Prob-GParareal: A Probabilistic Numerical Parallel-in-Time Solver for Differential Equations
Guglielmo Gattiglio [email protected], Lyudmila Grigoryeva [email protected], Massimiliano Tamborrino [email protected] · 2025
Read Prob-GParareal: A Probabilistic Numerical Parallel-in-Time Solver for Differential Equations by Guglielmo Gattiglio [email protected], Lyudmila Grigoryeva [email protected], Massimiliano Tamborrino [email protected] (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-03945.txt.
Introduction
Guglielmo Gattiglio and two co-authors devote a 2025 research paper to Prob-GParareal, a solver that adds uncertainty estimates to parallel-in-time solving. Solving an evolution equation step by step is serial by nature, wasting modern parallel hardware. Parareal methods split the time axis across processors and fix the seams in rounds, and GParareal uses Gaussian processes to model those fixes.
Prob-GParareal extends the idea so uncertainty travels with the answer. Gaussian processes model the fix, letting numerical uncertainty move across time segments and yielding forecasts with probabilities attached, not just one path. Uncertain starting conditions are supported too. The method still works with classical solvers, so it fits existing Parareal setups. A theory analysis derives error bounds and cost. Five benchmark systems test accuracy and robustness, covering chaotic, stiff, and bifurcation problems. A nearest-neighbour variant trades some statistical machinery for speed, shown on a PDE example.
Catalog record
| CREATOR | Guglielmo Gattiglio [email protected], Lyudmila Grigoryeva [email protected], Massimiliano Tamborrino [email protected] |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 152 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-03945