Carleton-SCSI-2019
Some of the Code from my time at the Summer Computer Science Institute at Carleton
Research:
This code has to do with my research project on Parallel and Distributed Systems. Here we test how adding more processors would effect 3 different problems (using the MPI Framework), these problems runtime where then compared between two different systems.
1. Parallel Conway's Game of Life .
- Built in C, it splits the board into segments for each worker to process.
- Information is sent to and from the head node at the start and end of each generation.
- Works by checking the cells around it, updating its own state if enough are alive.
.
- We found that the processing time for each generation was very small.
- We found that the amount of time spent communicating became the bottle neck.
- Adding more processors made the game compute much slower.
2. Parallel Water Path Finding .
- Built in C, it splits the board into segments for each worker to process.
- Information is sent to and from the head node at the start and end of each generation.
- Works by checking if the end position has already been reached, if not the water will expand
.
- We found that the processing time for each generation was very small.
- We found that the amount of time spent communicating became the bottle neck.
- Adding more processors made the code compute much slower.
2. Parallel Intigration.
- Built in C, it splits the problem at the start into segments.
- Information is only sent once, and received once.
.
- We found that the processing time was the major bottle neck.
- Communication time was fast.
- Adding more processors made the code compute much faster.
Natural Language Processing .
- What I did week 2 of the program! (week 3 was computer music... I forgot to save the code).
- Used python to do data abstraction on different text files.
- Also created an AI to generate text.