Planet Labs PBC (NYSE: PL) announced a collaboration with NVIDIA to develop a GPU-native AI engine for processing satellite imagery. The partnership aims to transform raw satellite data into analysis-ready insights within seconds rather than hours.

The collaboration centers on three main areas of development. Planet is working with NVIDIA to accelerate satellite data ground processing components including compositing, orthorectification, and atmospheric compensation using NVIDIA GPUs. The company plans to deploy this processing capability in cloud environments, at edge ground stations using NVIDIA Blackwell, or directly in space on the NVIDIA IGX Thor platform.

Planet is applying NVIDIA CorrDiff, a generative AI diffusion model, to its PlanetScope imagery to achieve super-resolution capabilities. The companies are also developing architectures to convert daily planetary data streams into AI embeddings, creating what they describe as a multi-dimensional vector map of Earth for intelligent searches.

"Planet is solving a massive, planetary-scale data problem by imaging the Earth every single day," said Will Marshall, co-founder and CEO of Planet. "By developing a GPU-native AI engine, we are unlocking the full potential of our petabyte-scale archive, delivering orders of magnitude speed improvements."

Planet has tested the NVIDIA IGX Jetson Thor module for space applications and plans to integrate the GPU into its next-generation Pelican satellites and upcoming Owl constellation to enable real-time insights directly from space.

The companies will present these technologies at NVIDIA GTC in San Jose. Planet will host a session titled "Accelerate Geospatial Workflows for Planetary Insight" on March 17 at 3:00 PM PDT.