The public defence of Nandinbaatar Tsog’s licentiate thesis in Computer Science

Datum: 2019-12-18
Tid: 09.15
Plats: room Kappa

The public defence of Nandinbaatar Tsog’s licentiate thesis in Computer Science will take place at Mälardalen University on December 18 2019 at 09.15 in room Kappa, MDH Västerås.

Title: “Improving On-Board Data Processing using CPU-GPU Heterogeneous Architectures for Real-Time Systems”.  

Serial number: 286.

The examining committee consists of Adjunct Professor Masoumeh Ebrahimi, University of Turku and Professor Sasikumar Punnekkat, Mälardalen University. Among the members of the examining committee, Professor Juha Plosila, University of Turku has been appointed the faculty examiner.

Reserve; Doctor Farhang Nemati, Örebro University.

Abstract:

This thesis investigates the efficacy of heterogeneous computing architectures in real-time systems. The goal of the thesis is twofold. First, investigate various characteristics of the Heterogeneous System Architectures (HSA) compliant reference platforms focusing on computing performance and power consumption. The investigation is focused on the new technologies that could boost on-board data processing systems in satellites and spacecraft. Second, enhance the usage of heterogeneous processing units by introducing a technique for static allocation of parallel segments of tasks.

The investigation and experimental evaluation shows that an efficient use of GPU allocation for parallel segments of the tasks is the most energy efficient compared to any other allocation of these segments. The investigations are conducted and considered under different types of environments, such as process level isolated environment, different software stacks, including kernels, and various task set scenarios. Furthermore, the evaluation results indicate that a balanced use of heterogeneous processing units (CPU and GPU) could improve the schedulability of task sets up to 90% with the proposed allocation techniques.