Application of ASP for Automatic Synthesis of Flexible Multiprocessor Systems from Parallel Programs

Authors:
Harold Ishebabi;Philipp Mahr;Christophe Bobda;Martin Gebser;Torsten Schaub
Affiliations:
Institute for Informatics, University of Potsdam, Potsdam D-14482;Institute for Informatics, University of Potsdam, Potsdam D-14482;Institute for Informatics, University of Potsdam, Potsdam D-14482;Institute for Informatics, University of Potsdam, Potsdam D-14482;Institute for Informatics, University of Potsdam, Potsdam D-14482
Venue:
LPNMR '09 Proceedings of the 10th International Conference on Logic Programming and Nonmonotonic Reasoning
Year:
2009

Citing 5
Cited 1

Extending and implementing the stable model semantics

Artificial Intelligence
Design of adaptive multiprocessor on chip systems

Proceedings of the 20th annual conference on Integrated circuits and systems design
Automatic Synthesis of Multiprocessor Systems from Parallel Programs under Preemptive Scheduling

RECONFIG '08 Proceedings of the 2008 International Conference on Reconfigurable Computing and FPGAs
Automated architecture synthesis for parallel programs on FPGA multiprocessor systems

Microprocessors & Microsystems
Conflict-driven answer set solving

IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence

ASP-based optimized mapping in a simulink-to-MPSoC design flow

Journal of Systems Architecture: the EUROMICRO Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

Configurable on chip multiprocessor systems combine advantages of task-level parallelism and the flexibility of field-programmable devices to customize architectures for parallel programs, thereby alleviating technological limitations due to memory bandwidth and power consumption. Given the huge size of the design space of such systems, it is important to automatically optimize design parameters in order to facilitate wide and disciplined explorations. Being a combinatorial problem, system design can be modeled and solved as such, but the amount of parameters renders the problem difficult to solve for large instances. However, as the synthesis problem usually exhibits structure, Answer Set Programming (ASP), for which solvers utilizing techniques from the propositional satisfiability domain are available, can be effectively employed. This paper presents a design flow based on ASP that uses the solver clasp as back-end engine. Synthesis experiments demonstrate the effectiveness of the approach.