Exploration on Task Scheduling Strategy for CPU-GPU Heterogeneous Computing System

Fang, Juan; Zhang, Jiaxing; Lu, Shuaibing; Hui, Zhao

doi:10.1109/isvlsi49217.2020.00063

Cited by 5 publications

(4 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…✓ The amount of processing time required depends on the number of million instructions (IMs), which are calculated at compilation time, and the processing power of the underlying hardware in MIPS. The processing time for task 𝑛 𝑖 when it is run on 𝑆 𝑗 is determined by equation (1). Equation ( 1) also shows that the node 𝑛 𝑖 's execution time is influenced by the underlying used 𝑆.…”

Section: Application Modelmentioning

confidence: 99%

“…This can include using a combination of CPUs, GPUs, and other specialized hardware devices such as FPGAs or DSPs. Heterogeneous computing platforms are designed to allow these different types of hardware to work together in order to increase the overall performance and efficiency of the system [1]. One of the common applications of heterogeneous computing is in the field of high-performance computing, where the combination of different types of hardware can be used to solve complex scientific and engineering problems.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Production Scheduling on Heterogeneous Computing Environment Using Modified GRASP

Kafafy

Saad

2023

Preprint

View full text Add to dashboard Cite

Heterogeneous computing environment refers to the use of multiple computing Sockets with different capabilities or characteristics in a parallel computing system. The production of task scheduling is one of the key issues with heterogeneous computing systems. This production of task scheduling problem desires to map tasks to heterogeneous machines in a way that will optimize the system's overall performance, such as minimization the schedule length of execution time. Because the task scheduling problem is NP-hard, intelligent algorithms are used to solve it, allowing us to achieve at a somewhat optimal result. To handle task scheduling in heterogeneous computing systems, this work adopted two algorithms one of them is a Greedy Randomized-based Simulated Annealing algorithm and the other is a GRASP-based Tabu Search algorithm. Additionally, greedy initial solutions with relatively optimized have taken the place of the random starting population. To enhance the capabilities of the Simulated Annealing or Tabu search Algorithm, the random initial solution has also been replaced by greedy initial solution with relatively optimal solutions. Results from testing the proposed approach on random graphs and graphs from real-world applications in heterogeneous computing systems with a variety of features showed that GRASP based Tabu Search was significantly more efficient than GRASP based Simulated annealing and the two algorithms more efficient than previous scheduling algorithms.

show abstract

Section: Application Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Production Scheduling on Heterogeneous Computing Environment Using Modified GRASP

Kafafy

Saad

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In a more recent study, authors in [13] propose a task scheduling strategy based on a genetic algorithm for CPU-GPU heterogeneous computing platforms. Bao et al [5] propose a dynamic task scheduling stratgy for Heterogeneous System Architectures (HSA) and evaluate their approach on data-parallel applications.…”

Section: Related Workmentioning

confidence: 99%

Scheduling for heterogeneous systems in accelerator-rich environments

Yesil

Öztürk

2021

J Supercomput

View full text Add to dashboard Cite

The world is creating ever more data and the applications are required to deal with ever-increasing datasets. To process such datasets heterogeneous and manycore accelerators are being deployed in various computing systems to improve energy efficiency. In this work, we present a runtime management system designed for such heterogeneous systems with manycore accelerators. More specifically, we design a resource-based runtime management system that considers application characteristics and respective execution properties on the nodes and accelerators. We propose scheduling heuristics and run time environment solutions to achieve better throughput and reduced energy in computing systems with different accelerators. We give implementation details about our framework; show different scheduling algorithms, and present experimental evaluation of our system. We also compare our approaches with an optimal scheme where integer linear programming approach has been implemented for mapping applications on the heterogeneous system. While it is possible to extend the proposed framework to a wide variety of accelerators, our initial focus is on Graphics Processing Units (GPUs). Our experimental evaluations show that including accelerator support in the management framework improves energy consumption and execution time significantly. We believe that this approach has the potential to provide an effective solution for next generation accelerator-based computing systems.

show abstract

“…Dynamic scheduling aims to effectively partition work across devices during execution, which has attracted more and more attentions recently. Many researches have concentrated on dynamic scheduling strategies designed for taskparallel applications, such as work-stealing scheduling [18], speedup-based scheduling [19], locality-aware scheduling [20], feature-aware scheduling [21], load-aware scheduling [22], energy-aware scheduling [23]. Recently some dynamic scheduling strategies designed for data-parallel applications have also been proposed.…”

Section: Introductionmentioning

confidence: 99%

Efficient Inter-Device Task Scheduling Schemes for Multi-Device Co-Processing of Data-Parallel Kernels on Heterogeneous Systems

2021

View full text Add to dashboard Cite

Heterogeneous systems consisting of multiple multi-core CPUs and many-core accelerators have recently come into wide use, and more and more parallel applications are developed in such a heterogeneous system. To fully utilize multiple compute devices to cooperatively and concurrently execute data-parallel kernels on heterogeneous systems, a feedback-based dynamic and elastic task scheduling scheme is proposed, which can provide a better load balance, a greater device utilization, and a lower scheduling overhead by flexibly and dynamically adjusting the workload between devices during execution. The proposed method is more suitable for data-parallel kernels whose computation and data are uniformly distributed, but is less suitable for data-parallel kernels whose computation and data are non-uniformly distributed. Thus, an asynchronous-based dynamic and elastic task scheduling scheme is proposed, which can avoid device underutilization, load imbalance across devices, and frequent kernel launches, interdevice data transfers and inter-device synchronizations by dynamically adjusting the chunk size according to the performance change during runtime. A series of experiments are conducted with 8 representative parallel applications on a hybrid CPU-GPU-MIC system, the results show that the proposed two interdevice task scheduling schemes can achieve the efficient CPU-GPU-MIC co-processing of different parallel applications by effectively partitioning work across devices.INDEX TERMS Data-parallel kernels, heterogeneous systems, many-core accelerators, multi-core CPUs, multi-device co-processing, parallel applications, task scheduling.

show abstract

Exploration on Task Scheduling Strategy for CPU-GPU Heterogeneous Computing System

Cited by 5 publications

References 19 publications

Production Scheduling on Heterogeneous Computing Environment Using Modified GRASP

Production Scheduling on Heterogeneous Computing Environment Using Modified GRASP

Scheduling for heterogeneous systems in accelerator-rich environments

Efficient Inter-Device Task Scheduling Schemes for Multi-Device Co-Processing of Data-Parallel Kernels on Heterogeneous Systems

Contact Info

Product

Resources

About