Re: New ECC Paper (fast GF(p) point mul)


Tom St Denis wrote:
Definitely nice to see advances in the GF(p) world. Unfortunately it
doesn't [???] apply to NIST curves as the curve he uses is of the form

Scratch that. The author sent me more info ... hehehe.

Turns out you can map from the NIST style to Montgomery style if two
things are true

1. There is at least one root to the equation x^3 - 3x + b in Fp [call
it \alpha]
2. The root plugged into 3\alpha^2 + a is a QR in Fp

Now I know how to find roots over Z and R [Newton comes in handy] but
how do you do it over Fp? Factoring it gives us

x(x^2 - 3) = -b

Any refreshers? hehehe

Tom

.