Chiyuan Zhang, PhD student at MIT talks about his joint project with Shell using GPUs and deep learning to automatically detect subsurface faults from seismic traces for oil and gas exploration.
Using a Tesla K80 GPU, CUDA, cuBLAS and the cuDNN-accelerated Mocha.jl deep learning framework, the researchers were able to speed-up up their experiments nearly 40% over their CPU-only solution.
To learn more, watch Chiyuan’s presentation at the 2016 GPU Technology Conference.
Share your GPU-accelerated science with us at http://nvda.ly/Vpjxr and with the world on #ShareYourScience.
Watch more scientists and researchers share how accelerated computing is benefiting their work at http://nvda.ly/X7WpH
Share Your Science: Using Deep Learning to Automatically Detect Geophysical Features
Aug 19, 2016
Discuss (0)

Related resources
- GTC session: Borrowing from Astronomy to Improve Deep Learning Weather Prediction (Spring 2023)
- GTC session: Learn How to Create Features from Tabular Data and Accelerate your Data Science Pipeline* (Spring 2023)
- GTC session: GPU-Accelerating End-to-End Geospatial Workflows (Spring 2023)
- Webinar: Accelerating Research in the Earth and Space Sciences with AI
- Webinar: Open Tools for Machine Learning in Energy
- Webinar: AI for Weather and Climate Modeling at NVIDIA