This toolkit adds 100+ Pothos blocks that provide optimized implementations of various mathematical operations by offloading intensive processing onto a GPU, resulting in significant speed improvements.
Under the hood, the ArrayFire library uses a heavily optimized CUDA or OpenCL implementation, depending on what is available. If neither are available, it falls back onto a default CPU implementation.
This module is licensed under the BSD 3-Clause license. To view the full license, view LICENSE.txt.