Low-Cost Approximation Algorithms for Scheduling Independent Tasks on Hybrid Platforms

Canon, Louis-Claude; Marchal, Loris; Vivien, Frédéric

doi:10.1007/978-3-319-64203-1_17

Cited by 7 publications

(15 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Low complexity algorithms, which are closer to our work, have been studied in [5,11] and achieve approximation ratios respectively equal to 2 and 2 + √ 2. For tasks with precedence constraints, Kedad-Sidhoum et al [18] provided a tight 6-approximation algorithm based on linear programming.…”

Section: Related Worksupporting

confidence: 61%

“…Schedulers for independent tasks on hybrid platforms have first been proposed [5,8,11]. Some of them have been adapted for task graphs: [20] extends the algorithm of [5] to the (offline) scheduling of task graphs, while [2] adapts an online scheduler for independent tasks on hybrid platforms [17] to obtain a competitive online scheduler for task graphs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Online Scheduling of Task Graphs on Hybrid Platforms

Canon

Marchal

Simon

et al. 2018

Euro-Par 2018: Parallel Processing

Self Cite

View full text Add to dashboard Cite

Abstract. Modern computing platforms commonly include accelerators. We target the problem of scheduling applications modeled as task graphs on hybrid platforms made of two types of resources, such as CPUs and GPUs. We consider that task graphs are uncovered dynamically, and that the scheduler has information only on the available tasks, i.e., tasks whose predecessors have all been completed. Each task can be processed by either a CPU or a GPU, and the corresponding processing times are known. Our study extends a previous 4 m/k-competitive online algorithm [2], where m is the number of CPUs and k the number of GPUs (m ≥ k). We prove that no online algorithm can have a competitive ratio smaller than m/k. We also study how adding flexibility on task processing, such as task migration or spoliation, or increasing the knowledge of the scheduler by providing it with information on the task graph, influences the lower bound. We provide a (2 m/k + 1)-competitive algorithm as well as a tunable combination of a system-oriented heuristic and a competitive algorithm; this combination performs well in practice and has a competitive ratio in Θ( m/k). Finally, simulations on different sets of task graphs illustrate how the instance properties impact the performance of the studied algorithms and show that our proposed tunable algorithm performs the best among the online algorithms in almost all cases and has even performance close to an offline algorithm.

show abstract

Section: Related Worksupporting

confidence: 61%

Section: Introductionmentioning

confidence: 99%

Online Scheduling of Task Graphs on Hybrid Platforms

Canon

Marchal

Simon

et al. 2018

Euro-Par 2018: Parallel Processing

Self Cite

View full text Add to dashboard Cite

show abstract

“…Finally, there has recently been a number of papers targeting both a low computing cost and a bounded approximation ratio, such as related works . Cheriere and Saule propose a 2‐approximation algorithm under the assumption that no task is longer on any resource than the optimal makespan.…”

Section: Related Workmentioning

confidence: 99%

Fast approximation algorithms for task‐based runtime systems

Beaumont

Eyraud-Dubois

Kumar

2018

Concurrency and Computation

View full text Add to dashboard Cite

Summary In High Performance Computing, heterogeneity is now the norm with specialized accelerators like GPUs providing efficient computational power. Resulting complexity led to the development of task‐based runtime systems, where complex computations are described as task graphs, and scheduling decisions are made at runtime to perform load balancing between all the resources of the platforms. Developing good scheduling strategies, even at the scale of a single node, and analyzing them both theoretically and in practice are expected to have a very high impact on the performance of current HPC systems. The special case of 2 kinds of resources, typically CPUs and GPUs, is already of great practical interest. The scheduling policy HeteroPrio has been proposed in the context of fast multipole computations (FMM) and has been extended to general task graphs with very promising results. In this paper, we provide a theoretical study of the performance of HeteroPrio, by proving approximation bounds compared to the optimal schedule, both in the case of independent tasks and in the case of general task graphs. Interestingly, our results establish that spoliation (a technique that enables resources to restart uncompleted tasks on another resource) is enough to prove bounded approximation ratios for a list scheduling algorithm on two unrelated resources, which is known to be impossible otherwise. This result holds true both for the independent and dependent tasks graphs. Additionally, we provide an experimental evaluation of HeteroPrio on real task graphs from dense linear algebra computation, which establishes its strong performance in practice.

show abstract

“…For hybrid architectures, a 6‐approximation algorithm was proposed by Kedad‐Sidhoum et al In the case of independent tasks, there is a

false(\frac{4}{3} + \frac{1}{3 k} false)

‐approximation algorithm based on dynamic programming. Several other faster algorithms have been proposed, achieving approximation ratio of two (in the work of Canon et al) and 3.41 (in the work of Beaumont et al). If the tasks arrive in an on‐line order, a 4‐competitive algorithm was presented by Chen et al for hybrid architectures without precedence relations.…”

Section: Related Workmentioning

confidence: 99%

“…In the case of independent tasks, a systematic comparison of various heuristics has been performed by Braun et al Specifically, the authors examined 11 different heuristics and provided a good basis for comparison and insights on circumstances why a technique outperforms another. Finally, experimental comparisons of the algorithms in the more recent literature for independent tasks can be found …”

Section: Related Workmentioning

confidence: 99%

Generic algorithms for scheduling applications on heterogeneous platforms

Amarís

Lucarelli

Mommessin

et al. 2018

Concurrency and Computation

View full text Add to dashboard Cite

Summary We study the problem of executing an application represented by a precedence task graph on a parallel machine composed of standard computing cores and accelerators. Both off‐line and on‐line settings are addressed by proposing generic scheduling approaches. In the first case, we establish strong lower bounds on the worst‐case performance of a known approach based on Linear Programming and replace the greedy List Scheduling policy used in this approach by a better task ordering. Although this modification leads to the same approximability guarantees, it performs much better in practice. We also extend this algorithm to more types of computing units, achieving an approximation ratio which depends on the number of different types. In the on‐line case, tasks arrive in any order which respects the precedence relations and the scheduler has to take irrevocable decisions about their allocation and execution. We propose the first on‐line scheduling algorithm taking into account precedences, which is based on adequate rules for selecting the type of processor where to allocate the tasks. Finally, all the previous algorithms have been experimented on a large number of simulations built on actual libraries, assessing their good practical behavior with respect to the state‐of‐the‐art solutions and baseline algorithms.

show abstract

Low-Cost Approximation Algorithms for Scheduling Independent Tasks on Hybrid Platforms

Cited by 7 publications

References 28 publications

Online Scheduling of Task Graphs on Hybrid Platforms

Online Scheduling of Task Graphs on Hybrid Platforms

Fast approximation algorithms for task‐based runtime systems

Generic algorithms for scheduling applications on heterogeneous platforms

Contact Info

Product

Resources

About