kernel_config#

MCEq.config.kernel_config = 'numpy'#

euler kernel implementation (numpy/MKL/CUDA/accelerate). With serious nVidia GPUs CUDA a few times faster than MKL autodetection of fastest kernel below