You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

65 lines
2.3 KiB

7 years ago
# opencl_fdfd
7 years ago
**opencl_fdfd** is a 3D Finite Difference Frequency Domain (FDFD)
7 years ago
electromagnetic solver implemented in Python and OpenCL.
7 years ago
7 years ago
* Arbitrary distributions of the following:
3 years ago
* Dielectric constant (`epsilon`)
* Magnetic permeabilty (`mu`)
* Perfect electric conductor (`PEC`)
* Perfect magnetic conductor (`PMC`)
7 years ago
* Variable-sized rectangular grids
* Stretched-coordinate PMLs (complex cell sizes allowed)
Currently, only periodic boundary conditions are included.
PEC/PMC boundaries can be implemented by drawing PEC/PMC cells near the edges.
Bloch boundary conditions are not included but wouldn't be very hard to add.
3 years ago
The default solver `opencl_fdfd.cg_solver(...)` located in
7 years ago
implements the E-field wave operator directly (ie, as a list of OpenCL
instructions rather than a matrix). Additionally, there is a slower
3 years ago
(and slightly more versatile) solver in `` which attempts to solve
7 years ago
an arbitrary sparse matrix in compressed sparse row (CSR) format using
the same conjugate gradient method as the default solver. The CSR solver
is significantly slower, but can be very useful for testing alternative
formulations of the FDFD electromagnetic wave equation.
7 years ago
Currently, this solver only uses a single GPU or other OpenCL accelerator;
generalization to multiple GPUs should be pretty straightforward
(ie, just copy over edge values during the matrix multiplication step).
## Installation
7 years ago
* python 3 (written and tested with 3.7)
7 years ago
* numpy
* pyopencl
* jinja2
* [meanas]( (>=0.5)
Install with pip, via git:
pip install git+
## Use
3 years ago
See the documentation for `opencl_fdfd.cg_solver(...)`
7 years ago
(located in ``````) for details about how to call the solver.
7 years ago
The FDFD arguments are identical to those in
3 years ago
`meanas.solvers.generic(...)`, and a few solver-specific
7 years ago
arguments are available.
2 years ago
An alternate (slower) FDFD solver and a general gpu-based sparse matrix
3 years ago
solver is available in ``. These aren't particularly
7 years ago
well-optimized, and something like
[MAGMA]( would probably be a
better choice if you absolutely need to solve arbitrary sparse matrices
7 years ago
and can tolerate writing and compiling C/C++ code. Still, they're
usually quite a bit faster than the scipy.linalg solvers.