Abstract
Stereotomography is a tomographic technique that can be used in the process of subsurface imaging, especially useful for oil and gas exploration. In this paper, we apply parallelization techniques on a stereotomography algorithm in an attempt to make it run faster on systems that support this paradigm. The program was subjected to hybrid parallelization using OpenMPand OpenMPI, decreasing runtimes for systems with both shared and distributed memory models. A scalability analysis was performed thereafter, with the aid of a state-of-the-art supercomputer, showing satisfactory speed-ups and scalability. The resulting program remained consistent with the sequential version and shows increase in efficiency for larger problem sizes.