Analysis and Deployment of Applications Acceleration Environment for Xilinx Hardware-Accelerated Platforms
Adrian Santiago, Leire Muguira, Jaime Jiménez, Le Sun, Jesús Lázaro · 2022
High Performance Computing is a practice that consists of adding computing power above what any typical computer or workstation can provide. Greater performance is achieved by solving complex problems. The aim is to obtain a more efficient, reliable, and faster processing than traditional methods. The foundation of this technology is based on the parallelization of the processing units so that a higher performance per unit of time is achieved. This can be get with highperformance hardware or by combining simpler hardware, giving rise to clusters or supercomputers, depending on the form of grouping. Once the general acceleration process of an application and the methodology on which it is based have been understood, a complete analysis of the performance of a specific algorithm or program has been carried out and implemented using different programming languages and technologies. Besides, a series of conclusions and results related to the acceleration of applications through acceleration hardware have been obtained, determining its potential and the possible inconveniences it entails.