--- title: A Faster Merge Sort For Mesa 26.3 Vulkan Drivers To... description: Valve's Natalie Vock developed a custom merge-sort algorithm to replace the shared radix sort used by RADV, ANV, Turnip, and Lavapipe Mesa Vulkan drivers for... canonical: https://daily.dev/posts/a-faster-merge-sort-for-mesa-26-3-vulkan-drivers-to-benefit-ray-tracing-wzcmsd0hj twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: A Faster Merge Sort For Mesa 26.3 Vulkan Drivers To Benefit Ray-Tracing | daily.dev og:description: Valve's Natalie Vock developed a custom merge-sort algorithm to replace the shared radix sort used by RADV, ANV, Turnip, and Lavapipe Mesa Vulkan drivers for... og:url: https://daily.dev/posts/a-faster-merge-sort-for-mesa-26-3-vulkan-drivers-to-benefit-ray-tracing-wzcmsd0hj og:image: https://api.daily.dev/og/posts/wZcmsd0hj.png og:image:alt: A Faster Merge Sort For Mesa 26.3 Vulkan Drivers To Benefit Ray-Tracing 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. # A Faster Merge Sort For Mesa 26.3 Vulkan Drivers To Benefit Ray-Tracing **[Phoronix](https://daily.dev/sources/phoronix)** · 2 min read · 0 upvotes · 0 comments ## Summary Valve's Natalie Vock developed a custom merge-sort algorithm to replace the shared radix sort used by RADV, ANV, Turnip, and Lavapipe Mesa Vulkan drivers for sorting Morton code keys used in BVH construction for ray-tracing. The new sort operates within a single dispatch without intermediate barriers, avoids VRAM roundtrips for small problems (under 512 elements on RADV with a subgroup size of 64), and is faster and more VRAM-efficient for small BVHs, though slightly worse for larger problem sizes. After three months of review, the code merged into Mesa 26.3, benefiting all open-source Vulkan drivers. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What sorting algorithm change did Mesa 26.3 make for Vulkan ray-tracing BVH construction? Mesa 26.3 replaced the shared radix sort used by RADV, ANV, Turnip, and Lavapipe drivers with a custom merge-sort for Morton code keys used in BVH construction. Written by Natalie Vock of Valve's Linux graphics team, the merge-sort runs within a single dispatch with no intermediate barriers and skips VRAM roundtrips entirely for problems under 512 elements on RADV with a 64-wide subgroup, improving speed and VRAM efficiency for small BVHs. _Vulkan and Linux GPU driver developers can follow low-level graphics optimizations like this via daily.dev._ ### Which Mesa Vulkan drivers benefit from the new Morton code sorting implementation? The new merge-sort benefits all Mesa Vulkan drivers that previously shared the same radix sort routine for Morton code keys: RADV (Radeon), ANV (Intel), Turnip (Adreno), and Lavapipe. Since the sort was a shared piece of code across these drivers, the improvement to BVH construction performance applies uniformly across AMD, Intel, Qualcomm, and the software Vulkan implementation. _Developers tracking open-source GPU driver improvements can follow this progress through daily.dev._ ## Similar posts on daily.dev - [Marek Olšák At Valve Lands RADV Code That Can "Double Performance" With Some VRS Cases](https://daily.dev/posts/marek-ol-k-at-valve-lands-radv-code-that-can-double-performance-with-some-vrs-cases-21gz16lua) · Phoronix · 0 upvotes · 0 comments --- Tags: [#linux](https://daily.dev/tags/linux), [#game-development](https://daily.dev/tags/game-development), [#gpu](https://daily.dev/tags/gpu), [#vulkan](https://daily.dev/tags/vulkan), [#mesa](https://daily.dev/tags/mesa) [View this post on daily.dev](https://daily.dev/posts/a-faster-merge-sort-for-mesa-26-3-vulkan-drivers-to-benefit-ray-tracing-wzcmsd0hj) ```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":"A Faster Merge Sort For Mesa 26.3 Vulkan Drivers To Benefit Ray-Tracing","url":"https://daily.dev/posts/a-faster-merge-sort-for-mesa-26-3-vulkan-drivers-to-benefit-ray-tracing-wzcmsd0hj","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/a-faster-merge-sort-for-mesa-26-3-vulkan-drivers-to-benefit-ray-tracing-wzcmsd0hj"},"datePublished":"2026-08-31T10:12:27.759Z","dateModified":"2026-08-31T10:13:07.004Z","description":"Valve's Natalie Vock developed a custom merge-sort algorithm to replace the shared radix sort used by RADV, ANV, Turnip, and Lavapipe Mesa Vulkan drivers for...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/4e230ba58d5890854a06779c80b8404f?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/4e230ba58d5890854a06779c80b8404f?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Phoronix","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":"Phoronix","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/6a26ba379dd64932830aa0c9803dd7ee","url":"https://daily.dev/sources/phoronix"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/a-faster-merge-sort-for-mesa-26-3-vulkan-drivers-to-benefit-ray-tracing-wzcmsd0hj","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"linux,game-development,gpu,vulkan,mesa","timeRequired":"PT2M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Phoronix","item":"https://daily.dev/sources/phoronix"},{"@type":"ListItem","position":3,"name":"A Faster Merge Sort For Mesa 26.3 Vulkan Drivers To Benefit Ray-Tracing"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/a-faster-merge-sort-for-mesa-26-3-vulkan-drivers-to-benefit-ray-tracing-wzcmsd0hj#faq","mainEntity":[{"@type":"Question","name":"What sorting algorithm change did Mesa 26.3 make for Vulkan ray-tracing BVH construction?","acceptedAnswer":{"@type":"Answer","text":"Mesa 26.3 replaced the shared radix sort used by RADV, ANV, Turnip, and Lavapipe drivers with a custom merge-sort for Morton code keys used in BVH construction. Written by Natalie Vock of Valve's Linux graphics team, the merge-sort runs within a single dispatch with no intermediate barriers and skips VRAM roundtrips entirely for problems under 512 elements on RADV with a 64-wide subgroup, improving speed and VRAM efficiency for small BVHs. Vulkan and Linux GPU driver developers can follow low-level graphics optimizations like this via daily.dev."}},{"@type":"Question","name":"Which Mesa Vulkan drivers benefit from the new Morton code sorting implementation?","acceptedAnswer":{"@type":"Answer","text":"The new merge-sort benefits all Mesa Vulkan drivers that previously shared the same radix sort routine for Morton code keys: RADV (Radeon), ANV (Intel), Turnip (Adreno), and Lavapipe. Since the sort was a shared piece of code across these drivers, the improvement to BVH construction performance applies uniformly across AMD, Intel, Qualcomm, and the software Vulkan implementation. Developers tracking open-source GPU driver improvements can follow this progress through daily.dev."}}]} ```