<![CDATA[This systematic mapping review examines application-driven parallel Differential Evolution (DE) as a scalable optimization approach for engineering, energy, computing, and intelligent systems. The review is based on a Scopus-only corpus searched on 24 May 2026 using TITLE-ABS-KEY queries related to DE, parallel computing, distributed computing, GPU acceleration, and scalable optimization. From 220 records, 25 unique studies published between 2021 and 2026 were included after screening by year, document type, language, relevance, and methodological alignment. Because the corpus contains heterogeneous benchmark studies, application studies, and hybrid intelligent-system frameworks, the evidence was synthesized thematically rather than through meta-analysis. The synthesis distinguishes explicit parallel-DE implementations from broader scalable DE applications in which scalability is achieved through decomposition, model reformulation, or integration with computationally expensive systems. The findings indicate that GPU acceleration, CUDA, MPI migration, cooperative coevolution, Spark/Hadoop distribution, and resource-aware dispatch frequently report reduced computational cost while preserving or improving solution quality under the evaluated conditions. However, cross-study comparison remains limited by heterogeneous benchmarks, incomplete hardware reporting, inconsistent scalability metrics, and uneven baseline selection. This review contributes a cross-domain taxonomy linking application constraints, DE mechanisms, computational architectures, evaluation metrics, and reported outcomes.]]>
Copyrights © 2026
