Subsequence generation of a string is an important task in various disciplines of Computer Science and Bio-informatics. Speeding is desirable in subsequence generation from a string and is essential for several applications. The most effective way of extracting substring from a string is by parallel processing principles. This paper uses Message Passing Interface (MPI) and Compute Unified Device Architecture (CUDA) techniques to investigate the parallelization of subsequence extraction from strings. In this study, for data distribution and coordination, MPI and CUDA are used in the best possible way. A novel parallel approach involving data segmentation, load balancing, and Graphical Processing Unit (GPU) acceleration as part of the design is addressed. The scalability and efficiency of the approach are demonstrated by the experimental findings on CPU-GPU system, which show significant speedup and better performance. Through quicker parallelized string data analysis, this work tackles the need for high-performance computing in subsequence analysis and offers potential benefits for data analytics, natural language processing, and genomics.