Closed CPerezz closed 5 years ago
As discussed in private with @autholykos who already implemented modular sqrt
algorithms on a BN256 lib.
Cipolla is faster if p is 3%4
Atkins is faster if p is 5%8
For other cases Tonnelli-Shanks
performs better which is our case.
So I'll close the issue.
This issue lives under the item: https://gitlab.dusk.network/dusk-org/tech/issues/2
This means that
Cipolla's
algorithm should be faster thanTonelli-Shanks
, so we should implement it and benchmark both to get the most performant one.