To enhance the performance of audio compression and decompression, the proposed research aims at
optimising Apple Lossless Audio Codec (ALAC) algorithm by using CUDA architecture provided by NVIDIA. The
CPU-intensive processes of ALAC such as prediction, residual coding and entropy encoding take more time;
therefore, the performance is increased by using graphics processing units through parallel processing with CUDA.
The research saves a lot of time in processing in comparison with traditional implementations of CPUs, and this is
done by redesigning key algorithm elements in order to utilize GPU parallelism. Experiment outings reveal the
performance is more efficient and higher throughput and audio quality remains unaffected. The opportunity to encode
and decode high-fidelity audio on faster real-time is beneficial to applications, such as streaming, storage, and
playback, on GPU-capable devices. The paper focuses on the way CUDA-based acceleration can enhance the work
of audio codecs.
Keywords : SIMT, PCM, CUDA, ALAC, Audio decoding, Audio encoding
Author : K Prabhakar Reddy
Title : NVIDIA CUDA Architecture for Apple Lossless Audio Codec Algorithm Optimisation
Volume/Issue : 2025;01(02)
Page No : 05-07