Practical design space exploration
(2019) 27th IEEE International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems, MASCOTS 2019 In Proceedings - IEEE Computer Society's Annual International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunications Systems, MASCOTS 2019-October. p.347-358- Abstract
Multi-objective optimization is a crucial matter in computer systems design space exploration because real-world applications often rely on a trade-off between several objectives. Derivatives are usually not available or impractical to compute and the feasibility of an experiment can not always be determined in advance. These problems are particularly difficult when the feasible region is relatively small, and it may be prohibitive to even find a feasible experiment, let alone an optimal one. We introduce a new methodology and corresponding software framework, HyperMapper 2.0, which handles multi-objective optimization, unknown feasibility constraints, and categorical/ordinal variables. This new methodology also supports injection of... (More)
Multi-objective optimization is a crucial matter in computer systems design space exploration because real-world applications often rely on a trade-off between several objectives. Derivatives are usually not available or impractical to compute and the feasibility of an experiment can not always be determined in advance. These problems are particularly difficult when the feasible region is relatively small, and it may be prohibitive to even find a feasible experiment, let alone an optimal one. We introduce a new methodology and corresponding software framework, HyperMapper 2.0, which handles multi-objective optimization, unknown feasibility constraints, and categorical/ordinal variables. This new methodology also supports injection of the user prior knowledge in the search when available. All of these features are common requirements in computer systems but rarely exposed in existing design space exploration systems. The proposed methodology follows a white-box model which is simple to understand and interpret (unlike, for example, neural networks) and can be used by the user to better understand the results of the automatic search. We apply and evaluate the new methodology to the automatic static tuning of hardware accelerators within the recently introduced Spatial programming language, with minimization of design run-time and compute logic under the constraint of the design fitting in a target field-programmable gate array chip. Our results show that HyperMapper 2.0 provides better Pareto fronts compared to state-of-the-art baselines, with better or competitive hypervolume indicator and with 8x improvement in sampling budget for most of the benchmarks explored.
(Less)
- author
- Nardi, Luigi LU ; Koeplinger, David and Olukotun, Kunle
- publishing date
- 2019-10
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- keywords
- Design space exploration, Hardware design, Machine learning driven optimization, Optimizing compilers, Pareto-optimal front, Performance modeling
- host publication
- 2019 IEEE 27th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems, MASCOTS 2019
- series title
- Proceedings - IEEE Computer Society's Annual International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunications Systems, MASCOTS
- volume
- 2019-October
- article number
- 8843094
- pages
- 12 pages
- publisher
- IEEE - Institute of Electrical and Electronics Engineers Inc.
- conference name
- 27th IEEE International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems, MASCOTS 2019
- conference location
- Rennes, France
- conference dates
- 2019-10-22 - 2019-10-25
- external identifiers
-
- scopus:85077791679
- ISSN
- 1526-7539
- ISBN
- 9781728149509
- DOI
- 10.1109/MASCOTS.2019.00045
- language
- English
- LU publication?
- no
- additional info
- Publisher Copyright: © 2019 IEEE.
- id
- 70256564-dce0-4bd0-b423-2d577f5e0fca
- date added to LUP
- 2022-09-16 18:05:24
- date last changed
- 2022-09-19 12:29:53
@inproceedings{70256564-dce0-4bd0-b423-2d577f5e0fca, abstract = {{<p>Multi-objective optimization is a crucial matter in computer systems design space exploration because real-world applications often rely on a trade-off between several objectives. Derivatives are usually not available or impractical to compute and the feasibility of an experiment can not always be determined in advance. These problems are particularly difficult when the feasible region is relatively small, and it may be prohibitive to even find a feasible experiment, let alone an optimal one. We introduce a new methodology and corresponding software framework, HyperMapper 2.0, which handles multi-objective optimization, unknown feasibility constraints, and categorical/ordinal variables. This new methodology also supports injection of the user prior knowledge in the search when available. All of these features are common requirements in computer systems but rarely exposed in existing design space exploration systems. The proposed methodology follows a white-box model which is simple to understand and interpret (unlike, for example, neural networks) and can be used by the user to better understand the results of the automatic search. We apply and evaluate the new methodology to the automatic static tuning of hardware accelerators within the recently introduced Spatial programming language, with minimization of design run-time and compute logic under the constraint of the design fitting in a target field-programmable gate array chip. Our results show that HyperMapper 2.0 provides better Pareto fronts compared to state-of-the-art baselines, with better or competitive hypervolume indicator and with 8x improvement in sampling budget for most of the benchmarks explored.</p>}}, author = {{Nardi, Luigi and Koeplinger, David and Olukotun, Kunle}}, booktitle = {{2019 IEEE 27th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems, MASCOTS 2019}}, isbn = {{9781728149509}}, issn = {{1526-7539}}, keywords = {{Design space exploration; Hardware design; Machine learning driven optimization; Optimizing compilers; Pareto-optimal front; Performance modeling}}, language = {{eng}}, pages = {{347--358}}, publisher = {{IEEE - Institute of Electrical and Electronics Engineers Inc.}}, series = {{Proceedings - IEEE Computer Society's Annual International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunications Systems, MASCOTS}}, title = {{Practical design space exploration}}, url = {{http://dx.doi.org/10.1109/MASCOTS.2019.00045}}, doi = {{10.1109/MASCOTS.2019.00045}}, volume = {{2019-October}}, year = {{2019}}, }