Adaptive statistical scheduling of divisible workloads in heterogeneous systems

González–Vélez, Horacio; Cole, Murray

doi:10.1007/s10951-009-0138-4

Cited by 15 publications

(9 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nevertheless, it is important to assert that there is a significant number of real problems in computational science, which can be modelled as divisible workloads of independent tasks, with or without precedence relations [17,18].…”

Section: Discussionmentioning

confidence: 99%

Adaptive structured parallelism for distributed heterogeneous architectures: a methodological approach with pipelines and farms

González–Vélez

Cole

2010

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

SUMMARYAlgorithmic skeletons abstract commonly used patterns of parallel computation, communication, and interaction. Based on the algorithmic skeleton concept, structured parallelism provides a high-level parallel programming technique that allows the conceptual description of parallel programs while fostering platform independence and algorithm abstraction. This work presents a methodology to improve skeletal parallel programming in heterogeneous distributed systems by introducing adaptivity through resource awareness. As we hypothesise that a skeletal program should be able to adapt to the dynamic resource conditions over time using its structural forecasting information, we have developed adaptive structured parallelism (ASPARA). ASPARA is a generic methodology to incorporate structural information at compilation into a parallel program, which will help it to adapt at execution. ASPARA comprises four phases: programming, compilation, calibration, and execution. We illustrate the feasibility of this approach and its associated performance improvements using independent case studies based on two algorithmic skeletons-the task farm and the pipeline-evaluated in a non-dedicated heterogeneous multi-cluster system.

show abstract

Section: Discussionmentioning

confidence: 99%

Adaptive structured parallelism for distributed heterogeneous architectures: a methodological approach with pipelines and farms

González–Vélez

Cole

2010

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

show abstract

“…Divisible Workloads. This approach builds on previous work from several partners including that of Robert Gordon University on statistical scheduling of divisible workloads [30] and the systematic introduction of adaptivity into parallel patterns and skeletons [29,31,4].…”

Section: Related Workmentioning

confidence: 99%

The ParaPhrase Project: Parallel Patterns for Adaptive Heterogeneous Multicore Systems

Hammond

Aldinucci²,

Brown

et al. 2013

Formal Methods for Components and Objects

Self Cite

View full text Add to dashboard Cite

Abstract. This paper describes the ParaPhrase project, a new 3-year targeted research project funded under EU Framework 7 Objective 3.4 (Computer Systems), starting in October 2011. ParaPhrase aims to follow a new approach to introducing parallelism using advanced refactoring techniques coupled with high-level parallel design patterns. The refactoring approach will use these design patterns to restructure programs defined as networks of software components into other forms that are more suited to parallel execution. The programmer will be aided by high-level cost information that will be integrated into the refactoring tools. The implementation of these patterns will then use a wellunderstood algorithmic skeleton approach to achieve good parallelism.A key ParaPhrase design goal is that parallel components are intended to match heterogeneous architectures, defined in terms of CPU/GPU combinations, for example. In order to achieve this, the ParaPhrase approach will map components at link time to the available hardware, and will then re-map them during program execution, taking account of multiple applications, changes in hardware resource availability, the desire to reduce communication costs etc. In this way, we aim to develop a new approach to programming that will be able to produce software that can adapt to dynamic changes in the system environment. Moreover, by using a strong component basis for parallelism, we can achieve potentially significant gains in terms of reducing sharing at a high level of abstraction, and so in reducing or even eliminating the costs that are usually associated with cache management, locking, and synchronisation.

show abstract

“…More recent research from the Edinburgh group has addressed the problem of adaptation on structured parallel programming 85, 86, in particular for the pipe skeleton 87, 88 and the farm 89.Known as adaptive structured parallelism, this methodological approach puts particular emphasis on the automatic scheduling of algorithmic skeletons 90.It instruments a skeletal program with a series of rules, which depend on particular performance thresholds based on the nature of the skeleton, the computation/communication ratio of the program, and the availability of resources in the system.Employing the pipe and farm skeletons as a basis for experimentation, this methodology has been successfully applied to allocate divisible workloads to real multinode configurations 91 and to solve complex parameter sweeps in the biomedical sciences92.…”

Section: Individualized Description Of Featuresmentioning

confidence: 99%

A survey of algorithmic skeleton frameworks: high‐level structured parallel programming enablers

2010

Self Cite

View full text Add to dashboard Cite

Structured parallel programs ought to be conceived as two separate and complementary entities: computation, which expresses the calculations in a procedural manner, and coordination, which abstracts the interaction and communication. By abstracting commonly used patterns of parallel computation, communication, and interaction, algorithmic skeletons enable programmers to code algorithms without specifying platform‐dependent primitives. This article presents a literature review on algorithmic skeleton frameworks (ASKF), parallel software development environments furnishing a collection of parameterizable algorithmic skeletons, where the control flow, nesting, resource monitoring, and portability of the resulting parallel program is dictated by the ASKF as opposed to the programmer. Consequently, the ASKF can be positioned as high‐level structured parallel programming enablers, as their systematic utilization permits the abstract description of programs and fosters portability by focusing on the description of the algorithmic structure rather than on its detailed implementation. Copyright © 2010 John Wiley & Sons, Ltd.

show abstract

Adaptive statistical scheduling of divisible workloads in heterogeneous systems

Cited by 15 publications

References 28 publications

Adaptive structured parallelism for distributed heterogeneous architectures: a methodological approach with pipelines and farms

Adaptive structured parallelism for distributed heterogeneous architectures: a methodological approach with pipelines and farms

The ParaPhrase Project: Parallel Patterns for Adaptive Heterogeneous Multicore Systems

A survey of algorithmic skeleton frameworks: high‐level structured parallel programming enablers

Contact Info

Product

Resources

About