Case Study: Oak Ridge National Laboratory accelerates global settlement mapping with NVIDIA GPU-powered AI

A 21st Century Planning Tool Built On Ai

Oak Ridge National Laboratory (ORNL) needed far more computing power to analyze massive volumes of satellite imagery and large-scale geographic data for understanding human settlement patterns, population distribution, and future resource needs. NVIDIA helped support this work with Titan Supercomputer systems powered by NVIDIA Tesla GPUs, enabling ORNL to move beyond CPU-based limits and process imagery fast enough to keep up with growing data demands.

Using NVIDIA GPU acceleration, ORNL was able to run computationally intensive feature extraction and settlement-mapping algorithms at speeds up to 200x faster than before. This let researchers process a city like Addis Ababa in under 20 seconds instead of days, rapidly identify buildings and settlements, support humanitarian response and disaster mapping, and uncover previously uncounted populations.

Oak Ridge National Laboratory

Budhendra Bhaduri

Director of the Urban Dynamics Institute and Group Leader, Geographic Information Science and Technology (GIST) Group