Development of an Efficient and Fast Computational Tool for Multi-Precision Floating Point Numbers

This paper presents a MATLAB toolbox for multiple-precision arithmetic, enabling high-precision numerical computations beyond the standard double-precision format. The toolbox integrates seamlessly with the MATLAB computational ecosystem, offering a user-friendly interface and a diverse set of operations for applications that require enhanced precision, such as numerical analysis, scientific computing, and applied mathematics. The toolbox supports arbitrary precision for real and complex numbers and matrix operations in a computationally efficient manner. Features include adaptive precision control and automatic precision propagation. The benchmark results illustrate the computational accuracy and performance of the toolbox, highlighting its ability to mitigate numerical instabilities and round-off errors in sensitive computations. The toolbox can empower researchers and practitioners to tackle problems requiring extreme numerical accuracy without leaving MATLAB's familiar environment.