GPU Development Engineer

Develop CPU-GPU parallelization architectures for different multi-physics modules based on a deep understanding of the underlying solution algorithms, data flows, and memory usage of each module. Implement the parallel processing architectures designed for specific modules, and ensure high parallel efficiency through profiling, optimization, and tuning on the full suite of representative application cases. An M.Sc, M.Tech, or PhD degree in computer science, applied mathematics, physics, or a computational modeling discipline, with emphasis on GPU parallelization and programming. Demonstrable talent and skill in software design, development, and testing. A high level of skill in profiling, debugging, and tuning GPU programs. A high level of proficiency in C/C++, Python, CUDA, and OpenCL, and a readiness to learn other languages, such as Fortran 90. A high level of proficiency in MPI and OpenMP. Desire to learn about different computational physics modeling methods and technologies. Ability to work independently as well as in close collaboration with other computational scientists. You will work with other talented and capable peers developing various modules of our flagship multi-physics software. The work environment is stimulating, collaborative, and rewarding, with good opportunities for learning, growth, and diversification of expertise and knowledge. This position is ideal for someone with a passion for scientific computing, high-performance computing, and taking parallel processing techniques to their limits.