SimWave FPGA - Acoustic Wave Propagator for Seismic Image Processing
- 0 Collaborators
Oil and Gas companies use seismic processing to create the subsurface models using computational methods like Full Waveform Inversion (FWI). This project aims to design and measure the performance of an FPGA-based accelerator that implements an acoustic wave propagator for seismic image processing. ...learn more
Project status: Concept
oneAPI, HPC, Artificial Intelligence
Intel Technologies
DevCloud,
DPC++,
Intel FPGA,
Intel vTune,
AI DevCloud / Xeon,
Intel Python
Overview / Usage
This project aims to design and measure the performance of an FPGA-based accelerator that implements an acoustic wave propagator for seismic image processing. The propagator will implement the finite difference method to solve the differential equations used to model the acoustic wave propagation.
If the final accelerator design improves the propagator performance, it will contribute to seismic image processing that uses methods like Reverse Time Migration (RTM) and Full Waveform Inversion (FWI).
Methodology / Approach
The idea is to use the Intel OneAPI framework to design the accelerator. The propagator algorithm will be written in DPC++ and compiled to target an FPGA. Simulation and measurements tools will be fundamental to change the design in order to optimize the propagator performance.
Technologies Used
We expect to use the OneAPI framework, DPC++, profiling and measurements tools, and eventually OpenCL too.