Project detail

Přibližná ekvivalence pro aproximativní počítání

Duration: 01.01.2016 — 31.12.2018

Funding resources

Czech Science Foundation - Standardní projekty

- part funder (2016-01-01 - 2018-12-31)

On the project

Přibližné počítání je velmi slibným přístupem k vývoji energeticky úsporných výpočetních systémů. Využívá se při něm skutečnosti, že v řadě aplikací je možno tolerovat jistou chybu výsledků. Otevřeným problémem zůstává, jak efektivně vyvíjet aproximace systémů, které by byly dobrým kompromisem mezi mírou chybovosti, spotřebou a výkonem. Použití evolučního návrhu vedlo v této oblasti k prvním slibným výsledkům, ale naráží na problém se škálovatelností při vyhodnocování kandidátních řešení. K řešení tohoto problému, který označujeme jako ověřování přibližné shody, navrhujeme nový přístup: využití pokročilých technik formální verifikace zvlášť upravených k rychlému výpočtu vzdálenosti mezi kandidátními a referenčními řešeními. Projekt směřuje k následujícím přínosům: (1) návrh efektivních algoritmů pro ověřování přibližné shody kombinačních (bezestavových) i sekvenčních (stavových) systémů, (2) návrh aproximačních algoritmů založených na genetickém programování a na vyvinutých algoritmech ověřování přibližné shody a (3) experimentální vyhodnocení navržených metod.

Description in English
Approximate computing is a promising approach to obtain energy-efficient computer systems. It exploits the fact that many applications are error resilient, i.e., do not require a perfect output to be produced. An open problem is how to effectively obtain approximations that are good compromises between the error ratio, power consumption, and performance. Using evolutionary algorithms for the approximation has led to promising results, but it suffers from scalability problems in evaluating candidate solutions. For that, we propose a novel way: using advanced methods of formal verification redesigned to quickly calculate distances between candidate approximations and the reference implementation, which we call relaxed equivalence checking. The project seeks the following original contributions: (1) efficient algorithms for relaxed equivalence checking of combinational (stateless) and sequential (stateful) systems, (2) approximation algorithms based on genetic programming using the proposed relaxed equivalence checking, (3) experimental evaluation of the proposed approximation methods.

Keywords
aproximativní počítání; genetické programování; vyvíjející se hardware; ověřování přibližné shody; automaty; logika

Key words in English
approximate computing; genetic programming; evolvable hardware; relaxed equivalence checking; automata; logic

Mark

GA16-17538S

Default language

Czech

People responsible

Holík Lukáš, doc. Mgr., Ph.D. - fellow researcher
Lengál Ondřej, Ing., Ph.D. - fellow researcher
Rogalewicz Adam, doc. Mgr., Ph.D. - fellow researcher
Sekanina Lukáš, prof. Ing., Ph.D. - fellow researcher
Vašíček Zdeněk, doc. Ing., Ph.D. - fellow researcher
Vojnar Tomáš, prof. Ing., Ph.D. - principal person responsible

Units

Department of Intelligent Systems
- (2015-03-24 - 2018-12-31)
Department of Computer Systems
- (2015-03-24 - 2018-12-31)

Results

DVOŘÁČEK, P.; SEKANINA, L. Evolutionary Approximation of Edge Detection Circuits. In 19th European Conference on Genetic programming. Lecture Notes in Computer Science. Berlin: Springer International Publishing, 2016. p. 19-34. ISBN: 978-3-319-30667-4.
Detail

FIEDOR, T.; HOLÍK, L.; JANKŮ, P.; LENGÁL, O.; VOJNAR, T. Lazy Automata Techniques for WS1S. arXiv:1701.06282: 2017. p. 0-0.
Detail

CHEN, Y.; HSIEH, C.; LENGÁL, O.; LII, T.; TSAI, M.; WANG, B.; WANG, F. PAC Learning-Based Verification and Model Synthesis. In Proceedings of the 38th International Conference on Software Engineering. Austin, TX: Association for Computing Machinery, 2016. p. 714-724. ISBN: 978-1-4503-3900-1.
Detail

SEKANINA, L.; KAPUSTA, V. Visualisation and Analysis of Genetic Records Produced by Cartesian Genetic Programming. In GECCO'16 Companion. New York: Association for Computing Machinery, 2016. p. 1411-1418. ISBN: 978-1-4503-4323-7.
Detail

ABATE, A.; ČEŠKA, M.; KWIATKOWSKA, M. Approximate Policy Iteration for Markov Decision Processes via Quantitative Adaptive Aggregations. In Proceedings of 14th International Symposium on Automated Technology for Verification and Analysis. Lecture Notes in Computer Science. Heidelberg: Springer Verlag, 2016. p. 13-31. ISBN: 978-3-319-46519-7.
Detail

VAVERKA, F.; HRBÁČEK, R.; SEKANINA, L. Evolving Component Library for Approximate High Level Synthesis. In 2016 IEEE Symposium Series on Computational Intelligence. Athens: IEEE Computational Intelligence Society, 2016. p. 1-8. ISBN: 978-1-5090-4240-1.
Detail

