cuda_fp_precision#

MCEq.config.cuda_fp_precision = 64#

CUDA Floating point precision (default 32-bit ‘float’)