Re: Distributed DES crack

["Perry E. Metzger" <perry@piermont.com>]

Matt Blaze writes:
> Here are my back-of-the-calculator numbers:
[...]
> MAYBE, somehow, you could do 100000 ECB/sec on "average" workstation
> (average = 100mhz Pentium).
> 
> That's 11000 Pentium-100 years for half the DES keyspace.  

Hmmm...

Lets assume 20,000 P100 Years to give a bit more breathing
room. 100,000 machines would be needed to get the thing into striking
distance. I think that is potentially doable. Hard, but
doable. Managing to avoid search failure (that is, having someone find
the key but somehow fail to report back) is the biggest problem, I
think.

> Well, I'm working on getting the funds to build (or support someone
> to build) some kind of parallel DES engine.  I can probably scrape
> together an FPGA-based machine that can do a key in less than 6 months.
> I'm very serious about this project, but I can't say for sure when or if
> I'll be ready to start.

If you can manage to do that, then I'd say that the software only
approach could be abandoned. Meanwhile, I think its time to try to
build those DES cracking screensavers for Windows...

Perry