VAŠÍČEK, Z.; MRÁZEK, V.; SEKANINA, L. Evolutionary Functional Approximation of Circuits Implemented into FPGAs. In 2016 IEEE Symposium Series on Computational Intelligence. Athens: Institute of Electrical and Electronics Engineers, 2016. p. 1-8. ISBN: 978-1-5090-4240-1.
Detail

HRBÁČEK, R.; MRÁZEK, V.; VAŠÍČEK, Z. Automatic Design of Approximate Circuits by Means of Multi-Objective Evolutionary Algorithms. In Proceedings of the 11th International Conference on Design & Technology of Integrated Systems in Nanoscale Era. Istanbul: Istanbul Sehir University, 2016. p. 239-244. ISBN: 978-1-5090-0335-8.
Detail

MRÁZEK, V.; VAŠÍČEK, Z. Automatic Design of Arbitrary-Size Approximate Sorting Networks with Error Guarantee. In Power and Timing Modeling, Optimization and Simulation (PATMOS), 2016 26rd International Workshop on. Bremen: Institute of Electrical and Electronics Engineers, 2016. p. 221-228. ISBN: 978-1-5090-0733-2.
Detail

MRÁZEK, V.; HRBÁČEK, R.; VAŠÍČEK, Z.; SEKANINA, L. EvoApprox8b: Library of Approximate Adders and Multipliers for Circuit Design and Benchmarking of Approximation Methods. In Proc. of the 2017 Design, Automation & Test in Europe Conference & Exhibition (DATE). Lausanne: European Design and Automation Association, 2017. p. 258-261. ISBN: 978-3-9815370-9-3.
Detail

VAŠÍČEK, Z.; MRÁZEK, V.; SEKANINA, L. Towards Low Power Approximate DCT Architecture for HEVC Standard. In Proc. of the 2017 Design, Automation & Test in Europe Conference & Exhibition (DATE). Lausanne: European Design and Automation Association, 2017. p. 1576-1581. ISBN: 978-3-9815370-9-3.
Detail

FIEDOR, T.; HOLÍK, L.; JANKŮ, P.; LENGÁL, O.; VOJNAR, T. Lazy Automata Techniques for WS1S. In Proceedings of TACAS'17. Lecture Notes in Computer Science. Lecture Notes in Computer Science. Heidelberg: Springer Verlag, 2017. p. 407-425. ISBN: 978-3-662-54576-8. ISSN: 0302-9743.
Detail

MINAŘÍK, M.; SEKANINA, L. On Evolutionary Approximation of Sigmoid Function for HW/SW Embedded Systems. In 20th European Conference on Genetic Programming, EuroGP 2017. Lecture Notes in Computer Science. Berlin: Springer International Publishing, 2017. p. 343-358. ISBN: 978-3-319-55696-3.
Detail

ČEŠKA, M.; CALINESCU, R.; GERASIMOU, S.; KWIATKOWSKA, M.; PAOLETTI, N. Designing Robust Software Systems through Parametric Markov Chain Synthesis. In Proceedings of 14th IEEE International Conference On Software Architecture. New Jersey: IEEE Computer Society, 2017. p. 131-140. ISBN: 978-1-5090-5729-0.
Detail

VAŠÍČEK, Z. Relaxed equivalence checking: a new challenge in logic synthesis. In Proceedings 2017 IEEE 20th International Symposium on Design and Diagnotics of Electronic Circuit & Systems. Dresden: IEEE Computer Society, 2017. p. 1-6. ISBN: 978-1-5386-0472-4.
Detail

ČEŠKA, M.; MATYÁŠ, J.; MRÁZEK, V.; VAŠÍČEK, Z.; SEKANINA, L.; VOJNAR, T. Approximating Complex Arithmetic Circuits with Formal Error Guarantees: 32-bit Multipliers Accomplished. In Proceedings of 36th IEEE/ACM International Conference On Computer Aided Design (ICCAD). Irvine, CA: Institute of Electrical and Electronics Engineers, 2017. p. 416-423. ISBN: 978-1-5386-3093-8.
Detail

SEKANINA, L.; VAŠÍČEK, Z.; MRÁZEK, V. Approximate Circuits in Low-Power Image and Video Processing: The Approximate Median Filter. Radioengineering, 2017, vol. 26, no. 3, p. 623-632. ISSN: 1210-2512.
Detail

LENGÁL, O.; VOJNAR, T.; ENEA, C.; SIGHIREANU, M. Compositional Entailment Checking for a Fragment of Separation Logic. FORMAL METHODS IN SYSTEM DESIGN, 2017, vol. 2017, no. 51, p. 575-607. ISSN: 0925-9856.
Detail

ČEŠKA, M.; CARDELLI, L.; FRANZLE, M.; KWIATKOWSKA, M.; LAURENTI, L.; PAOLETTI, N.; WHITBY, M. Syntax-Guided Optimal Synthesis for Chemical Reaction Networks. In Proceedings of the 29th International Conference on Computer Aided Verification. Lecture Notes in Computer Science. Heidelberg: Springer Verlag, 2017. p. 375-395. ISBN: 978-3-319-63390-9.
Detail

