[2184] in cryptography@c2.net mail archive
Re: The secret message is...
daemon@ATHENA.MIT.EDU (Stephen Cobb, CISSP)
Wed Feb 25 18:38:24 1998
Date: Wed, 25 Feb 1998 17:57:56 -0500
To: Amanda Walker <amanda@intercon.com>
From: "Stephen Cobb, CISSP" <stephen@iu.net>
Cc: cryptography@c2.net
In-Reply-To: <199802252200.RAA03235@mail.intercon.com>
At 05:00 PM 2/25/98 -0500, you wrote:
>> Why is that interesting? Macs are not particularly efficient for this. NSA
>> has its own chip foundary and remember [...]
>
>I find it interesting because it demonstrates that a private company
>can buy existing off-the-shelf computing hardware and use freely
>available software to build a 1-2 month DES key-cracker for only a few
>millions of dollars, requiring no crypto-specific hardware or software
>expertise.
>...
>Brute force key search on DES is no longer the monopoly of governments
>or even hackers. It's now simply a cash flow problem.
I agree. Obviously, NSA designing and building crypto-specific chips is a significant data point in some contexts. In other contexts, such as competitive commercial environments, the rapidly shifting economics of home-brewed, code-breaking hardware are acutely important.
Rack-mounting Pentium motherboards would deliver considerable economies of scale, and more direct, possibly less-expensive routes are entirely feasible. In 1985 I speculated that large quantities of Intel CPUs in a passive backplane arrangement could lead to affordable, private-sector, software-based key-cracking without custom crypto-chips.
The economics of commercial code-breaking include not only the dimensions of how cheap and how fast, but also how accessible. With results such as we have seen recently, the scale in all three dimensions is heading towards "very."
Stephen
<<<<<<<<<<<<<<<Stephen Cobb, CISSP>>>>>>>>>>>>>>>>>
Certified Information Systems Security Professional
tel: 1.407.383.0977 fx: 0336 email: scobb@miora.com
For information security info, http://www.miora.com
