Re: [deal.II] Parallel implementation

2017-09-21 Thread Anna Avdeeva
Dear Wolfgang, to compute values of the electric field at the receivers I follow the strategy of ASPECT code as you suggested To do this I sum the current_point_values across processors and divide by the number of processors that contain point p as following // Reduce all collected values into

Re: [deal.II] Re: about deal.II with CUDA C programming acceleration

2017-09-21 Thread Chih-Che Chueh
> I'm going to add that installing a compiler will really only be the >> first >> step. After that, of course, comes the task of understanding how to >> program these graphics cards, and all of the other things that require >> interacting with the appropriate software levels. You wi

[deal.II] Re: Parallel distributed triangulation: informations on part of the mesh which are not owned

2017-09-21 Thread Bruno Turcksin
Giovanni, It looks like you are doing a lot of things similar to what we do in this project https://github.com/ORNL-CEES/DataTransferKit In particular, one of our subroutine allows the user to ask for the value of the field at arbitrary points. We need our code to use GPUs, so we based our algo

Re: [deal.II] Parallel distributed triangulation: informations on part of the mesh which are not owned

2017-09-21 Thread Wolfgang Bangerth
- Does Cellid contain some information about the owners? Could it be useful for a process to pass Cellid to another process CellId? I mean: modifying the find_point_location could the Cellid give information about where to start looking for that point? CellIds are globally unique, so you can

[deal.II] Distribute constraints on PETSc wrapper

2017-09-21 Thread Jaekwang Kim
Hi, all I was practicing using Petsc library through deal.ii wrappers, While, step 17 gives me clear explanation, I am having trouble in distributing hanging node constraints on Petsc system matrix, I was modifying step-9 hyperbolic equation solvers with PETSc matrix, but when I compile my

Re: [deal.II] restrict a solution in some range?

2017-09-21 Thread Yiyang Zhang
Hello Prof. Bangerth, Thank you very much for the explanation. It really helps to understand FEM and the dealii library. I am glad to find this group to discuss things on FEM and dealii. Best, Yiyang On Thursday, September 21, 2017 at 9:45:48 AM UTC-5, Wolfgang Bangerth wrote: > > > > Suppose

Re: [deal.II] restrict a solution in some range?

2017-09-21 Thread Wolfgang Bangerth
Suppose we need to solve some PDE about some function u(r). Following the weak form of an equation, we obtain and solve a linear system (A*x=b). What we really solve are the coefficients U_j (this corresponds to the vector x in the linear system). Thus I am assuming that the solution vector (n

Re: [deal.II] Re: about deal.II with CUDA C programming acceleration

2017-09-21 Thread Wolfgang Bangerth
I'm going to add that installing a compiler will really only be the first step. After that, of course, comes the task of understanding how to program these graphics cards, and all of the other things that require interacting with the appropriate software levels. You will need to