ČEŠKA, M.; CALINESCU, R.; GERASIMOU, S.; KWIATKOWSKA, M.; PAOLETTI, N. RODES: A Robust-Design Synthesis Tool for Probabilistic Systems. In Proceedings of 14th International Conference on Quantitative Evaluation of SysTems. Heidelberg: Springer Verlag, 2017. p. 304-308. ISBN: 978-3-319-66335-7.
Detail

ČEŠKA, M.; HAVLENA, V.; HOLÍK, L.; LENGÁL, O.; VOJNAR, T. Approximate Reduction of Finite Automata for High-Speed Network Intrusion Detection. In Proceedings of TACAS'18. Lecture Notes in Computer Science. Thessaloniki: Springer Verlag, 2018. p. 155-175. ISSN: 0302-9743.
Detail

MRÁZEK, V.; VAŠÍČEK, Z.; HRBÁČEK, R. Role of circuit representation in evolutionary design of energy-efficient approximate circuits. IET Computers and Digital Techniques, 2018, vol. 2018, no. 4, p. 139-149. ISSN: 1751-8601.
Detail

HUSA, J.; KALKREUTH, R. A Comparative Study on Crossover in Cartesian Genetic Programming. In Genetic Programming 21st European Conference, EuroGP 2018, Proceedings. Lecture Notes in Computer Science. Cham: Springer International Publishing, 2018. p. 203-219. ISBN: 978-3-319-77553-1. ISSN: 0302-9743.
Detail

MRÁZEK, V.; VAŠÍČEK, Z. Evolutionary Design of Large Approximate Adders Optimized for Various Error Criteria. In Proceedings of the Genetic and Evolutionary Computation Conference Companion (GECCO '18). Kyoto: Association for Computing Machinery, 2018. p. 294-295. ISBN: 978-1-4503-5764-7.
Detail

MRÁZEK, V.; VAŠÍČEK, Z.; SEKANINA, L.; JIANG, H.; HAN, J. Scalable Construction of Approximate Multipliers With Formally Guaranteed Worst Case Error. IEEE Trans. on VLSI Systems., 2018, vol. 26, no. 11, p. 2572-2576. ISSN: 1063-8210.
Detail

SEKANINA, L.; VAŠÍČEK, Z.; MRÁZEK, V. Automated Search-Based Functional Approximation for Digital Circuits. In Approximate Circuits - Methodologies and CAD. Heidelberg: Springer International Publishing, 2019. p. 175-203. ISBN: 978-3-319-99322-5.
Detail

WIGLASZ, M.; SEKANINA, L. Cooperative Coevolutionary Approximation in HOG-based Human Detection Embedded System. In 2018 IEEE Symposium Series on Computational Intelligence (SSCI 2018). Bengaluru: Institute of Electrical and Electronics Engineers, 2018. p. 1313-1320. ISBN: 978-1-5386-9276-9.
Detail

CALINESCU, R.; ČEŠKA, M.; GERASIMOU, S.; KWIATKOWSKA, M.; PAOLETTI, N. Efficient Synthesis of Robust Models for Stochastic Systems. JOURNAL OF SYSTEMS AND SOFTWARE, 2018, vol. 2018, no. 143, p. 140-158. ISSN: 0164-1212.
Detail

HOLÍK, L.; LENGÁL, O.; SÍČ, J.; VOJNAR, T.; VEANES, M. Simulation Algorithms for Symbolic Automata. In Proc. of 16th International Symposium on Automated Technology for Verification and Analysis. Lecture Notes in Computer Science. Heidelberg: Springer Verlag, 2018. p. 109-125. ISBN: 978-3-030-01089-8. ISSN: 0302-9743.
Detail

HOLÍK, L.; LENGÁL, O.; SÍČ, J.; VOJNAR, T.; VEANES, M. Simulation Algorithms for Symbolic Automata (Technical Report). Ithaca: 2018. p. 1-23.
Detail

HOLÍK, L.; LENGÁL, O.; ROGALEWICZ, A.; SEKANINA, L.; VAŠÍČEK, Z.; VOJNAR, T. Towards Formal Relaxed Equivalence Checking in Approximate Computing Methodology. 2nd Workshop on Approximate Computing (WAPCO 2016). Prague: 2016. p. 1-6.
Detail

ČEŠKA, M.; CALINESCU, R.; GERASIMOU, S.; KWIATKOWSKA, M.; PAOLETTI, N. Recent Advances in Designing Robust Probabilistic Systems. 2nd International Workshop on Design and Analysis of Robust Systems (Extended Abstract). Berlin: 2017. p. 1-3.
Detail

FIEDOR, T.; HOLÍK, L.; JANKŮ, P.; LENGÁL, O.; VOJNAR, T.: gaston; Gaston - Symbolic WS1S Solver. Nástroj a dodatečné informace se nacházejí na http://www.fit.vutbr.cz/research/groups/verifit/tools/gaston/ a https://github.com/tfiedor/gaston. URL: https://www.fit.vut.cz/research/product/511/. (software)
Detail

Responsibility: Vojnar Tomáš, prof. Ing., Ph.D.