scholarly journals Communication and Computation Aware Task Scheduling Framework for Heterogeneous Computing

2019 ◽  
Vol 8 (2) ◽  
pp. 1-12
Author(s):  
Suhelah Sandokji Suhelah Sandokji
Keyword(s):  
Task Scheduling ◽  
High Performance ◽  
Heterogeneous Computing ◽  
Scheduling Algorithm ◽  
Programming Models ◽  
Critical Aspect ◽  
The Third ◽  
Computing Unit ◽  
Task Scheduling Algorithm ◽  
Performance Computing

The heterogeneous Computing (HC) is the promised paradigm for high performance computing. In HC the vastly different architectures and programming models of each type of the computing unit, present several challenges in achieving collaborative computing. Task scheduling is the main critical aspect in managing these challenges. In this paper, a Communication and Computation Aware task scheduler framework (CCATSF) is introduced. The proposed task scheduling framework consist of four parts; the first of which is the resource monitor, the second is the resources manager, the third is the task scheduler and the fourth the dispatcher. We also introduce DVR-HEFT algorithm a new hybrid task scheduling algorithm, on which the framework is based. Our results indicate that CCATSF framework based on algorithm is able to reduce the scheduler's makespan without increasing the algorithm's time complicity.

scholarly journals HSIP: A Novel Task Scheduling Algorithm for Heterogeneous Computing

2016 ◽  
Vol 2016 ◽  
pp. 1-11 ◽  
Author(s):  
Guan Wang ◽  
Yuxin Wang ◽  
Hui Liu ◽  
He Guo
Keyword(s):  
Task Scheduling ◽  
High Performance ◽  
Heterogeneous Computing ◽  
Scheduling Algorithm ◽  
Scheduling Algorithms ◽  
Selection Policy ◽  
Task Scheduling Algorithm ◽  
Task Duplication ◽  
Task Priority ◽  
Low Efficiency

High-performance heterogeneous computing systems are achieved by the use of efficient application scheduling algorithms. However, most of the current algorithms have low efficiency in scheduling. Aiming at solving this problem, we propose a novel task scheduling algorithm for heterogeneous computing named HSIP (heterogeneous scheduling algorithm with improved task priority) whose functionality relies on three pillars: (1) an improved task priority strategy based on standard deviation with improved magnitude as computation weight and communication cost weight to make scheduling priority more reasonable; (2) an entry task duplication selection policy to make the makespan shorter; and (3) an improved idle time slots (ITS) insertion-based optimizing policy to make the task scheduling more efficient. We evaluate our proposed algorithm on randomly generated DAGs, using some real application DAGs by comparison with some classical scheduling algorithms. According to the experimental results, our proposed algorithm appears to perform better than other algorithms in terms of schedule length ratio, efficiency, and frequency of best results.

Task Scheduling in Heterogeneous Computing Systems Based on Machine Learning Approach

2020 ◽  
Vol 34 (12) ◽  
pp. 2051012
Author(s):  
Hui Xie ◽  
Li Wei ◽  
Dong Liu ◽  
Luda Wang
Keyword(s):  
Task Scheduling ◽  
Heterogeneous Computing ◽  
Scheduling Algorithm ◽  
Computing System ◽  
Scheduling Problem ◽  
Computing Systems ◽  
Task Scheduling Algorithm ◽  
Machine Learning Approach ◽  
Heterogeneous Computing Systems ◽  
Regression Theory

Task scheduling problem of heterogeneous computing system (HCS), which with increasing popularity, nowadays has become a research hotspot in this domain. The task scheduling problem of HCS, which can be described essentially as assigning tasks to the proper processor for executing, has been shown to be NP-complete. However, the existing scheduling algorithm suffers from an inherent limitation of lacking global view. Here, we reported a novel task scheduling algorithm based on Multi-Logistic Regression theory (called MLRS) in heterogeneous computing environment. First, we collected the best scheduling plans as the historical training set, and then a scheduling model was established by which we could predict the following schedule action. Through the analysis of experimental results, it is interpreted that the proposed algorithm has better optimization effect and robustness.

Communication-aware task scheduling algorithm for heterogeneous computing

2017 ◽  
Vol 10 (4/5) ◽  
pp. 298 ◽  
Author(s):  
Yulu Yang ◽  
Wenjing Ma ◽  
Kezhao Zhao ◽  
Qiankun Dong ◽  
Tehui Huang ◽  
...  
Keyword(s):  
Task Scheduling ◽  
Heterogeneous Computing ◽  
Scheduling Algorithm ◽  
Task Scheduling Algorithm

ON MESSAGE PACKAGING IN TASK SCHEDULING FOR DISTRIBUTED MEMORY PARALLEL MACHINES

2001 ◽  
Vol 12 (03) ◽  
pp. 285-306 ◽  
Author(s):  
NORIYUKI FUJIMOTO ◽  
TOMOKI BABA ◽  
TAKASHI HASHIMOTO ◽  
KENICHI HAGIHARA
Keyword(s):  
Task Scheduling ◽  
High Performance ◽  
Distributed Memory ◽  
Parallel Machines ◽  
Scheduling Algorithm ◽  
Parallel Programs ◽  
Parallel Program ◽  
Interprocessor Communication ◽  
Task Scheduling Algorithm ◽  
Software Overhead

In this paper, we report a performance gap betweeen a schedule with small makespan on the task scheduling model and the corresponding parallel program on distributed memory parallel machines. The main reason of the gap is the software overhead in the interprocessor communication. Therefore, speedup ratios of schedules on the model do not approximate well to those of parallel programs on the machines. The purpose of the paper is to get a task scheduling algorithm that generates a schedule with good approximation to the corresponding parallel program and with small makespan. For this purpose, we propose algorithm BCSH that generates only bulk synchronous schedules. In those schedules, no-communication phases and communication phases appear alternately. All interprocessor communications are done only in the latter phases, and thus the corresponding parallel programs can make better use of the message packaging technique easily. It reduces many software overheads of messages form a source processor to the same destination processor to almost one software overhead, and improves the performance of a parallel program significantly. Finally, we show some experimental results of performance gaps on BCSH, Kruatrachue's algorithm DSH, and Ahmad et al's algorithm ECPFD. The schedules by DSH and ECPFD are famous for their small makespans, but message packaging can not be effectively applied to the corresponding program. The results show that a bulk synchronous schedule with small makespan has advantages that the gap is small and the corresponding program is a high performance parallel one.

Sign in / Sign up

Export Citation Format

Share Document