1996-09-26 - Re: Mitsubishi MISTY LSI

Header Data

From: Adam Shostack <adam@homeport.org>
To: jya@pipeline.com (John Young)
Message Hash: 595580bd310befb4f1ed34a27159420a4151cb73e333ab7c918f3172be9352ae
Message ID: <199609261246.HAA03620@homeport.org>
Reply To: <199609252037.UAA11310@pipe2.ny3.usa.pipeline.com>
UTC Datetime: 1996-09-26 15:00:14 UTC
Raw Date: Thu, 26 Sep 1996 23:00:14 +0800

Raw message

From: Adam Shostack <adam@homeport.org>
Date: Thu, 26 Sep 1996 23:00:14 +0800
To: jya@pipeline.com (John Young)
Subject: Re: Mitsubishi MISTY LSI
In-Reply-To: <199609252037.UAA11310@pipe2.ny3.usa.pipeline.com>
Message-ID: <199609261246.HAA03620@homeport.org>
MIME-Version: 1.0
Content-Type: text


John Young wrote:
|    Mitsubishi Electric Corp. said it has developed an LSI 
|    that can encrypt data at a speed of 450 megabits per 
|    second, which is four times faster than any other 
|    encryption chip developed in Japan and brings domestic 
|    technology in line with DES, the U.S.-developed 
|    encryption system that has become the standard in the 
|    U.S. and Europe. 

	Its worth noting that even if the chip is as fast as they
claim (quite possible), and if its highly key agile (less probable,
but not unlikely), it may not be feasable to build devices that work
at anything near that speed, due to bus issues, i/o barriers, etc.

	Also, in response to Bill Frantz's snake oil questions, we
need to wait a few years before using it.  Unfortunately, in the real
world, there are people installing point to point ATM links, who need
this speed today, and might choose to install as evesdropping
protection.  Weak crypto, acknowledged and understood as such, can be
better than plaintext.

	Strong crypto is obviously better, and I'll happily take
pointers to vendors of 100mb ethernet, fddi or >=155mb atm crypto
hardware thats available today.

Adam

-- 
"It is seldom that liberty of any kind is lost all at once."
					               -Hume






Thread