The last preparation of CUDA programming, two key points of CUDA programming are introduced.

The architecture and design mode of GPU is introduced with examples, in comparison with the CPU architecture.

Warning: a set of stupid errors occur when compiling c++ and OpenFOAM.

A crash into Parallelism, some terms are listed and very basic concepts are introduced.

2D wave equation with Dirichlet boundary condition is solved in C++ (with blas and Lapack), Python (with numpy), Matlab and Julia. Euler backward in time and central finite difference in

This is a very quick review on the CPU's fundamentals, in order to compare with GPU's when learning CUDA.

Need to program on CUDA, so try to compile some examples first.

1D shockTube case is tested with different schemes on OpenFOAM. And a recommended differencing scheme setting in OpenFOAM is proposed.

Last blog is only 1-equation. In this blog we are going to multi-equation and artificial viscosity.

This is the essence of CFD, advecting with the discontinuities due to inviscid fluid PDEs. Several computational schemes are introduced with fortran scripts.