--- title: WWDC26: Optimize custom machine learning operations with... description: A WWDC26 session covering advanced custom ML kernel development on Apple Silicon using Metal's TensorOps library. Topics include native support for quantized... canonical: https://daily.dev/posts/wwdc26-optimize-custom-machine-learning-operations-with-metal-tensors-apple-2d1pjak4c twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: WWDC26: Optimize custom machine learning operations with Metal tensors | Apple | daily.dev og:description: A WWDC26 session covering advanced custom ML kernel development on Apple Silicon using Metal's TensorOps library. Topics include native support for quantized... og:url: https://daily.dev/posts/wwdc26-optimize-custom-machine-learning-operations-with-metal-tensors-apple-2d1pjak4c og:image: https://api.daily.dev/og/posts/2D1PjaK4c.png og:image:alt: WWDC26: Optimize custom machine learning operations with Metal tensors | Apple 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. # WWDC26: Optimize custom machine learning operations with Metal tensors | Apple **[Apple Developer](https://daily.dev/sources/appledeveloper)** · 16 min read · 0 upvotes · 0 comments ## Summary A WWDC26 session covering advanced custom ML kernel development on Apple Silicon using Metal's TensorOps library. Topics include native support for quantized data types (4/8-bit integers, FP8) with MX scaling formats arriving in macOS/iOS 27, techniques for building a FlashAttention kernel using cooperative tensors and row reductions, and integrating custom Metal kernels into Core AI models via Python. The session also highlights the new neural accelerator in the M5 chip family and how TensorOps automatically leverages hardware acceleration across GPU generations. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [Teaching Automatic1111 to Speak Metal on an M1](https://daily.dev/posts/teaching-automatic1111-to-speak-metal-on-an-m1-45rpgxe1o) · Hacker News · 0 upvotes · 0 comments --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#data-science](https://daily.dev/tags/data-science), [#gpu](https://daily.dev/tags/gpu) [View this post on daily.dev](https://daily.dev/posts/wwdc26-optimize-custom-machine-learning-operations-with-metal-tensors-apple-2d1pjak4c) ```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":"WWDC26: Optimize custom machine learning operations with Metal tensors | Apple","url":"https://daily.dev/posts/wwdc26-optimize-custom-machine-learning-operations-with-metal-tensors-apple-2d1pjak4c","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/wwdc26-optimize-custom-machine-learning-operations-with-metal-tensors-apple-2d1pjak4c"},"datePublished":"2026-06-08T21:25:25.471Z","dateModified":"2026-06-08T21:34:14.927Z","description":"A WWDC26 session covering advanced custom ML kernel development on Apple Silicon using Metal's TensorOps library. Topics include native support for quantized...","image":"https://i.ytimg.com/vi/toO9j5NKg0c/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/toO9j5NKg0c/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Apple Developer","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":"Apple Developer","logo":"https://media.daily.dev/image/upload/s--CAtMViBl--/f_auto/v1711726806/logos/appledeveloper","url":"https://daily.dev/sources/appledeveloper"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/wwdc26-optimize-custom-machine-learning-operations-with-metal-tensors-apple-2d1pjak4c","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,data-science,gpu","timeRequired":"PT16M","video":{"@type":"VideoObject","name":"WWDC26: Optimize custom machine learning operations with Metal tensors | Apple","description":"A WWDC26 session covering advanced custom ML kernel development on Apple Silicon using Metal's TensorOps library. Topics include native support for quantized...","thumbnailUrl":"https://i.ytimg.com/vi/toO9j5NKg0c/sddefault.jpg","uploadDate":"2026-06-08T21:25:25.471Z","duration":"PT16M","url":"https://api.daily.dev/r/2D1PjaK4c","embedUrl":"https://www.youtube.com/embed/toO9j5NKg0c"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Apple Developer","item":"https://daily.dev/sources/appledeveloper"},{"@type":"ListItem","position":3,"name":"WWDC26: Optimize custom machine learning operations with Metal tensors | Apple"}]} ```