Performance and power consumption evaluation of concurrent queue implementations in embedded systems

Abstract

Embedded and high performance computing (HPC) systems face many common challenges. One of them is the synchronization of the memory accesses in shared data. Concurrent queues have been extensively studied in the HPC domain and they are used in a wide variety of HPC applications. In this work, we evaluate a set of concurrent queue implementations in an embedded platform, in terms of execution time and power consumption. Our results show that by taking advantage of the embedded platform specifications, we achieve up to 28.2 % lower execution time and 6.8 % less power dissipation in comparison with the conventional lockbased queue implementation. We show that HPC applications utilizing concurrent queues can be efficiently implemented in embedded systems and that synchronization algorithms from the HPC domain can lead to optimal resource utilization of embedded platforms.

A Systematic Methodology for Optimization of Applications Utilizing Concurrent Data Structures

“…The last category is the Lock-less Synchronization Decisions that groups the design options for lock-less data structures. Client-Server synchronization model refers to the synchronization method proposed in [12]: An on-chip core is dedicated to play the role of the "server" and it is the only one that accesses the concurrent data structure directly. The rest of the on-chip cores are "clients", which send request messages to the server for specific operations (e.g., insert, remove, etc.)…”

“…The communication takes place through on-chip message passing between the cores. The application interface for this model is implemented through a set of macros, as described in [12]. Elimination-backoff Stack is an algorithm for concurrent stacks proposed in [13].…”

“…Each SHAVE has its own copy of these registers and the size of each one is 4Â64 bit words. They can be used to implement a client-server synchronization model, as described in Section 3.1 and presented in [12]. Therefore, with respect to the hardware specifications that can be used for synchronization, Myriad chip provides Custom/Platform-specific locks and Message Passing communication support (Table 3).…”

Mobile GPU Computing Based Filter Bank Convolution for Three-Dimensional Wavelet Transform

“…Existing research topics of SoC include performance analysis (A. Alejandro Acosta & Francisco Almeida, 2014b;Papadopoulos et al, 2014), power consumption (Grasso, Radojkovic, Rajovic, Gelado, & Ramirez, 2014;Papadopoulos et al, 2014;Zhan, Lung, & Srivastava, 2014), etc. Similar with regular GPU in desktop or notebook, SoC embedded GPU (SoC GPU) is responsible for graphics processing for SoC (Giles & Reguly, 2014).…”