Alert button
Picture for James Paul Ahrens

James Paul Ahrens

Alert button

Deep Learning-Based Feature-Aware Data Modeling for Complex Physics Simulations

Dec 08, 2019
Qun Liu, Subhashis Hazarika, John M. Patchett, James Paul Ahrens, Ayan Biswas

Figure 1 for Deep Learning-Based Feature-Aware Data Modeling for Complex Physics Simulations
Figure 2 for Deep Learning-Based Feature-Aware Data Modeling for Complex Physics Simulations
Figure 3 for Deep Learning-Based Feature-Aware Data Modeling for Complex Physics Simulations
Figure 4 for Deep Learning-Based Feature-Aware Data Modeling for Complex Physics Simulations

Data modeling and reduction for in situ is important. Feature-driven methods for in situ data analysis and reduction are a priority for future exascale machines as there are currently very few such methods. We investigate a deep-learning based workflow that targets in situ data processing using autoencoders. We propose a Residual Autoencoder integrated Residual in Residual Dense Block (RRDB) to obtain better performance. Our proposed framework compressed our test data into 66 KB from 2.1 MB per 3D volume timestep.

* Accepted as a research poster at the International Conference for High Performance Computing, Networking, Storage, and Analysis (SC19) 
Viaarxiv icon