Kaplan, MücahitAkman, Ali2024-04-222024-04-222024https://hdl.handle.net/11467/7250https://doi.org/10.1007/978-3-031-52760-9_20The High-Efficiency Video Coding (HEVC) standard represents a significant advancement in video compression technology, offering superior compression efficiency compared to its predecessors. This study investigates into the parallel implementation and efficacy of the Discrete Cosine Transform (DCT) process, a pivotal component of the HEVC standard’s transform step, when executed on GPU hardware. To enhance the performance of DCT, two distinct methods were introduced and implemented, making efficient use of shared memory resources within the GPU. A comprehensive analysis was conducted by measuring the execution times of these methods using a standardized test image. The study rigorously compared the performance of these methods across various scenarios and meticulously assessed their impact. In summary, this research contributes valuable insights into optimizing the DCT process within the HEVC standard through the parallel implementation on GPU hardware. The proposed methods exhibit promising results, demonstrating the potential to enhance the efficiency of HEVC’s transform step.eninfo:eu-repo/semantics/restrictedAccessDCT; GPU; HEVCConference Object281293N/A2-s2.0-8518953646610.1007/978-3-031-52760-9_20