Direct Finite-Time Contraction (Step-Log) Profiling--Driven Optimization of Parallel Schemes for Nonlinear Problems on Multicore Architectures

TL;DR

This paper introduces a data-driven, profiling-based parameter tuning method for parallel iterative schemes solving nonlinear problems, improving convergence, stability, and robustness without extensive problem-specific tuning.

Contribution

It proposes a novel, training-free tuning framework using finite-time contraction profiling for parallel nonlinear solvers, enhancing performance and reproducibility.

Findings

Consistent convergence rate improvements across test problems

Enhanced stability and robustness demonstrated

Effective parameter ranking via stability scores

Abstract

Efficient computation of all distinct solutions of nonlinear problems is essential in many scientific and engineering applications. Although high-order parallel iterative schemes offer fast convergence, their practical performance is often limited by sensitivity to internal parameters and the lack of reproducible tuning procedures. Classical parameter selection tools based on analytical conditions and dynamical-system diagnostics can be problem-dependent and computationally demanding, which motivates lightweight data-driven alternatives. In this study, we propose a parameterized single-step bi-parametric parallel Weierstrass-type scheme with third-order convergence together with a training-free tuning framework based on Direct finite-time contraction (step-log) profiling. The approach extracts Lyapunov-like finite-time contraction information directly from solver trajectories via step norms and step-log ratios, aggregates the resulting profiles over micro-launch ensembles, and ranks parameter candidates using two compact scores: the stability minimum S_min and the stability moment S_mom. Numerical results demonstrate consistent improvements in convergence rate, stability, and robustness across diverse nonlinear test problems, establishing the proposed profiling-based strategy as an efficient and reproducible alternative to classical parameter tuning methods.