The Fast Fourier Transform (FFT) is one of the most widely used algorithms in engineering and scientific applications and therefore its analysis and performance is of much importance to a range of research fields. On PRACE Tier-0 systems a parallel environment with a great deal of processing power (large number of CPU cores or accelerators such as GPUs) is at disposal for researchers. The FFTs investigated are both in-code and through various numerical libraries, where the algorithm is implemented in both serial and parallel form. The implementations of FFT investigated range from pure MPI, OpenMP versions for multicore, hybrid (OpenMP/MPI) to GPU-based. The objective of this project is to assess the suitability, performance and scalability of various implementations of FFT for the PRACE large-scale scientific applications Quantum ESPRESSO and DL_POLY.