A performance study of multiprocessor task scheduling algorithms

Jin, Shidai; Schiavone, Guy A.; Turgut, Damla

doi:10.1007/s11227-007-0139-z

Cited by 74 publications

(21 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They can be classified based on the computing platforms as mentioned above, or based on the characteristics of the real time applications. Another classification to scheduling approaches could be based on additional performance metrics along with satisfying timing requirements such as minimizing number of processors as in [12], reducing power consumption in processors with dynamic voltage scaling as in [1], [18], [17], [23] or achieving effective fault-tolerant in real time systems as in [22], [24].…”

Section: Related Workmentioning

confidence: 99%

Real-Time Workload Allocation on a Uni-Processor

Hamdy¹,

Youssef²,

Ammar³

2012

IJCA

View full text Add to dashboard Cite

The paper presents a novel allocation algorithm to allocate independent real time tasks on a processor in a way that improves the processor's throughput (Processor's throughput is the number of tasks the processor can accept for execution). The proposed approach allocates tasks' workloads (task's workload is the percentage of work required by the processor to execute the task) instead of their processing powers (Processing power assigned to a task is a percentage of the processor reserved to execute the task such that its deadline is satisfied). To achieve our objective a variable processing power is assigned to the task under consideration over its deadline to satisfy its timing requirements instead of rejecting it if a constant processing power cannot be guaranteed as in previous CPU reservation approaches. Simulation results revealed that the acceptance rate of the admitted tasks to a certain processor using the new approach is superior to that achieved using the traditional processing power reservation approach.

show abstract

Section: Related Workmentioning

confidence: 99%

Real-Time Workload Allocation on a Uni-Processor

Hamdy¹,

Youssef²,

Ammar³

2012

IJCA

View full text Add to dashboard Cite

show abstract

“…Various effective algorithms exist for scheduling applications with dependent tasks on a set of machines [29,43,44]. The input for these algorithms is a directed acyclic graph (DAG) that represents the dependencies between the tasks of an application.…”

Section: Type Of Parallel Applicationsmentioning

confidence: 99%

A strategy for scheduling tightly coupled parallel applications on clusters

Sanjay

Vadhiyar

2009

Concurrency and Computation

View full text Add to dashboard Cite

SUMMARYAlthough various strategies have been developed for scheduling parallel applications with independent tasks, very little work exists for scheduling tightly coupled parallel applications on cluster environments. In this paper, we compare four different strategies based on performance models of tightly coupled parallel applications for scheduling the applications on clusters. In addition to algorithms based on existing popular optimization techniques, we also propose a new algorithm called Box Elimination that searches the space of performance model parameters to determine the best schedule of machines. By means of real and simulation experiments, we evaluated the algorithms on single cluster and multi-cluster setups. We show that our Box Elimination algorithm generates up to 80% more efficient schedules than other algorithms. We also show that the execution times of the schedules produced by our algorithm are more robust against the performance modeling errors.

show abstract

“…Control flow graph and data dependence analyses are used to make this parallelization. Many task scheduling tools exist [17] [18], but they are totally black boxes and produce object codes directly. There is no way to check that the executed code is correct as it could be possible from the sources of parallel codes.…”

Section: Introductionmentioning

confidence: 99%

Automatic Code Generation of Distributed Parallel Tasks

Lossing

Ancourt

Irigoin

2016

2016 IEEE Intl Conference on Computational Science and Engineering (CSE) and IEEE Intl Conference on Embedded and Ubiquitous Co

View full text Add to dashboard Cite

Abstract-With the advent of clustered systems, more and more parallel computing is required. However a lot of programming skills is needed to write a parallel codes, especially when you want to benefit from the various parallel architectural resources, with heterogeneous units and complex memory organizations. We present in this paper a method that generates automatically, step by step, a task-parallel distributed code from a sequential program. It has been implemented in an existing source-to-source compiler PIPS. Our approach provides two main advantages 1) all the program transformations are simple and applied on source code, thus are visible by the user, 2) a proof of correctness of the parallelization process can be made. This ensures that we end up with a correct distributed-task program for distributed-memory machines. To our knowledge, it is the first tool that automatically generates a distributed code for task parallelization.

show abstract

A performance study of multiprocessor task scheduling algorithms

Cited by 74 publications

References 49 publications

Real-Time Workload Allocation on a Uni-Processor

Real-Time Workload Allocation on a Uni-Processor

A strategy for scheduling tightly coupled parallel applications on clusters

Automatic Code Generation of Distributed Parallel Tasks

Contact Info

Product

Resources

About