1998-03-05 - Brickell and Sandia

Header Data

From: bill payne <billp@nmol.com>
To: j orlin grabbe <kalliste@aci.net>
Message Hash: df624b5ab0fda75bd6bee08d1e7024e2e5e84de627234e489f1d8557cef06f98
Message ID: <34FE12B4.6021@nmol.com>
Reply To: N/A
UTC Datetime: 1998-03-05 02:51:16 UTC
Raw Date: Wed, 4 Mar 1998 18:51:16 -0800 (PST)

Raw message

From: bill payne <billp@nmol.com>
Date: Wed, 4 Mar 1998 18:51:16 -0800 (PST)
To: j orlin grabbe <kalliste@aci.net>
Subject: Brickell and Sandia
Message-ID: <34FE12B4.6021@nmol.com>
MIME-Version: 1.0
Content-Type: text/plain


Wednesday 3/4/98 6:59 PM

Orlin

While trying to recover from the stomach flu, I am going over your stuff
summarized at 

  http://www.aci.net/kalliste/dcguide.htm

You wrote at [Brickell-Gemmell-Kravitz.]
http://www.aci.net/kalliste/sandia.htm

  Sandia National Laboratories have created the digital cash equivalent
of the Clipper chip: an "anonymous"     
  digital cash system that would give participants privacy from all
viewers, except for the government agencies 
  that would control the secret keys required for backdoor access.

and  

  Why is Sandia interested in digital cash systems? Well, Sandia is
responsible for all non-nuclear components    
  of nuclear weapons.  The security of nuclear weapons depends partly on
cryptology. The code- breaking 
  National Security Agency (NSA), for example,  is responsible for the
communication security of the 
  Minuteman missile, as well as the codes by which the President must
identify himself to authorize a nuclear   
  strike.

Brickell and Simmons were in the COMPUTATIONAL/COMPUTER SCIENCES & MATH
CENTER,
directorate 1400, at Sandia.

I worked in the ELECTRONIC SUBSYSTEMS CENTER, directorate 2300, 
division 2311, when I was project leader of the Missile Secure
Cryptographic Unit, the small missile,  [between about 1982-86].

All of the nuclear bomb crypto implmentation work was done in 2300, not
1400.

The MAIN difference is that the 2300 people were a bunch of PRACTICAL
engineers and REAL-PRACTICAL software types.  NOT theoreticians.

Gus Simmon once tried to get into the implentation business.  

Simmons bought an Intel 320 [?] development system.  The Intel 320 was a
piece of junk and nothing happened with Simmons' implementation work.

Simmons' try, naturally, caught the attention of 2300 management, for
business reasons, of course.

I LOVE reading all of this stuff from a Sandia historical standpoint.

The REAL WORLD of Sandia crypto stuff in pockmarked by some REAL
SCREW-UPS.  Which all
of us implementers shared so as not to commit the same mistakes again.

I asked my department manager, Kent Parsons, how the screw-ups affected
him.  He responded that
it made him sleepy.  These were MULTI- if not HUNDREDS- million dollar
screw-ups.

Later
bill







Thread