#CME-253A Introduction to high-performance-computing (HPC) for mathematical, computational, physical sciences and engineering within the field of computational fluid dynamics (CFD) and mechanics; particular focus on parallel computing using GPU accelerators and 3-D code development with application to nonlinear PDEs related to Earth science dynamical systems; evaluation of performance limiters and discussion on basic optimization techniques; finite-difference discretization (stencil codes) and accelerated iterative methods. Programming languages: MATLAB, CUDA C, (MPI). Hands-on approach: starting from diffusion and wave propagation physics, the goal being to achieve a 3-D Stokes flow utilizing C CUDA in order to leverage the GPU computing power. Pre-requisites: basic programming skills (e.g. MATLAB) and Scientific computing. Basic knowledge of compiled languages (C) and CFD are a plus. Terms: Sum 2019 | Units: 1 | Grading: Satisfactory/No Credit Instructors: Raess, L. (PI)
somonesummers / cme-253a Goto Github PK
View Code? Open in Web Editor NEWIntroduction to High Performance Computing and Parallel (GPU) Computing