Scheduling independent jobs on partitionable hypercubes

Chen, Guaning; Lai, Ten-Hwang

doi:10.1016/0743-7315(91)90031-4

Cited by 19 publications

(8 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is called largest dimension largest processing time (LDLPT) algorithm [3], which can slightly improve the worst-case performance. However, this improvement does not affect our analytical results.…”

Section: Independent Random Tasksmentioning

confidence: 99%

“…Our problem addressed in this paper is to find nonpreemptive schedules for nonmalleable parallel tasks. A special case of our problem is scheduling parallel computations with independent parallel tasks [2,16,21], especially in binarily partitionable systems such hypercubes [3,30]. For precedence constrained parallel tasks, an earlier work [22] showed that if all the tasks have the same execution time, a worst-case performance ratio of ð2P 2 DÞ=ðP 2 D þ 1Þ can be achieved, where P is the total number of processors available and D is the maximum task parallelism.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Scheduling DAGs with random parallel tasks on binarily partitionable systems

Li¹

2005

International Journal of Parallel, Emergent and Distributed Sys

View full text Add to dashboard Cite

We investigate the problem of scheduling dags with parallel tasks on binarily partitionable systems by analyzing a n ðLLÞ; the average-case performance ratio of the level-by-level (LL) scheduling algorithm, where n is the number of tasks. It is shown that for arbitrary probability distributions of task parallelisms and task execution times, the asymptotic average-case performance ratio of algorithm LL in scheduling parallel computations with wide dags is a 1 ðLLÞ ¼ 1: In particular, for exponential distributions of task execution times, the average-case performance ratio of LL in scheduling iterative computations, complete trees, partitioning algorithms and diamond dags is a n ðLLÞ ¼ 1 þ Oðlog n=nÞ; 1 þ Oððlog nÞ 2 =nÞ; 1 þ Oððlog nÞ 2 =nÞ and 1 þ Oððlog nÞ 2 = ffiffi ffi n p Þ; respectively.

show abstract

Section: Independent Random Tasksmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Scheduling DAGs with random parallel tasks on binarily partitionable systems

Li¹

2005

International Journal of Parallel, Emergent and Distributed Sys

View full text Add to dashboard Cite

show abstract

“…Some features of distributed memory hypercube computers were described in [13]. Other network topologies can be Hypercubes, Lines, Meshes, Rings, Trees [2,5,15]. A parallel system is a PRAM if its underlying networking topology is a complete graph.…”

Section: Introductionmentioning

confidence: 99%

Maximizing the throughput of parallel jobs on hypercubes

Zhang

2007

Information Processing Letters

View full text Add to dashboard Cite

“…Static algorithms are appropriate to a very specialized class of systems in which this information is available, including some real-time systems. Numerous static algorithms have been proposed, including [2,3,21]. Dynamic scheduling, on the other hand, assumes a continuous stochastic stream of arriving jobs.…”

Section: Introductionmentioning

confidence: 99%