streaming Applications",
author = "{Da Silva Gomez}, Bruno and An Braeken and Abdellah Touhafi",
year = "2019",
month = apr,
day = "9",
doi = "10.1007/978-3-030-17227-5_7",
language = "English",
isbn = "978-3-030-17227-5",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "81--95",
editor = "Pedro Diniz and Roger Woods and Christian Hochberger and Andreas Koch and Brent Nelson",
booktitle = "Applied Reconfigurable Computing - 15th International Symposium, ARC 2019, Proceedings",
note = "International Symposium on Applied Reconfigurable Computing, ARC ; Conference date: 09-04-2019 Through 11-04-2019",
url = "https://www.arc2019.tu-darmstadt.de/",
}