Task Tree Partition and Subtree Allocation for Heterogeneous Multiprocessors

He, Suna; Wu, Jigang; Wei, Bing; Wu, Jiaxin

doi:10.1109/ispa-bdcloud-socialcom-sustaincom52081.2021.00084

Cited by 3 publications

(4 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Note that the problem of makespan minimization of a tree of tasks, by partitioning the tree so that each part fits (memory-wise) onto a processor, has already been tackled in the case of homogeneous processors [10]. As pointed out in Section 1, recent work by He et al has attempted to extend this approach to heterogeneous architectures [11]. Their work leaves several important questions open, though: (i) the experiments seem to be on a system with homogeneous memories only, and (ii) the descriptions regarding the subroutine FitMemory are not sufficient for a reimplementation.…”

Section: Related Workmentioning

confidence: 99%

“…1 Thus, to adapt the scheduling algorithm to variable memory constraints is very relevant. Yet, maybe with the exception of He et al [11], there are no scheduling algorithms in the literature tailored to the problem of scheduling tree-shaped workflows on memory-heterogeneous architectures. And while He et al [11] design their algorithm with heterogeneity in mind, their experimental setup and results do not consider memory-heterogeneous architectures, which are our focus.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mapping Tree-Shaped Workflows on Memory-Heterogeneous Architectures

Kulagina

Meyerhenke

Benoît

2023

Euro-Par 2022: Parallel Processing Workshops

View full text Add to dashboard Cite

Directed acyclic graphs are commonly used to model scientific workflows, by expressing dependencies between tasks, as well as the resource requirements of the workflow. As a special case, rooted directed trees occur in several applications, for instance in sparse matrix computations. Since typical workflows are modeled by huge trees, it is crucial to schedule them efficiently, so that their execution time (or makespan) is minimized. Furthermore, it might be beneficial to distribute the execution on several compute nodes, hence increasing the available memory, and allowing us to parallelize parts of the execution. To exploit the heterogeneity of modern clusters in this context, we investigate the partitioning and mapping of tree-shaped workflows on target architectures where each processor can have a different memory size. Our three-step heuristic adapts and extends previous work for homogeneous clusters [Gou et al., TPDS 2020]. The changes we propose concern the assignment to processors (which considers the different memory sizes) and the availability of suitable processors when splitting or merging subtrees. We evaluate our approach with extensive simulations and demonstrate that exploiting the heterogeneity in the cluster reduces the makespan significantly compared to the state of the art for homogeneous memory.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Mapping Tree-Shaped Workflows on Memory-Heterogeneous Architectures

Kulagina

Meyerhenke

Benoît

2023

Euro-Par 2022: Parallel Processing Workshops

View full text Add to dashboard Cite

show abstract

“…Note that the problem of makespan minimization of a tree of tasks, by partitioning the tree so that each part fits (memory‐wise) onto a processor, has already been tackled in the case of homogeneous processors by Gou et al 3 As pointed out in Section 1, recent work by He et al has attempted to extend this approach to heterogeneous architectures 4 . Their work leaves several important questions open, though: (i) the experiments seem to be on a system with homogeneous memories only, and (ii) the descriptions regarding the subroutine FitMemory are not sufficient for a reimplementation.…”

Section: Related Workmentioning

confidence: 99%

“…Thus, adapting the mapping algorithm to different memory sizes and processor speeds is very relevant. Yet, maybe with the exception of He et al, 4 there are no algorithms in the literature tailored to the problem of mapping tree‐shaped workflows on memory‐ or speed‐heterogeneous architectures (i.e., where the memory sizes/processor speeds differ). And, while He et al 4 design their algorithm with heterogeneity in mind, their experimental setup and results do not consider architectures with different memory sizes or different processor speeds, which are in our focus.…”

Section: Introductionmentioning

confidence: 99%

Mapping tree‐shaped workflows on systems with different memory sizes and processor speeds

Kulagina

Meyerhenke

Benoît

2023

Concurrency and Computation

View full text Add to dashboard Cite

Directed acyclic graphs are commonly used to model scientific workflows, by expressing dependencies between tasks, as well as the resource requirements of the workflow. As a special case, rooted directed trees occur in several applications, for instance in sparse matrix computations. Since typical workflows are modeled by large trees, it is crucial to schedule them efficiently, so that their execution time (or makespan) is minimized. Furthermore, it is usually beneficial to distribute the execution on several compute nodes, hence increasing the available memory, and allowing us to parallelize parts of the execution. To exploit the heterogeneity of modern clusters in this context, we investigate the partitioning and mapping of tree‐shaped workflows on two types of target architecture models: in AM1, each processor can have a different memory size, and in AM2, each processor can also have a different speed (in addition to a different memory size). We design a three‐step heuristic for AM1, which adapts and extends previous work for homogeneous clusters [Gou C, Benoit A, Marchal L. Partitioning tree‐shaped task graphs for distributed platforms with limited memory. IEEE Trans Parallel Dist Syst 2020; 31(7): 1533–1544]. The changes we propose concern the assignment to processors (accounting for the different memory sizes) and the availability of suitable processors when splitting or merging subtrees. For AM2, we extend the heuristic for AM1 with a two‐phase local search approach. Phase A is a swap‐based hill climber, while (the optional) Phase B is inspired by iterated local search. We evaluate our heuristics for AM1 and AM2 with extensive simulations, and we demonstrate that exploiting the heterogeneity in the cluster significantly reduces the makespan, compared to the state of the art for homogeneous processors.

show abstract

Algorithms for tree-shaped task partition and allocation on heterogeneous multiprocessors

Wei

et al. 2023

J Supercomput

View full text Add to dashboard Cite

Task Tree Partition and Subtree Allocation for Heterogeneous Multiprocessors

Cited by 3 publications

References 23 publications

Mapping Tree-Shaped Workflows on Memory-Heterogeneous Architectures

Mapping Tree-Shaped Workflows on Memory-Heterogeneous Architectures

Mapping tree‐shaped workflows on systems with different memory sizes and processor speeds

Algorithms for tree-shaped task partition and allocation on heterogeneous multiprocessors

Contact Info

Product

Resources

About