Re: Optimized key rates?

Darrell Kindred (dkindred@cmu.edu)
Sat, 3 May 1997 23:51:15 -0400


Corbett J. Klempay writes:
> [...]
> >I'm not doing the code, so I can't tell you for sure, but everyone
> >working on it is aware of the Biham-type optimizations for 64-bit
> >processors, which should provide still more performance. They should
> >kick butt.
> >
>
> Cool! Can someone provide some quick info on the Biham bit-slicing
> approach to me, or at least tell me where to go? Will Biham's method only
> work for 64-bit processors (or will it work on 32-bit, but with greatly
> reduced efficiency?)? Thanks.

I'm the person working on a bit-sliced DESCHALL client.
Given the amount of apparent interest in fast 64-bit
clients, I think my time would be better spent right now
finishing it up than explaining Biham's method, but I
promise I'll post an explanation once the client is out the
door.

Just a quick answer to your last question: the method does
work with 32-bit processors, but roughly half as fast as it
does on an "equivalent" 64-bit processor. At the moment it
looks like the current DESCHALL clients will do better on
32-bit machines.

Back to work!

- Darrell