# Seven terabytes in three minutes

A real geospatial pipeline: 192,000 satellite files, resampled and written, start to finish.

Source: https://university.eugo.io/videos/seven-terabytes-in-three-minutes

---

- Level: Advanced
- Duration: 9m
- Topics: data-io, distributed-compute, performance
- Status: not yet recorded — the transcript below covers the same material

## What this video covers

This run processes roughly seven terabytes of Sentinel satellite imagery, about 192,000 files, applying a geospatial resampling step and writing results back to object storage. End to end it completes in about three minutes. The interesting part is not the headline number but where the time goes: most of it is I/O, which is why partitioning strategy mattered more than worker count past a certain point.
