A Framework for Application-Guided Task Management on Heterogeneous Embedded Systems

Abstract

In this article, we propose a general framework for fine-grain application-aware task management in heterogeneous embedded platforms, which allows integration of different mechanisms for an efficient resource utilization, frequency scaling, and task migration. The proposed framework incorporates several components for accurate runtime monitoring by relying on the OS facilities and performance self-reporting for parallel and iterative applications. The framework efficiency is experimentally evaluated on a real hardware platform, where significant power and energy savings are attained for SPEC CPU2006 and PARSEC benchmarks, by guiding frequency scaling and intercluster migrations according to the runtime application behavior and predefined performance targets. CCS Concepts: r Computer systems organization → Multicore architectures; Heterogeneous (hybrid) systems; r Software and its engineering → Process management;

Compiler-assisted adaptive program scheduling in big.LITTLE systems

“…Y-axis shows code size (Kb). machine (VM)/OS level [2,10,20,22,27,34]. By leveraging runtime information, the system can use environment information, unknown at compilation time, to solve SPha.…”

Application Control and Monitoring in Heterogeneous Multiprocessor Systems

“…Most of the frameworks are based on the Heartbeats concept and inherit its features, e.g. application monitors [9], [11], [13], [15], [18], [19]. In order to perform tradeoffs within a single application, multiple monitors of different types must be exposed, e.g., throughput, accuracy or error rate, but many frameworks do not support this functionality.…”

“…A small number of frameworks support this feature but they do not indicate as to whether the target is a maximisation or minimisation objective [9], [11], [13], [15], [19]. Furthermore, only some frameworks support concurrent applications, which is a common scenario in real-world systems [10], [13], [18], [19]. As a result, these approaches cannot be considered broadly applicable because they do not allow fully application-agnostic behaviour.…”

“…Current frameworks provide partial abstraction of RTM to device communication, but do not include both knobs and monitors to control hardware components at runtime [9], [12], [13], [19]. Moreover, most existing works do not operate on heterogeneous platforms, which provide both highperformance and energy-efficiency by combining conventional CPUs with other accelerators.…”

Performance prediction of application executed on GPUs using a simple analytical model and machine learning techniques

“…32 samples were collected of each bi-dimensional application. The matrix sizes of these applications were from 2 8 until 2 13 , increasing the matrix size in a step of 2 8 . Vector-Matrix applications were well characterized in Section 2.3.…”