phc-discussions - bcrypt in hardware (Re: [PHC] A review per day

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date: Sun, 31 Aug 2014 10:27:49 +0400
From: Solar Designer <solar@...nwall.com>
To: discussions@...sword-hashing.net
Subject: bcrypt in hardware (Re: [PHC] A review per day - TwoCats)

Bill,

On bcrypt:

On Sat, Aug 30, 2014 at 08:50:38AM +0400, Solar Designer wrote:
> On Fri, Aug 29, 2014 at 03:12:47PM -0400, Bill Cox wrote:
> > I think a round should be doable in a clock cycle
> 
> Yes.
> 
> > without any memory accesses using an async SRAM,
> 
> You're the expert here.  I can only say that a round is doable in a
> single clock cycle with memory accesses, as has been implemented in
> FPGAs already.  The addresses may be computed at the beginning of this
> cycle, or at the end of the previous cycle.  Yes, combining both
> computation and memory lookups on the same cycle limits clock rate.
> Some 2-cycle designs might be faster, if 2x more memory is available.

I finally got around to placing the slides from Katja Malvoni's latest
talks online:

http://www.openwall.com/presentations/Passwords14-Energy-Efficient-Cracking/

This also includes a link to our WOOT '14 paper.  1 cycle/round bcrypt
designs are mentioned on the slides and in the paper.

As we learned in July, another team worked on bcrypt cracking on
ZedBoard as well.  They have a 1 cycle/round design too.  Curiously,
Ralf Zimmermann from that other team is also co-author of AntCrypt, a
PHC candidate.  It's a small world.

Alexander