Post-Quantum signatures zoo

[Thom Wiggers]

Hi all,

I have taken all the public key and signature sizes, as well as all reported performance metrics, from the submission documents and put them in a bunch of tables. Then, I wrote a bunch of JavaScript (poorly) to allow one to sort, filter and compare these schemes to heart's content.

Find our comparison tool at https://pqshield.github.io/nist-sigs-zoo/

Note that the scatter plot for key and signature sizes will reflect the selected filtering options, so it can be useful to get a birds' eye view of the schemes that are relevant to your use case.

If anyone has suggestions for improvements or wants to write some JavaScript themselves, we welcome comments and contributions here or on GitHub.

I hope that this is a helpful and fun way to drill down into the long list of submissions.

Cheers,

Thom

[Anjan Roy]

Hi Thom and all,

This is actually very useful tool for comparing among all these signatures schemes. Thanks for putting this together.

Regards,

Anjan 

--
You received this message because you are subscribed to the Google Groups "pqc-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pqc-forum+...@list.nist.gov.
To view this discussion on the web visit https://groups.google.com/a/list.nist.gov/d/msgid/pqc-forum/CABzBS7kPpmDdFaLxdguyj3z-dgTL%2B-pQi2aSnR9e_ZY9mEJCiA%40mail.gmail.com.

[Gustavo Banegas]

Hi Thom,

It seems that the filters are not exactly filtering correctly. It might
be that my browser is blocking some JS. Also, the graph does not grow
for bigger parameters in certain schemes.

All the best,

Gustavo

> <https://groups.google.com/a/list.nist.gov/d/msgid/pqc-forum/CABzBS7kPpmDdFaLxdguyj3z-dgTL%2B-pQi2aSnR9e_ZY9mEJCiA%40mail.gmail.com?utm_medium=email&utm_source=footer>.

>
> --
> You received this message because you are subscribed to the Google
> Groups "pqc-forum" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pqc-forum+...@list.nist.gov.
> To view this discussion on the web visit

> https://groups.google.com/a/list.nist.gov/d/msgid/pqc-forum/CANhqc%2BNQXarYgE41pjohZ%3DF2Xfi0swNEvbnM2YH3Hx79YjjMdQ%40mail.gmail.com
> <https://groups.google.com/a/list.nist.gov/d/msgid/pqc-forum/CANhqc%2BNQXarYgE41pjohZ%3DF2Xfi0swNEvbnM2YH3Hx79YjjMdQ%40mail.gmail.com?utm_medium=email&utm_source=footer>.

[Thom Wiggers]

Hi all,

I have just released an all-in-one table version. It is best enjoyed on the widest screen you own, but might be a bit easier to drill down into the relation between key/sig sizes and cycles.

https://pqshield.github.io/nist-sigs-zoo/wide.html

Cheers,

Thom

Op wo 19 jul 2023 om 09:44 schreef Thom Wiggers <th...@thomwiggers.nl>:

[D. J. Bernstein]

Thom Wiggers writes:
> https://pqshield.github.io/nist-sigs-zoo/wide.html

In the graph, a factor 10 is given less visual space for sigbytes than
for pkbytes (at least on my screen); I'd suggest equalizing the scales.
It might also be helpful to use different sizes for the plus marks to
indicate claimed security levels.

---D. J. Bernstein

[Dan Collins]

Thanks for making this! I'd also suggest trying to make the plot use more of the available screen width. Currently it appears to max out at 1000 px, giving the plot `width: 100%;` (rather than `max-width: 100%;`) improves the clarity, though the plot is still constrained by the `max-width: 75rem;` on the main grid-container div.

--

Dan Collins

--
You received this message because you are subscribed to the Google Groups "pqc-forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pqc-forum+...@list.nist.gov.

To view this discussion on the web visit https://groups.google.com/a/list.nist.gov/d/msgid/pqc-forum/20230720165018.601307.qmail%40cr.yp.to.

[Bo Lin]

Very nice. Thanks..

From application point of view, considering the low communication bandwidth of small devices, any signature size or public key size greater than 1000 bytes are hardly to see any realistic use.