side-channel analysis, parallel computing, CUDA, pyecsca, GPU computing, performance optimization, Numba