Scuby
Signal-Cuda-Blocks-Py. SimuLink (Signals and Blocks) in Python with CUDA support.
InputBuffer
are rounded hexogons). Color
reflect the order of updating from the earliest (dark blue) to the latest (dark red).
Arrows indicate output-to-input routing between components, label around the
beginning/end of arrows indicate the name of the output/input variables respectively.
SCuBy is a CuPy-accelearted Python Library for designing and simulating network of processors (Filters, Operators) with emphasis on temporal processing. It is designed to support definition of network components (kernels) and processing that are tightly integrated with mathematical constructs like Hilbert Spaces.
SCuBy is heavily influenced by Neural. The high level API for circuit specification in SCuBy is designed to be identical as that in Neural, with the intention of merging the libraries to support unified scientific computation across these two levels of abstractions.
Note, As SCuBy contains code used for unpublished projects, it remains a private repository on GitHub as of March 2022. It will be made publicly available as soon as the projects are published. The documentation for SCuBy is publicly accessible.