# Meta's Open Vision Models Cut a Month of Beamline Annotation to Fifteen Minutes at Berkeley Lab

Segment Anything 3 and DINOv3 run on 300 Nvidia A100s at the National Energy Research Scientific Computing Center inside the Department of Energy's Genesis Mission.

- Published: 2026-08-23T06:19:43.709Z
- Canonical: https://polylog.news/ai/2026-08-23/meta-s-open-vision-models-cut-a-month-of-beamline-annotation
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Meta AI](https://ai.meta.com/blog/genesis-mission-lawrence-berkeley-national-laboratory-segment-anything-dino/), [Berkeley Lab News Center](https://newscenter.lbl.gov/2026/07/22/berkeley-lab-to-lead-13-new-genesis-mission-ai-projects/), [Tech Times](https://www.techtimes.com/articles/321456/20260724/meta-open-source-ai-cuts-doe-beamline-analysis-month-minutes.htm)

X-ray and neutron imaging produces more data than anyone can label. Until recently, a single time step of a beamline experiment took a human expert roughly a month to annotate, which meant that scientists analyzed a small sample of what the…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-08-23/meta-s-open-vision-models-cut-a-month-of-beamline-annotation (subscription information: https://polylog.news/pricing).