CollatzConjecture

This project is an exhaustive check of numbers against the Collatz Conjecture (explained here).

Although the usefulness of this project is very questionable, it's become a great source of coding practice for myself.

Requirements

• CMake
• Boost
• CUDA (only if building/running the CUDA executable)

Building

This project uses CMake, so build using this format after cloning:

mkdir build
cd build
cmake ..
make

By default, this will build the C and C++ versions of the code. To enable build of the CUDA executable, turn on the Collatz_CUDA flag using either ccmake . or cmake .. -DCollatz_CUDA=ON.

Running

C++

To run the C++ version, run the built CollatzConjectureCPP executable. This executable can take the following options:

  -h [ --help ]         Display this message
  -n [ --numproc ] arg  Number of cpu threads to use
  -s [ --server ] arg   Start a CollatzServer on this machine
  -c [ --client ] arg   Point this machine as a client to the given server

Server/Client Operation

To start an executable in the serving mode, run with the option -s <port>, where port is the number of the port you want to serve from. To connect a client to this server, run with the option -c <ip:port>, where ip is the address of the machine running the server and port is the port number the server was told to serve from.

C

The C executable is not built with client/server capabilities, even though it claims it is. Run it with the option --server <port> to get it to run, but it will not serve anything. This executable, being the original, is just called CollatzConjecture.

CUDA

The CUDA version of the conjecture is also a standalone application, and acts like it. Run it by executing the CollatzConjectureCUDA executable.

Documentation

The C++ code has adopted a Doxygen-compatible documentation standard, so documentation can be generated by just running Doxygen in the checkout directory.