<redsh4de:matrix.org> There is a new isogeny scheme that claimsto be several orders of magnitude faster and more resistant to PQ attacks than CSIDH, with slightly bigger 80 byte public keys. Should be relevant for the Jamtis-PQ discussion: eprint.iacr.org/2026/2377.pdf

<redsh4de:matrix.org> "For NIST level~1, MIKE has very compact public keys of 80B, and using our constant time C implementation, the key generation takes 0.65ms and the shared secret takes 4.9ms (including key validation) on a AMD CPU at 3.3GHz."

This looks like a major advancement, promising!

<redsh4de:matrix.org> Exponential quantum attack complexity due to being immune to Kuperberg's algorithm, which CSIDH is vulnerable to 👀

<redsh4de:matrix.org> Really good reading material

<syntheticbird> eprint.iacr.org/2026/2377.pdf

<syntheticbird> thats one chonky paper holy shit

<redsh4de:matrix.org> they do acknowledge the that fact on page 7 hahahh

<redsh4de:matrix.org> while reassuring that the Rust impl is only 3000 loc

<syntheticbird> > To the humans reading this paper. We are aware that this paper is probably

<syntheticbird> > longer than it should be. Let us reassure you by the fact that our implementation

<syntheticbird> > is much shorter than this paper: only around 3000 lines of rust code (not counting

<syntheticbird> > tests nor the finite-field arithmetic implementatio

<syntheticbird> nice

<spirobel:kernal.eu> they didnt cite townforge as prior work ... > <@ircmouse:matrix.org> Not sure if it’s been posted already: ia.cr/2026/1205

<spirobel:kernal.eu> tbh it seems like a paper specific to zano