Computational Science Stack Exchange is a question and answer site for scientists using computers to solve scientific problems. It only takes a minute to sign up.
Sign up to join this community
I am looking for a simple usage example/demo of OpenFOAM + CUDA and would like to understand how exactly OpenFOAM benefits from CUDA. The thing is I do not have any background in fluid dynamics and so finding it hard to get a high-level understanding of the matter.
Appreciate if you can point me to a relevant link. I think a kind of "Hello World" would be sufficient.
I suggest to start form this page, where you can find different project related to OpenFoam that use GPU.
Keep in mind that OpenFoam was born with the intent to resolve practical case in applied problems, think its typical user as an engineer. In this kind of program is normal to develop a layer that hide the implementation, so the user principal work is to set options. In fact if you try OpenFoam essentially you set option: describe the mesh proprieties, choose solver, ecc.. in its syntax, but it is not a real programming task as you do in normal language.
Imagine OpenFoam as a front-end and the implementation, over cpu or gpu, as a back-end. So when you use a gpu the goal is to not note difference from the front-end.
Another consideration is that in this kind of program is important to know what are you doing, i.e. know the software that has got an entry level barrier, but more important know physics and CFD. For example there are a lot of solver developed along the years and every solver is for a specific case. In addition, to use it with awareness it would be important to know the finite volume method that it is what OpenFoam uses.
At the end I report the link to Rapid CFD a gpu version of OpenFoam, so you can navigate into the code.
I would prefer to re-write a finite volume code directly using a library compatible with GPU. In fact, OpenFOAM has been written in C++ and optimized for CPU parallel usage.
In my opinion, it is better either to use an existing finite volume code optimized for GPU usage or to re-write one from zero while paying attention to the possible optimizations.
