Abstract:Energy-efficiency has become a major challenge in modern computer systems. To address this challenge, candidate systems increasingly integrate heterogeneous cores in order to satisfy diverse computation requirements by selecting cores with suitable features. In particular, single-ISA heterogeneous multicore processors such as ARM <a class="link-external link-http" href="http://big.LITTLE" rel="external noopener nofollow">this http URL</a> have become very attractive since they offer good opportunities in terms of performance and power consumption trade-off. While existing works already showed that this feature can improve system energy-efficiency, further gains are possible by generalizing the principle to higher levels of heterogeneity. The present paper aims to explore these gains by considering single-ISA heterogeneous multicore architectures including three different types of cores. For this purpose, we use the Samsung Exynos Octa 5422 chip as baseline architecture. Then, we model and evaluate Cortex A7, A9, and A15 cores using the gem5 simulation framework coupled to McPAT for power estimation. We demonstrate that varying the level of heterogeneity as well as the different core ratio can lead to up to 2.3x gains in energy efficiency and up to 1.5x in performance. This study further provides insights on the impact of workload nature on performance/energy trade-off and draws recommendations concerning suitable architecture configurations. This contributes in fine to guide future research towards dynamically reconfigurable HSAs in which some cores/clusters can be disabled momentarily so as to optimize certain metrics such as energy efficiency. This is of particular interest when dealing with quality-tunable algorithms in which accuracy can be then traded for compute effort, thereby enabling to use only those cores that provide the best energy-efficiency for the chosen algorithm.

Optimizing the Energy Efficiency of Power Supply in Heterogeneous Multicore Chips with Integrated Switched-Capacitor Converters

Early-Stage Planning of Switched-Capacitor Converters in a Heterogeneous Chip

Design and Implementation of High-Efficiency and Low-Power DC-DC Converter with PWM/PFM Modes

Evaluation of Power Losses in Different CCM Mode Single-Phase Boost PFC Converters Via a Simulation Tool

Fast Performance Evaluation for Switched-Capacitor Converter Based on Convolutional Neural Networks

Multi-Optimization Power Management for Chip Multiprocessors

A Layout Optimization Method to Reduce Commutation Inductance of Multi-Chip Power Module Based on Genetic Algorithm

Optimal Scheduling of Integrated Energy Systems with Multiple CCHPs for High Efficiency and Low Emissions.

A Multiplexing Modulation Strategy Based On Single-Phase Ac/Dc Converter With Optimized E-Capless Function

Three Phase PFC Converter with Multilevel Inductive Switching Network

Optimization Scheme Based on High-Frequency Link Interconnection of Submodules

Exploration of Performance and Energy Trade-offs for Heterogeneous Multicore Architectures

A Fast Decoupling Capacitor Budgeting Algorithm for Robust On-Chip Power Delivery.

Resonant-Switched Capacitor Converters for Chip-Scale Power Delivery: Design and Implementation

Efficiency-Oriented Optimized Design and Control of Hybrid FSBB-CLLC Converters with Partial Power Processing Capability

Reduced Switching-Frequency Modulation and Circulating Current Suppression for Modular Multilevel Converters

HW/SW Co-Optimization for Stencil Computation: Beginning with a Customizable Core

Energy-Aware Loop Scheduling and Assignment for Multi-Core, Multi-Functional-Unit Architecture

Optimal Multivariable Control for Wide Output Regulation and Full-Range Efficiency Optimization in LCC–LCC Compensated Wireless Power Transfer Systems

Optimized Control for MMCs with Reduced Power Loss and Extended Lifetime

Switching Characterization and Power Loss Optimization for Modular Multiactive Bridge Converter under Common Phase Shift Control