Authors:Â R. K. Malhotra, Amit D. Choudhary, Umashanker Ray
Abstract: Scientific computing has become a core pillar of modern science and engineering, enabling the numerical solution of complex mathematical models that are otherwise analytically intractable. With the exponential growth of data and problem complexity, traditional sequential algorithms are no longer sufficient. This has led to the development of high-performance algorithms that exploit advanced computer architectures, including multicore processors, graphics processing units (GPUs), and distributed memory systems. This paper presents a comprehensive review of scientific computing and high-performance algorithms, focusing on their mathematical foundations, computational challenges, and practical implementations. Key algorithmic paradigms such as parallel numerical linear algebra, iterative solvers, domain decomposition methods, and optimization techniques are discussed. The role of software frameworks and performance models in achieving scalability and efficiency is also examined. Furthermore, emerging trends including heterogeneous computing, machine learning–assisted solvers, and exascale computing are highlighted. The study aims to provide researchers and practitioners with a structured understanding of how scientific computing methodologies evolve alongside high-performance architectures, while also identifying future research directions and unresolved challenges.
Keywords: Scientific Computing, High-Performance Algorithms, Parallel Computing, Numerical Methods, Scalability
Full Issue
| View or download the full issue | PDF 1-15 |