[Cryptography] let's kill md5sum!

[Ryan Carboni]

>
> The version of Tiger benchmarked onhttp://bench.cr.yp.to/results-hash.html is Wei Dai's implementation
> from Crypto++ (https://github.com/mmoss/cryptopp/blob/5a55f26b5e07e5effdbefe00df3a0588a11bfbf0/src/tiger.cpp).
> It has optimized assembly and SSE2. I don't know for sure if there
> could be further optimizations to Tiger, but I would be surprised if
> anyone could squeeze better than 10% more speed out of it.
>
>
64-bit multiplication with 32-bit integers requires about four
instructions... I think?

That has been reduced to one instruction in modern architectures, a savings
of about three instructions per round, 24 rounds, so 72 instructions? There
is one multiplication each round in Tiger, I think.


BLAKE (the immediate ancestor of BLAKE2) came out in 2008, and during
> the SHA-3 competition BLAKE was probably subjected to more
> cryptanalysis than Tiger has been in its entire 20-year life.
>
>
You may not be aware of this in economics, but there's a loss in efficiency
when you increase inputs per unit of time.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.metzdowd.com/pipermail/cryptography/attachments/20150608/1250719a/attachment.html>