[1170] in cryptography@c2.net mail archive
Re: Better DES challenge update
daemon@ATHENA.MIT.EDU (Peter Trei)
Mon Jul 7 11:36:58 1997
From: "Peter Trei" <trei@process.com>
To: Andreas Bogk <andreas@artcom.de>, die@die.com, eli@gs160.sp.cs.cmu.edu,
cryptography@c2.net, crisp@netcom.com, smb@research.att.com
Date: Mon, 7 Jul 1997 10:01:40 -6
Reply-to: trei@process.com
CC: trei@c2.net
Andreas Bogk <andreas@artcom.de>
> >>>>> "Steven" == Steven Bellovin <smb@research.att.com> writes:
>
> Steven> Let me point to two papers:
>
> Steven> ftp://ftp.research.att.com/dist/smb/recog.ps
>
> This is very interesting, but unfortunately there's not enough RAM on
> a typical FPGA to implement digraph statistics. On the other hand,
> there's plenty of RAM for single character statistics.
>
> The Wiener design hardly fits on todays FPGAs, the S-boxes already eat
> up 80% of the logic functions of an Altera 10K100. It's hard to tell
> wheter the rest of the logic fits as well.
> Andreas
People interested in doing this should look at the work that has been
done on 'bitslice' DES implementations in software. The gate count
for doing the S-boxes can be radically reduced from the
run-of-the-mill, 2 level AND/OR structure.
Peter Trei
trei@process.com
