Performance Analysis of Arm Architecture Based Mobile Processor With Multi-Thread Image Compression

Abstract

The increasing use of social media in recent years has brought the problem of transmitting picture and audio data on the Internet. In order to reduce the burden on the Internet during data transfer, it is necessary to either increase the bandwidth of the Internet or reduce the size of the data by compressing it. In this study, the performance change due to multi-core usage of an ARM architectured mobile processor is analyzed on the image compression process with Haar Wavelet. The results showed that the use of multi-cores in the compression process reduces runtime but leads to an increase in processor temperature. In addition, it is observed that the working time of the scenario where the number of threads and the number of cores is equal works faster than the other scenarios.