--- title: Walk on Decomposed Subdomains | daily.dev description: A research paper introducing 'Walk on Decomposed Subdomains', a hybrid Monte Carlo–deterministic solver for elliptic PDEs. The approach addresses the slow... canonical: https://daily.dev/posts/walk-on-decomposed-subdomains-83tselihn twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Walk on Decomposed Subdomains | daily.dev og:description: A research paper introducing 'Walk on Decomposed Subdomains', a hybrid Monte Carlo–deterministic solver for elliptic PDEs. The approach addresses the slow... og:url: https://daily.dev/posts/walk-on-decomposed-subdomains-83tselihn og:image: https://api.daily.dev/og/posts/83tSeliHN.png og:image:alt: Walk on Decomposed Subdomains og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Walk on Decomposed Subdomains **[Hacker News](https://daily.dev/sources/hn)** · 18 min read · 0 upvotes · 0 comments ## Summary A research paper introducing 'Walk on Decomposed Subdomains', a hybrid Monte Carlo–deterministic solver for elliptic PDEs. The approach addresses the slow convergence and high variance of grid-free Monte Carlo methods (like Walk on Spheres) by decomposing the domain into smaller subdomains, precomputing Poisson kernel-based solution operators per tile, and coupling subdomains via an absorbing Markov chain solved deterministically. This combines the geometry-handling advantages of grid-free Monte Carlo with the fast convergence of grid-based solvers. The method is trivially parallelizable per tile and exposes a tunable trade-off between Monte Carlo estimation cost and deterministic global solve cost. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is the Walk on Spheres algorithm used for in solving PDEs? Walk on Spheres (WoS) is a grid-free Monte Carlo method for solving elliptic PDEs like the Laplace equation with Dirichlet boundary conditions. It exploits the fact that the exit distribution of Brownian motion on a sphere is exactly uniform, letting a simulated walk hop from one inscribed sphere to the next until it nears the boundary, avoiding costly discretized Brownian motion simulation. _Explore practical breakdowns of Monte Carlo PDE solvers like Walk on Spheres on daily.dev when researching grid-free numerical methods._ ### How does decomposing a PDE domain into subdomains reduce the variance and cost of Monte Carlo random walk solvers? Splitting the domain into smaller non-overlapping tiles shortens individual random walks since they now stop at tile boundaries instead of traveling the full domain. Interface values between tiles are then resolved with a deterministic sparse linear solve over an absorbing Markov chain built from Poisson-kernel transition probabilities, eliminating the high variance of long walks while keeping per-tile Monte Carlo estimation embarrassingly parallel. _Developers weighing stochastic versus deterministic PDE solvers can track hybrid approaches like this via daily.dev._ ### Why do absorbing Markov chains let you solve for random walk expectations without simulating any walks? By the memoryless property of Markov chains, the expected absorbed value at a transient state equals the weighted average of expected absorbed values at its neighbors, giving a linear system Q x = R y where Q is the transient-to-transient transition block. Solving this system once yields the exact expected absorbed value everywhere, with zero sampling noise, since the transition matrix is invertible whenever every walk is eventually absorbed with probability one. _Anyone comparing sampling-based versus exact linear-solve techniques for random walks can follow this reasoning via daily.dev._ ## Similar posts on daily.dev - [Benders’ Decomposition 101: How to Crack Open a Stochastic Program That’s Too Big to Swallow Whole](https://daily.dev/posts/benders-decomposition-101-how-to-crack-open-a-stochastic-program-that-s-too-big-to-swallow-whole-ysaazrtpa) · Towards Data Science · 0 upvotes · 0 comments - [A hybrid quantum walk model unifying discrete and continuous quantum walks](https://daily.dev/posts/a-hybrid-quantum-walk-model-unifying-discrete-and-continuous-quantum-walks-2qylktl7e) · Nature · 0 upvotes · 0 comments - [GSWT](https://daily.dev/posts/gswt-jehhceujr) · Hacker News · 0 upvotes · 0 comments - [Why You Should Care About Partial Differential Equations \(PDEs\)](https://daily.dev/posts/why-you-should-care-about-partial-differential-equations-pdes--vknfoaeaq) · Hugging Face · 0 upvotes · 0 comments --- Tags: [#graphics-programming](https://daily.dev/tags/graphics-programming) [View this post on daily.dev](https://daily.dev/posts/walk-on-decomposed-subdomains-83tselihn) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Walk on Decomposed Subdomains","url":"https://daily.dev/posts/walk-on-decomposed-subdomains-83tselihn","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/walk-on-decomposed-subdomains-83tselihn"},"datePublished":"2026-08-03T17:22:51.722Z","dateModified":"2026-09-14T06:39:41.847Z","description":"A research paper introducing 'Walk on Decomposed Subdomains', a hybrid Monte Carlo–deterministic solver for elliptic PDEs. The approach addresses the slow...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/03a255189260b1083caaf0b8d66e16a0?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/03a255189260b1083caaf0b8d66e16a0?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hacker News","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Hacker News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/hn","url":"https://daily.dev/sources/hn"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/walk-on-decomposed-subdomains-83tselihn","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"graphics-programming","timeRequired":"PT18M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hacker News","item":"https://daily.dev/sources/hn"},{"@type":"ListItem","position":3,"name":"Walk on Decomposed Subdomains"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/walk-on-decomposed-subdomains-83tselihn#faq","mainEntity":[{"@type":"Question","name":"What is the Walk on Spheres algorithm used for in solving PDEs?","acceptedAnswer":{"@type":"Answer","text":"Walk on Spheres (WoS) is a grid-free Monte Carlo method for solving elliptic PDEs like the Laplace equation with Dirichlet boundary conditions. It exploits the fact that the exit distribution of Brownian motion on a sphere is exactly uniform, letting a simulated walk hop from one inscribed sphere to the next until it nears the boundary, avoiding costly discretized Brownian motion simulation. Explore practical breakdowns of Monte Carlo PDE solvers like Walk on Spheres on daily.dev when researching grid-free numerical methods."}},{"@type":"Question","name":"How does decomposing a PDE domain into subdomains reduce the variance and cost of Monte Carlo random walk solvers?","acceptedAnswer":{"@type":"Answer","text":"Splitting the domain into smaller non-overlapping tiles shortens individual random walks since they now stop at tile boundaries instead of traveling the full domain. Interface values between tiles are then resolved with a deterministic sparse linear solve over an absorbing Markov chain built from Poisson-kernel transition probabilities, eliminating the high variance of long walks while keeping per-tile Monte Carlo estimation embarrassingly parallel. Developers weighing stochastic versus deterministic PDE solvers can track hybrid approaches like this via daily.dev."}},{"@type":"Question","name":"Why do absorbing Markov chains let you solve for random walk expectations without simulating any walks?","acceptedAnswer":{"@type":"Answer","text":"By the memoryless property of Markov chains, the expected absorbed value at a transient state equals the weighted average of expected absorbed values at its neighbors, giving a linear system Q x = R y where Q is the transient-to-transient transition block. Solving this system once yields the exact expected absorbed value everywhere, with zero sampling noise, since the transition matrix is invertible whenever every walk is eventually absorbed with probability one. Anyone comparing sampling-based versus exact linear-solve techniques for random walks can follow this reasoning via daily.dev."}}]} ```