NCipher Takes Hardware Security To Network Level

[Perry E. Metzger]

I was asked by the author of this to forward it with the sender
information removed.

From: [someone]
Subject: Re: NCipher Takes Hardware Security To Network Level

On Mon, Oct 06, 2003 at 01:38:13PM -0400, R. A. Hettinga wrote:
> 
> --- begin forwarded text
> 
> 
> Status:  U
> Date: Mon, 06 Oct 2003 12:40:41 -0400
> From: Somebody
> To: "R. A. Hettinga" <rah at shipwright.com>
> Subject: Re: NCipher Takes Hardware Security To Network Level
> 
> Don't identify me, since I'm not sure what parts of my NDA are still in 
> force now that they've announced it.
> 
> It's really pretty clever.  All the expensive key-management is moved 
> off to their centralized server.  As each low-cost HSM (the things that 
> go into your server) comes up, it sends its "card identity" to the 
> server.  The server responds with the necessary keys, sent in 3DES 
> (maybe AES?  I forget details).  Their cards can now be fairly simple 
> accelerators, and need less key protection, less NVRAM, etc.

That's not particularly clever at all -- it's one of the standard tricks
in your typical FIPS-140 consultant's bag thereof.

In fact, if you're clever, you can manage to not trouble yourself to get
the key-management, etc. certified, getting only the simple, symmetric-cipher
stuff run through the process.  The government will still buy your 
"encryption devices" (FIPS-140 certified) and will conveniently ignore the
lack of certification on your "management device", even though it acts as
an administrative user towards the "encryption device".  It's somewhat
scary that this sort of skulduggery is possible, but it's also not really
anything new or exciting.

---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo at metzdowd.com