A program that uses wires to connect gates on a chip three-dimensionally and efficiently. Based on the case of Chips & Circuits.
Chips (or more precisely: integrated circuits) are found in your PC, MacBook, Android Phone and microwave oven where they perform a diversity of functions, ranging from timekeeping and motor control to arithmetic and logic. Basically a small plate of silicon, chips are usually designed logically and subsequentially transformed to a list of connectable gates. This list, commonly known as a net list is finally transformed into a 2-dimensional design on a silicon base.
This last step however, the physical real-world process of connecting the gates, is highly volatile. Good arrangements on the base lead to short connections, leading to faster circuits, whereas poor arrangements lead to slower circuits. It leads to no doubt that a good arrangement of logical gates and good wiring between them is of vital essence to the performance of the IC as a whole (source: http://heuristieken.nl/wiki/index.php?title=Chips_%26_Circuits).
The algorithms and content of the sample has been setup to create paths from predefined net lists and two boards that can be found in sample/data. So in short, the aim is to connect the dots (gates) in a efficient manner.
Board: The board a three dimensional grid with a presetted height, weight and depth.
Gate: A gate is placed on the surface of the board and can be connected to another gate.
Connection: Two gates that have to be linked, whereas gate A is the starting point and gate B is the end point.
Path: A path from gate to gate, walking over the board. It cannot move diagonally only in one of six directions: north, south, west, east, up and down.
Net list: A list of connections
Score of a board: The total sum of the lengths of all connected paths.
Result of a board: The percentage of made connections from the netlist.
Solvability of a board: The average calculated percentage of a board with randomly placed gates and random netlists.
The documentations of the classes and helper functions used in sample. The documentations are made by Sphinx, a docstring converter.
One experiment is documented at the moment. We used a random netlist creator and random board creator to see how the used algorithms preformed.
Sample is the core of our repository. _main._.py is the core of the core. It makes use of classes.py and helpers.py to drive the algorithms to solve the case. Additions are colors.py for adding terminal colors and settings.py. settings.py can be used to modify the program but more on that in de README.md in the sample directory.
Clone or download this repository (containing the pathfinder module) to a folder of your preference. Make sure the requirements above are installed to be able to run the program.
To run this program, you need the following:
- Python 3.6.3
- numpy 1.13.3
- matplotlib 2.1.0
We use SemVer for semantic versioning. For the versions available, see the tags on this repository.
- Jurre Brandsen - Developer - www.jurrebrandsen.nl
- Lennart Klein - Webdeveloper and webdesigner - www.lennartklein.nl
- Thomas de Lange - Developer - www.long-coding.nl
This project is not open for contributing at the moment, for it being a school assignment.
This project is licensed under the MIT License - see the LICENSE file for details
- Thanks to Daan van den Berg for creating the "Chips and circuits" case (as specified on this Wiki).