Hey, I was wondering if there are any neon optimizations that are coming down the pipeline that are relevant for mpz_powm mpz_powm_sec. I've done some benchmarking with the current trunk, but armcortexa7 and armcortexa7neon seem to perform identically. Rob.