Describe the example
The MATLAB doc says in this example that it you would give a 10x performance improvement if data is cast using "gpuArray-single", which uses MATLAB Parallel Computing Toolbox.
Is "gpuArray-single" supported and offers any improvement in this Python port?
option_factory() (from simulation_options) says that "gpuArray-single" is a valid data_cast, but simulation fails because post_init() (also from simulation_options) doesn't accept "gpuArray-single" as a valid parameter.
If it is supported, it would be nice to modify the example to show how.
Sources
Source: this example
Target: this example
Additional context
None. Just thanks for your hard-work!
Describe the example
The MATLAB doc says in this example that it you would give a 10x performance improvement if data is cast using "gpuArray-single", which uses MATLAB Parallel Computing Toolbox.
Is "gpuArray-single" supported and offers any improvement in this Python port?
option_factory() (from simulation_options) says that "gpuArray-single" is a valid data_cast, but simulation fails because post_init() (also from simulation_options) doesn't accept "gpuArray-single" as a valid parameter.
If it is supported, it would be nice to modify the example to show how.
Sources
Source: this example
Target: this example
Additional context
None. Just thanks for your hard-work!