Acceleration of multi-component multiple-precision arithmetic with branch-free algorithms and SIMD vectorization

2026/03/16 by Tomonori Kouya · 1 voice
Computer Science · Mathematics · #cs.MS #math.NA

paper · pdf · doi:10.1007/978-3-032-30521-3_32

arxiv published 2026/03/16 · arxiv updated 2026/05/07

Abstract

Multiple-precision floating-point branch-free algorithms can significantly accelerate multi-component arithmetic implemented by combining hardware-based binary64 and binary32, particularly for triple- and quadruple-precision computations. In this study, we achieved benchmark results on x86 and ARM CPU platforms to quantify the accelerations achieved in linear computations and polynomial evaluation by integrating these algorithms.

Citations

• Sparse Iterative Solvers Using High-Precision Arithmetic with Quasi Multi-Word Algorithms
• Automatic Verification of Floating-Point Accumulation Networks

Discussions

• Tomonori Kouya: Acceleration of multi-component multiple-precision arithmetic with branch-free algorithms and SIMD vectorization https://arxiv.org/abs/2603.14926 https://arxiv.org/pdf/2603.14926 https [bsky, 0 points, 0 comments]

Related

• A Survey of Numerical Methods Utilizing Mixed Precision Arithmetic
• Dynamic Precision Math Engine for Linear Algebra and Trigonometry Acceleration on Xtensa LX6 Microcontrollers
• MPCR: Multi-Precision Computations Package in R
• White Paper from Workshop on Large-scale Parallel Numerical Computing Technology (LSPANC 2020): HPC and Computer Arithmetic toward Minimal-Precision Computing
• A Pilot Study on Tunable Precision Emulation via Automatic BLAS Offloading