Table of Contents
Fetching ...

Mamute: high-performance computing for geophysical methods

João B. Fernandes, Antônio D. S. Oliveira, Mateus C. A. T. Silva, Felipe H. Santos-da-Silva, Vitor H. M. Rodrigues, Kleiton A. Schneider, Calebe P. Bianchini, João M. de Araujo, Tiago Barros, Ítalo A. S. Assis, Samuel Xavier-de-Souza

TL;DR

Mamute addresses the computational challenges of wave-equation-based geophysical methods by providing a high-performance, open-source C++ platform for 3D seismic modeling and full waveform inversion. The software combines MPI/OpenMP parallelism, advanced workload scheduling (including CTWS), fault tolerance via DeLIA, and memory-aware forward-wavefield management including checkpointing, to run efficiently on modern supercomputers. It solves the 3D acoustic wave equation with high-order spatial discretization and supports adjoint-state gradient computation through FWI using L-BFGS-b optimization, with configurable input parameters and data formats. The work demonstrates practical utility through an illustrative synthetic example and establishes Mamute as a scalable, widely usable tool for oil and gas exploration and reservoir monitoring, released under the MIT license with ongoing development and community contributions.

Abstract

Due to their high computational cost, geophysical applications are typically designed to run in large computing systems. Because of that, such applications must implement several high-performance techniques to use the computational resources better. In this paper, we present Mamute, a software that delivers wave equation-based geophysical methods. Mamute implements two geophysical methods: seismic modeling and full waveform inversion (FWI). It also supports high-performance strategies such as fault tolerance, automatic parallel looping scheduling, and distributed systems workload balancing. We demonstrate Mamute's operation using both seismic modeling and FWI. Mamute is a C++ software readily available under the MIT license.

Mamute: high-performance computing for geophysical methods

TL;DR

Mamute addresses the computational challenges of wave-equation-based geophysical methods by providing a high-performance, open-source C++ platform for 3D seismic modeling and full waveform inversion. The software combines MPI/OpenMP parallelism, advanced workload scheduling (including CTWS), fault tolerance via DeLIA, and memory-aware forward-wavefield management including checkpointing, to run efficiently on modern supercomputers. It solves the 3D acoustic wave equation with high-order spatial discretization and supports adjoint-state gradient computation through FWI using L-BFGS-b optimization, with configurable input parameters and data formats. The work demonstrates practical utility through an illustrative synthetic example and establishes Mamute as a scalable, widely usable tool for oil and gas exploration and reservoir monitoring, released under the MIT license with ongoing development and community contributions.

Abstract

Due to their high computational cost, geophysical applications are typically designed to run in large computing systems. Because of that, such applications must implement several high-performance techniques to use the computational resources better. In this paper, we present Mamute, a software that delivers wave equation-based geophysical methods. Mamute implements two geophysical methods: seismic modeling and full waveform inversion (FWI). It also supports high-performance strategies such as fault tolerance, automatic parallel looping scheduling, and distributed systems workload balancing. We demonstrate Mamute's operation using both seismic modeling and FWI. Mamute is a C++ software readily available under the MIT license.

Paper Structure

This paper contains 8 sections, 2 equations, 4 figures, 6 tables, 3 algorithms.

Figures (4)

  • Figure 1: (a) True velocity model and (b) acquisition geometry, source, and receiver coordinates, used in the Modeling and FWI examples.
  • Figure 2: Ricker source wavelet.
  • Figure 3: Seismogram for the shot $0$.
  • Figure 4: Initial (a) and final (b) velocity models from Mamute's FWI execution.