[5086] in cryptography@c2.net mail archive
Re: World's fastest hardware encryption
daemon@ATHENA.MIT.EDU (David Honig)
Fri Jul 9 01:10:24 1999
Date: Thu, 08 Jul 1999 17:58:47 -0700
To: koontz@no_spam_ariolimax.com, Udhay Shankar N <udhay@pobox.com>
From: David Honig <honig@sprynet.com>
Cc: cryptography@c2.net
In-Reply-To: <37842B1B.98109A87@ariolimax.com>
At 09:37 PM 7/7/99 -0700, David G. Koontz wrote:
>Udhay Shankar N wrote:
>>
>> found on slashdot, where it was headlined "The first step to cypherspace
You should subscribe to cpunks, where this has been mentioned.
>> http://www.sandia.gov/media/NewsRel/NR1999/encrypt.htm
>Should be able to do static key distribution accross the device, and
>11 gate levels per round. That gives a number around 880 ps per gate
>with routing. Should be able to go 3 or 4 times faster. At least
>twice as fast with routing distances.
>
>You could also do a dual pipline...
Your numbers are correct.
You could put about a dozen fully-pipelined
DES engines on an ASIC and get reasonable
(e.g., commercial) yields. A 1-hour brute-force
engine is the size of a laptop. Not counting SiGe.
