satellite imagery
3 published articles
Geospatial AI
Ai2's OlmoEarth platform turns satellite inference into a distributed systems problem
Ai2's OlmoEarth Platform tackles the unique scale of geospatial inference by splitting work into CPU-driven data prep, GPU inference, and CPU postprocessing. A continent-wide wildfire risk map ran with a 155x speedup at a fraction of a cent per square kilometer. The architecture, tolerance for failures, and roadmap show how the team is operationalizing foundation models for environmental organizations.
2026-08-02
Remote Sensing AI
Smarter tokens just cut satellite AI costs by 3x without losing accuracy
Ai4earth's OlmoEarth v1.1 cuts compute costs by up to 3x over v1 for satellite image analysis, using a smarter token merging technique that maintains performance. The updated models enable cheaper planet-scale map refreshes for partner organizations.
2026-07-13
Artificial Intelligence
Ai2 cuts satellite imagery AI costs by 3x with a smarter token trick
Ai2's OlmoEarth v1.1 reduces compute costs by up to 3x compared to v1, enabling cheaper large-scale map refreshes. The key innovation is merging resolution-based tokens for Sentinel-2 imagery, cutting token counts by a factor of three while preserving performance through modified pre-training.
2026-07-05