Energy‐aware load balancing of parallel evolutionary algorithms with heavy fitness functions in heterogeneous CPU‐GPU architectures

Abstract

Summary By means of the availability of mechanisms such as Dynamic Voltage and Frequency Scaling (DVFS) and heterogeneous architectures including processors with different power consumption profiles, it is possible to devise scheduling algorithms that are aware of both runtime and energy consumption in parallel programs. In this paper, we propose and evaluate a multi‐objective (more specifically, a bi‐objective) approach to distribute the workload among the processing cores in a given heterogeneous parallel CPU‐GPU architecture. The aim of this distribution may be either to save energy without increasing the running time or to reach a trade‐off among time and energy consumption. The parallel programs considered here are master‐worker evolutionary algorithms where the evaluation of the fitness function for the individuals in the population demands the most part of the computing time. As many useful bioinformatics and data mining applications exhibit this kind of parallel profile, the proposed energy‐aware approach for workload scheduling could be frequently applied.

Design of a Standard and Programmatically Accessible Interface for Smart Meters to Allow Monitoring Automation of the Energy Consumed by the Execution of Computer Software

“…Mainly, three approaches can be applied to obtain consumed energy, either for the execution of a program or for other scenarios [2][3][4]: estimating the consumption through a customized model, aggregating the power consumed by the different parts of a system, or metering the consumption of the whole system globally, with the latter being widely used to measure the power consumed by personal computers [5][6][7][8][9][10][11][12][13][14][15]. However, the collection of energy measurements for this approach becomes cumbersome since although device manufacturers provide human-friendly interfaces, neither of them meets any standard nor allows programmatic access to the meter, which makes it difficult or even impossible for programs to be aware of their consumed power while being executed [16]. This paper aims to design such an interface to allow any program to access its current energy consumption transparently, regardless of the smart meter connected between the computer and the wall outlet.…”

Empirical analysis of machine learning-based energy efficient cloud load balancing architectures: a quantitative perspective

“…The model is compared with First Fit Decreasing (FFD) & OTS models and it is observed that the GSO model has 25% better energy efficiency, 15% better throughput, and 18% better load balancing degree when compared with these algorithms. This efficiency can be further improved by performing computations on fog devices as observed from (Bhuvaneswari and Akila, 2019) [8], wherein comparison of different fog-based load balancing algorithms w.r.t. their energy performance is studied.…”

Correlation of Performance Optimizations and Energy Consumption for Stencil-Based Application on Intel Xeon Scalable Processors

“…Works [26], [27] demonstrated bi--objective approaches to distribute the workload among processing cores in heterogeneous parallel platforms. The goal is to benefit from the trade-off between execution time and energy consumption, aiming at reducing the energy without increasing the execution time.…”