K-means clustering is an unsupervised classification algorithm used to partition a feature space into clusters. MATLAB has been utilised to vary the input arguments and evaluate the performance of the serial and concurrent variations. The serial implementation has been developed on an intel 4770 quadcore 3.4GHz processor in C++. The parallel implementation up to v12 has been developed on a Nvidia GTX 680 with Microsoft Visual Studios 2012. v13 is optimised for a Tesla C2075 (the target platform) with the Nvidia Nsight Eclipse plug-in.