--- title: How Meta’s AI Models are Powering the First Wave of... description: Lawrence Berkeley National Laboratory's SYNAPS-I project is using Meta's open-source SAM 3 and DINOv3 models to tackle a massive scientific data bottleneck at... canonical: https://daily.dev/posts/how-meta-s-ai-models-are-powering-the-first-wave-of-genesis-mission-projects-q4bovdtfl twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: How Meta’s AI Models are Powering the First Wave of Genesis Mission Projects | daily.dev og:description: Lawrence Berkeley National Laboratory's SYNAPS-I project is using Meta's open-source SAM 3 and DINOv3 models to tackle a massive scientific data bottleneck at... og:url: https://daily.dev/posts/how-meta-s-ai-models-are-powering-the-first-wave-of-genesis-mission-projects-q4bovdtfl og:image: https://api.daily.dev/og/posts/Q4BOvDtFl.png og:image:alt: How Meta’s AI Models are Powering the First Wave of Genesis Mission Projects 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. # How Meta’s AI Models are Powering the First Wave of Genesis Mission Projects **[Hacker News](https://daily.dev/sources/hn)** · 5 min read · 1 upvotes · 0 comments ## Summary Lawrence Berkeley National Laboratory's SYNAPS-I project is using Meta's open-source SAM 3 and DINOv3 models to tackle a massive scientific data bottleneck at DOE light and neutron source facilities, which now generate tens of petabytes annually. The pipeline combines DINOv3's self-supervised contextual understanding with SAM's precise pixel-level segmentation, fine-tuned on scientific imaging data and deployed across 300 A100 GPUs at national supercomputing facilities. What previously took a month of expert annotation per time step now completes in ~15 minutes, enabling real-time 3D reconstruction during live experiments. A demonstration on grapevine drought response using micro-CT scans showed the pipeline automatically identifying xylem vessels at cellular scale. Meta's open-source approach is critical because national labs must keep prepublication data on secure government infrastructure, making downloadable and self-hostable models essential. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How are Meta's SAM 3 and DINOv3 models being used for scientific image analysis at national laboratories? The SYNAPS-I team at Lawrence Berkeley National Laboratory fine-tuned SAM 3 and DINOv3 on scientific imaging data from DOE beamlines, then deployed them across 300 A100 GPUs at NERSC. DINOv3 identifies structures and their context in an image while SAM 3 draws precise pixel-level boundaries around them, together producing fully labeled 3D volumes from raw scans in about 15 minutes. _Teams evaluating open-source vision models for domain-specific pipelines can track real deployments like this on daily.dev._ ### Why do national labs prefer open-source AI models like SAM and DINO over cloud-based AI services for scientific data? National laboratories must keep prepublication research data and AI models on government infrastructure rather than external cloud services for security reasons. Because SAM and DINO are open source, the SYNAPS-I team can download, fine-tune, and deploy them entirely within their own secure computing environments, adapting models trained on natural images to scientific domains like micro-CT scans. _Developers weighing open-source versus hosted AI models for sensitive workloads follow cases like this on daily.dev._ ### How much faster is AI-based image segmentation compared to manual annotation for scientific micro-CT scans? What previously required a month of expert annotation per time step now takes about 15 minutes using the fine-tuned SAM and DINO pipeline, demonstrated on micro-CT scans of grapevine stems to study drought response by tracking xylem vessel changes. This speed lets researchers study dynamic biological processes at the pace data is actually collected. _Anyone measuring real-world AI productivity gains in scientific workflows can follow examples like this on daily.dev._ ## Similar posts on daily.dev - [How DINO and SAM are Helping Modernize Essential Medical Triage Practices](https://daily.dev/posts/how-dino-and-sam-are-helping-modernize-essential-medical-triage-practices-5nfbljdar) · AI at Meta · 0 upvotes · 0 comments - [Genesis Open Models](https://daily.dev/posts/genesis-open-models-qkhtfua7x) · Hacker News · 2 upvotes · 0 comments --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#computer-vision](https://daily.dev/tags/computer-vision) [View this post on daily.dev](https://daily.dev/posts/how-meta-s-ai-models-are-powering-the-first-wave-of-genesis-mission-projects-q4bovdtfl) ```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":"How Meta’s AI Models are Powering the First Wave of Genesis Mission Projects","url":"https://daily.dev/posts/how-meta-s-ai-models-are-powering-the-first-wave-of-genesis-mission-projects-q4bovdtfl","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/how-meta-s-ai-models-are-powering-the-first-wave-of-genesis-mission-projects-q4bovdtfl"},"datePublished":"2026-07-21T18:08:17.353Z","dateModified":"2026-09-14T06:14:04.974Z","description":"Lawrence Berkeley National Laboratory's SYNAPS-I project is using Meta's open-source SAM 3 and DINOv3 models to tackle a massive scientific data bottleneck at...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f8f5f8232fb8d9c1a6215fe16947813b?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f8f5f8232fb8d9c1a6215fe16947813b?_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/how-meta-s-ai-models-are-powering-the-first-wave-of-genesis-mission-projects-q4bovdtfl","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,computer-vision","timeRequired":"PT5M"} {"@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":"How Meta’s AI Models are Powering the First Wave of Genesis Mission Projects"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/how-meta-s-ai-models-are-powering-the-first-wave-of-genesis-mission-projects-q4bovdtfl#faq","mainEntity":[{"@type":"Question","name":"How are Meta's SAM 3 and DINOv3 models being used for scientific image analysis at national laboratories?","acceptedAnswer":{"@type":"Answer","text":"The SYNAPS-I team at Lawrence Berkeley National Laboratory fine-tuned SAM 3 and DINOv3 on scientific imaging data from DOE beamlines, then deployed them across 300 A100 GPUs at NERSC. DINOv3 identifies structures and their context in an image while SAM 3 draws precise pixel-level boundaries around them, together producing fully labeled 3D volumes from raw scans in about 15 minutes. Teams evaluating open-source vision models for domain-specific pipelines can track real deployments like this on daily.dev."}},{"@type":"Question","name":"Why do national labs prefer open-source AI models like SAM and DINO over cloud-based AI services for scientific data?","acceptedAnswer":{"@type":"Answer","text":"National laboratories must keep prepublication research data and AI models on government infrastructure rather than external cloud services for security reasons. Because SAM and DINO are open source, the SYNAPS-I team can download, fine-tune, and deploy them entirely within their own secure computing environments, adapting models trained on natural images to scientific domains like micro-CT scans. Developers weighing open-source versus hosted AI models for sensitive workloads follow cases like this on daily.dev."}},{"@type":"Question","name":"How much faster is AI-based image segmentation compared to manual annotation for scientific micro-CT scans?","acceptedAnswer":{"@type":"Answer","text":"What previously required a month of expert annotation per time step now takes about 15 minutes using the fine-tuned SAM and DINO pipeline, demonstrated on micro-CT scans of grapevine stems to study drought response by tracking xylem vessel changes. This speed lets researchers study dynamic biological processes at the pace data is actually collected. Anyone measuring real-world AI productivity gains in scientific workflows can follow examples like this on daily.dev."}}]} ```