matkicail / hpc-project

A project that involved C programming. Where we made use of parallel implementations of OpenMPI and Cuda. We compared the performance of these two and attempted to perform K Means and Fuzzy C Means clustering on synthetic data.

HPC-Project

HPC-Project

The project's code segments are fairly well commented. This is largely only intended to be used as a reference. The focus of this project was to showcase the performance differnces in Fuzzy C-Means clustering and K-Means clustering which made use of OpenMPI and CUDA in each implementation. Thus, comparing the benefits of a chosen framework. It should be noted, this made use of global memory in the case of CUDA which is sub-optimal for this problem but served the purpose of a basic CUDA implementation.