In this work, the two most popular interference cancellation schemes, PIC and SIC, are implemented and compared in a graphics processing unit (GPU) using CUDA to speed up the most time-consuming process of a NOMA receiver. Numerical results show that PIC, due to its parallel architecture, is...