Othello Solver

NUS CS3211 Project: Parallel solver for Othello / Reversi using OpenMPI

This solver evaluates the best move for a current board position and for a specified player using either the minimax algorithm or minimax with alpha-beta pruning algorithm. The algorithm, the parallelization method (batch or job pooling) and the other parameters that the solver uses can be customised in the src/othello.cpp file.

Instructions

1. Download OpenMPI (https://www.open-mpi.org/)
2. Run make to build the executable othellox
3. Run make run or
   mpirun -np <NUMBER OF PROCESSORS TO USE> ./bin/othellox config/initialbrd.txt config/evalparams.txt

Configurations

1. Initial Board Position (config/initialbrd.txt)

• Specify the size of the board
• Specify the current positions of the White and Black disks
• Specify the current Player (Black or White)
• Specify the timeout for the evaluation of each board (Not implemented)

2. Evaluation Parameters (config/evalparams.txt)

• Specify the maximum depth of evaluation of the board (MaxDepth)
• Specify the maximum number of boards to evaluate (MaxBoards) (Not implemented)
• Specify the value given to a corner as heuristic to evaluate the board (CornerValue)
• Specify the value given to an edge as a heuristic to evaluate the board (EdgeValue)

3. Algorithm Details (src/othello.cpp)

• Specify the algorithm to use (Serial or Parallel Batch or Parallel Job Pool, Minimax or Alpha-beta Pruning)
• Specify the method of job distribution (Random or Sequential)
• Specify the number of jobs that each processor should work on
• Specify the number of jobs to send each processor if algorithm is Job Pooling