You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Field inverse is computing 1/x mod p. This is currently used for FFT, hash-to-curve, final exponentiation (easy part), coordinates conversion and Karabina's cyclotomic square decompression (final exp. hard part). This can also be possibly needed for BW6 Miller loop in 2-NAF (see here) and BatchAffineAddition (see #72). Particularly, for BatchAffineAddition and Karabina (see #75), the faster the field inversion is the better the the performance is.
Currently, gnark-crypto implements Algorithm 16 in [GKPP06]. Implementing this work [Pronin20] in pure Go should yield better performances.
The text was updated successfully, but these errors were encountered:
Field inverse is computing
1/x mod p
. This is currently used for FFT, hash-to-curve, final exponentiation (easy part), coordinates conversion and Karabina's cyclotomic square decompression (final exp. hard part). This can also be possibly needed for BW6 Miller loop in 2-NAF (see here) andBatchAffineAddition
(see #72). Particularly, forBatchAffineAddition
and Karabina (see #75), the faster the field inversion is the better the the performance is.Currently,
gnark-crypto
implements Algorithm 16 in [GKPP06]. Implementing this work [Pronin20] in pure Go should yield better performances.The text was updated successfully, but these errors were encountered: