B-ASIC’s documentation¶
B-ASIC is a toolbox for Python 3 that simplifies implementing application-specific circuits, primarily aimed at signal processing algorithms for both standard-cell and FPGA technology.
It was initially developed by a group of students to replace an existing Matlab toolbox in a course commonly referred to as “the ASIC course” at Division of Computer Engineering, Linköping University, Sweden. Hence, the name “Better ASIC toolbox” = B-ASIC.
It was initiated and is currently maintained by Oscar Gustafsson.
The development of B-ASIC happens at https://gitlab.liu.se/da/B-ASIC.
It is not yet fully functional, but several parts of the design flow works, while others are missing/buggy. The goal is to have a working design path from algorithm down to a HDL-description of a custom architecture. Once it becomes a bit more mature, we expect to make it available on pypi and conda-forge so that it will becomes easier to access.
To install B-ASIC, the currently preferred way is:
git clone https://gitlab.liu.se/da/B-ASIC.git
cd B-ASIC
python -m pip install -e .
This will install in editable mode, which is beneficial since you then easily can pull new changes without having to reinstall it. It also makes it easy to contribute any improvements.
In addition to the dependencies that are automatically installed, you will also need a Qt-binding, but you are free to choose from the available Qt5 and Qt6 bindings. See https://gitlab.liu.se/da/B-ASIC for more info.
If you use B-ASIC in a publication, please acknowledge it. Later on there will be a citation provided, but right now, please refer to this documentation or the repository. We will also maintain a list of publications that have used B-ASIC.