makepath is excited to announce that we recently made Xarray-Spatial’s raster calculations 300 times faster with Open Source libraries, Nvidia’s CUDA toolkit, and a Nvidia GPU.
This is part of makepath’s efforts to speed up processing times for common raster analysis functions, as well as leverage GPUs to speed up geospatial data processing and analysis.
The Results: Nvidia GPU + CUDA Toolkit
Viewshed operations using CUDA algorithms were over 300 times faster than traditional CPU algorithms.
To get these results, we used a combination of CUDA algorithms on a Nvidia T4 GPU, and ray tracing with OptiX through RTXpy.
Demo: Real-time Viewshed and Hillshade results on Crater Lake National Park. With each click, the Viewshed is calculated from that point. The illumination source for Hillshade calculations circles around the crater.
Open Source Projects Involved
Three Open Source projects are an active part of makepath’s work with GPUs:
NVIDIA’s source code behind OptiX is Open Source, which paves the way for the Open Source community to access GPU capabilities for Open Source projects.
RTXpy is an RTX wrapper for Python that the makepath team contributes to, built off the RTX source code.
RTXpy uses Nvidia’s RTX technology to enable Open Source ray tracing capabilities.
Xarray-Spatial is a GIS library for raster analysis and makepath’s current test ground for the capabilities of GPUs.
What We Worked On
We used Open Source GIS operations from Xarray-Spatial for testing.
Viewshed finds all visible locations in an input raster surface from a specific place within the raster. This geospatial operation answers the question of, “What can an observer see from this specific location?”
Hillshade makes a shaded relief from a surface raster, taking the illumination source angle and shadows into account. This geospatial operation answers the question of, “If an illumination source (like the sun) was over there, what would the terrain look like?”.
Want to learn more about the work makepath is doing with GPUs?
Send us your questions and comments at email@example.com
- Machine Learning for Change Detection: Part 1
- GPU-Enhanced Geospatial Analysis
- Open Source Machine Learning Tools (Updated for 2023)
- Getting Started with Open Source (Updated for 2023)
- The History of Open Source GIS: An Interactive Infographic (Updated for 2023)
- Superpowered GIS: ESRI’s ArcGIS + Open Source Spatial Analysis Tools.
- Seniors at Risk: Using Spatial Analysis to Identify Pharmacy Deserts
- Open Source Spatial Analysis Tools for Python: A Quick Guide (Updated for 2022)