Multi-stage Benders Decomposition for Optimizing Multicore Architectures

Benini, Luca; Lombardi, Michele; Mantovani, M.; Milano, Michela; Ruggiero, Martino

doi:10.1007/978-3-540-68155-7_6

Cited by 16 publications

(10 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In practice the solver moves towards promising part of the search space by learning from its mistakes, rather than taking very good decisions in the earlier stages. Experimental results (reported in Benini et al (2008a)) shown how in our case this choice pays off in terms of computation time, compared to using higher quality (but harder to optimize) heuristic objective functions, or less expensive (but weaker) cuts.…”

Section: Memory Device Allocationmentioning

confidence: 91%

“…Following this idea, in Benini et al (2008a), we have implemented a Logic Based Benders' Decomposition based approach, similarly to Benini et al (2005Benini et al ( , 2006) (see Fig. 4A).…”

Section: Multi-stage Logic Based Benders Decomposition (Ms-lbd)mentioning

confidence: 99%

“…The allocation part is extremely difficult to solve while the scheduling part is indeed easier. We propose in this paper three alternative approaches: the first is a three-stage decomposition, which is actually a recursive application of standard Logic based Benders' Decomposition (LBD), that aims at obtaining balanced and lighter components (Benini et al 2008a); the second is a pure Constraint Programming approach (Benini et al 2008b) and the third is a novel contribution derived by integrating the two solvers.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Optimal resource allocation and scheduling for the CELL BE platform

et al. 2010

Self Cite

View full text Add to dashboard Cite

Resource allocation and scheduling for multicore platforms is one of the most critical challenges in today's embedded computing. In this paper we focus on a well-known multicore platform, namely the Cell BE processor, and we address the problem of allocating and scheduling its processors, communication channels and memories, with the goal of minimizing execution time for complex data streaming applications.We propose three complete approaches that optimally solve the problem and prove optimality. The first is based on the recursive application of the Logic Based Benders decomposition, resulting in a three stage algorithm. The second is a pure CP approach while the third is a hybrid approach integrating the first two.Extensive experimental evaluation shows the features of each approach and its effectiveness on a specific instance structure.

show abstract

Section: Memory Device Allocationmentioning

confidence: 91%

“…Following this idea, in Benini et al (2008a), we have implemented a Logic Based Benders' Decomposition based approach, similarly to Benini et al (2005Benini et al ( , 2006) (see Fig. 4A).…”

Section: Multi-stage Logic Based Benders Decomposition (Ms-lbd)mentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Optimal resource allocation and scheduling for the CELL BE platform

et al. 2010

Self Cite

View full text Add to dashboard Cite

show abstract

“…The second is an exact optimization algorithm providing optimal solutions. It has been extensively described in [3] and it is based on a multi-stage decomposition. An extensive set of experimental results confirm that the multi-stage decomposition pays off in terms of efficiency w.r.t.…”

Section: Introductionmentioning

confidence: 99%

Cellflow: A Parallel Application Development Environment with Run-Time Support for the Cell BE Processor

Ruggiero

Lombardi

Milano

et al. 2008

2008 11th EUROMICRO Conference on Digital System Design Architectures, Methods and Tools

Self Cite

View full text Add to dashboard Cite

The Cell BE processor provides both scalable computation power and flexibility, and it is already being adopted for many computational intensive applications. Despite of its merits, it also presents many challenges, as it is now widely known that is very difficult to program the Cell BE in an efficient manner. Hence, the creation of an efficient software development framework is becoming the key challenge for this computational platform.We propose a novel software toolkit, called Cellflow, which enables developers to quickly build multi-task applications for Cellbased platform. We support programmers from the initial stage of their work, through a development-time software infrastructure, to the final stage of the application development, proposing a safe and easy-to-use explicit parallel programming model.Experimental results show that in Cellflow we reduced to minimum the abstraction gap between the optimization and development phases.

show abstract

“…The whole problem is decomposed into a master problem and a subproblem, that will then communicate. One of the two could be more easily solvable by an FD solver, while the other by a linear solver; this gives an interesting pattern to have the two solvers communicate [70,17,98].…”

Section: Integration With Operations Researchmentioning

confidence: 99%

Constraint Logic Programming

Rossi

2003

Constraint Processing

View full text Add to dashboard Cite

Abstract. Constraint Logic Programming (CLP) is one of the most successful branches of Logic Programming; it attracts the interest of theoreticians and practitioners, and it is currently used in many commercial applications. Since the original proposal, it has developed enormously: many languages and systems are now available either as open source programs or as commercial systems.Also, CLP has been one of the technologies able to recruit researchers from other communities to the declarative programming cause. Current CLP engines include technologies and results developed in other communities, which themselves discovered logic as an invaluable tool to model and solve real-life problems. The CLP ParadigmConstraint Logic Programming (CLP) [7] represents a successful attempt to merge the best features of logic programming (LP) and constraint solving.Constraint solving [127,6,56,31] includes a variety of expressive modelling frameworks and efficient solving tools for real-life problems that can be described via a set of variables and constraints over them. A constraint is just a restriction imposed over the combination of values of some variables of the problem. Solving a problem with constraints means finding a way to assign values to all its variables such that all constraints are satisfied. Constraint solving methods have been successfully applied to many application domains, such as scheduling, planning, resource allocation, vehicle routing, computer networks, and bioinformatics [137,127,51].Embedding the notion of constraint into a high-level programming language allows for a more flexible and practical constraint processing environment, where constraints can be represented as formulae and can be incrementally accumulated. Moreover, the presence of constraints in a programming language usually augments its expressive power, in the sense that some complex relations can be defined easily by means of constraints, and there are also efficient techniques to prove them.For these reasons, constraints have been embedded in many programming environments, but some are more suitable than others. For example, the fact that constraints can be seen as relations or predicates, that constraint solving can be seen as a generalized form of unification, that their conjunction can be seen as a logical and, and that backtracking search is the base methodology to A. Dovier, E. Pontelli (Eds.): 25

show abstract

Multi-stage Benders Decomposition for Optimizing Multicore Architectures

Cited by 16 publications

References 13 publications

Optimal resource allocation and scheduling for the CELL BE platform

Optimal resource allocation and scheduling for the CELL BE platform

Cellflow: A Parallel Application Development Environment with Run-Time Support for the Cell BE Processor

Constraint Logic Programming

Contact Info

Product

Resources

About