Task Scheduling for Energy Consumption Constrained Parallel Applications on Heterogeneous Computing Systems

Time‐constrained and reliability aware energy minimization scheduling algorithm for heterogeneous multiprocessor environments

“…Zhu and Aydin 22 proposed two heuristics based on task‐level utilization to minimize the energy consumption of periodic real‐time systems while maintaining the reliability of the system. The problem of optimizing time reliability under energy consumption constraints in heterogeneous multiprocessor systems was considered by Quan et al, 23 where the authors provided them with sub‐energy constraints based on the weights of each task and prove through mathematical derivation that their strategy guarantees the successful execution of all tasks.…”

“…In this article, we refer to previous work 23,27 and classify the energy consumption into static and dynamic energy consumption. The power of a processor at a frequency $$$ f $$$ is calculated as follow: $$$$ P(f)={P}_s+\beta \left({P}_{f_i}+{P}_d\right), $$$$ $$$$ {P}_d={C}_e{f}^m, $$$$ where $$$ {P}_s $$$ denotes the static power of the processor, which is generally caused by current leakage, $$$ {P}_{f_i} $$$ and $$$ {P}_d $$$ denote frequency‐independent and frequency‐dependent dynamic power, respectively.…”

“…Four real scientific workflows are used for the experimental dataset, 26 and the structure of these workflow applications is shown in Figure 4. The experimental setup is referred to previous works, 3,23 and we set the parameters of the workflow and processor as follows: $$$ 10\;\mathrm{ms}\le {w}_{i,k}\le 100\;\mathrm{ms} $$$, $$$ 10\;\mathrm{ms}\le d{t}_{i,j}\le 100\;\mathrm{ms} $$$, $$$ 0.03\le {P}_{k, fi}\le 0.07 $$$, $$$ 0.8\le {C}_{k,e}\le 1.2 $$$, $$$ 2.5\le {m}_k\le 3.0 $$$, $$$ {f}_{k,\max }=1 $$$ GHz, and all frequencies are discrete with an accuracy of 0.01 $$$ GHz $$$. We write Java simulation programs on the WorkflowSim 34 simulation platform to perform the experiments in this section.…”

Minimizing Energy Consumption Scheduling Algorithm of Workflows With Cost Budget Constraint on Heterogeneous Cloud Computing Systems

“…Cloud computing systems are becoming increasingly heterogeneous with the emergence of new and fast VMs [24]. The task scheduling problem in heterogeneous cloud computing systems is complex due to different execution rates among VMs [25].…”

New Matrix Based Algorithm for Calculation of Importance Measures

“…The application of BDD in importance analysis of system and IMs calculation have been considered in [5,8]. Other approach of this computational complexity decreasing is the use of parallel procedure [9,10]. The correlation of the parallel algorithms and matrix procedures has been studied in [11].…”