RESEARCH ON THE EFFICIENCY OF USING GPGU TECHNOLOGIES IN CUSTOM ADOBE AFTER EFFECTS PLUG-INS

Abstract

The paper examines the features of using parallel computing technologies based on the computer’s graphics subsystem in Adobe After Effects plug-ins from third-party developers. The analysis showed that when creating plug-ins for the Adobe After Effects environment, developers quite often complicate the code, trying to transfer parallel computing from traditional means of the central processor to the functional blocks of the GPU, which at the present stage is considered absolutely correct. However, upon detailed examination, it turns out that the efficiency that the user receives from the operation of such plug-ins is by no means higher than the efficiency of plug-ins that have much simpler code that traditionally runs on the CPU. This situation is due to the lack of analysis of the features of the operation of parallel code in the graphics subsystem environment when processing various visual effects, and, as a consequence, the lack of clear recommendations on the necessity or, conversely, the inexpediency of working with the GPU hardware part in the code. This article presents a study of factors that affect the effectiveness of using parallel computing on GPUs in Adobe After Effects plugins. Recommendations are given on the feasibility of using GPUs depending on the specialization of the plugin related to the processing of a particular visual effect.