2021.01.31 18:40:10 (1355933909788225536) from Daniel J. Bernstein:
Our inversion code is now generalized to handle common 256-bit primes at almost the same speed as 2^255-19, around 4000 Skylake cycles: https://gcd.cr.yp.to/software.html Still haven't fully verified, so don't put into production. Tracking verification progress: https://gcd.cr.yp.to/verif.html