README.md

GPU Pothos Blocks

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.

Documentation

• https://github.com/pothosware/PothosGPU/wiki (TODO)

Dependencies

• Pothos library (0.7+)
• ArrayFire (3.6+)
• Mako Python module (build-time only)
• YAML Python module (build-time only)

Licensing information

This module is licensed under the BSD 3-Clause license. To view the full license, view LICENSE.txt.