Efficient cloud deployment of large-scale finite-difference simulations research project. The aim of this project is to explore some novel approaches for efficient deployment of large-scale simulations on cloud-based resources
The finite difference method is widely used for large-scale seismic wave equation modelling in complex media. Many commercial and academic finite-difference modellers exist, and they are often highly optimised. Current research topics in seismology require finite-difference modellers to account for the effects of attenuation, anisotropy and elasticity in complex geological structures. Efficient implementations require a sophisticated domain decomposition approach designed to solve heterogeneous problems on heterogeneous compute clusters.
The work will extend an existing finite-difference scheme, exploring the implementation of several new ideas for rapid and flexible large-scale deployment. It will be written in C++17 and tested in cloud-based simulations for scalability.
Applications are invited from students with C++ programming skills who are interested in algorithm design and High-Performance Computing. The Ideal candidate will be working towards a BSc in Computer Science or related field (2:1/1st predicted)
Send your CV via the button below. Use the subject "Application – SCR –- HPC Summer Internship "