1994-08-24 - Re: Nuclear Weapons Material

Header Data

From: “Perry E. Metzger” <perry@imsi.com>
To: Phil Karn <karn@qualcomm.com>
Message Hash: ca7c7518f7353b13d730a9466b9f994703a6a56bde339111a845d2cf157ac6f3
Message ID: <9408241340.AA03320@snark.imsi.com>
Reply To: <199408240810.BAA27546@servo.qualcomm.com>
UTC Datetime: 1994-08-24 13:40:55 UTC
Raw Date: Wed, 24 Aug 94 06:40:55 PDT

Raw message

From: "Perry E. Metzger" <perry@imsi.com>
Date: Wed, 24 Aug 94 06:40:55 PDT
To: Phil Karn <karn@qualcomm.com>
Subject: Re: Nuclear Weapons Material
In-Reply-To: <199408240810.BAA27546@servo.qualcomm.com>
Message-ID: <9408241340.AA03320@snark.imsi.com>
MIME-Version: 1.0
Content-Type: text/plain



Phil Karn says:
> Now can we return to cryptography? How about a discussion of fast
> modular exponentiation algorithms, something we (or at least I) can
> put to more immediate and constructive use than nuclear bomb designs?

Indeed.

I've been wondering recently, by the way, about what advantages doing
some of this stuff on DSPs might have. DSPs are not magical chips, but
they are optimized for a few tasks, including, typically, fast integer
multiplies. IDEA and modular exponentiation both require lots of fast
integer multiplies. Would it make sense to use DSPs as co-processors
to things like Pentiums to speed up these processes?

Phil? You are the resident expert on DSPs, I believe...

Perry





Thread