> > Without detail knowledge of the current implementation, these numbers > > suggest to me that we could speed small operand gcd and gcdext. But I > > might be wrong.
not clear to me. For one word basically you can perform a multiply in a few cycles, whereas for a gcd you may need up to mp_bits_per_limb operations. If you find an algorithm that computes a gcd of two words in a few cycles, it would be a major improvement! Paul _______________________________________________ gmp-devel mailing list gmp-devel@gmplib.org https://gmplib.org/mailman/listinfo/gmp-devel
