1993-09-12 - Re: `supercomputer’ export control

Header Data

From: gnu (John Gilmore)
To: “L. Detweiler” <ld231782@longs.lance.colostate.edu>
Message Hash: 939b997f15b15d5b6d8f6d8c80661884b76105756f4cdb97dadcbf7c49c031f5
Message ID: <9309120728.AA28062@toad.com>
Reply To: <9309120637.AA09563@longs.lance.colostate.edu>
UTC Datetime: 1993-09-12 07:28:37 UTC
Raw Date: Sun, 12 Sep 93 00:28:37 PDT

Raw message

From: gnu (John Gilmore)
Date: Sun, 12 Sep 93 00:28:37 PDT
To: "L. Detweiler" <ld231782@longs.lance.colostate.edu>
Subject: Re: `supercomputer' export control
In-Reply-To: <9309120637.AA09563@longs.lance.colostate.edu>
Message-ID: <9309120728.AA28062@toad.com>
MIME-Version: 1.0
Content-Type: text/plain


> >From Washington Post newswire   08/27

Note the date; this is old news.

>     U.S. Acts to Ease Export Controls On Computers; Industry Officials 
>     Say Proposed Standard Falls Far Short of Need 

They're right.

> 				    The new Commerce Decision rules allow
> export of microprocessors rated at 67 Mops (million operations per second),
> a big boost from the previous limit of 12 Mops.  However, multiprocessor
> units are still on the forbidden list.

67 NOPS [-) is a single SuperSPARC chip or a fast Pentium.  Nothing
faster than that can be exported without being considered a
"supercomputer", requiring armed guards around it in the foreign
country, etc.  Like, a SPARCstation-10 containing TWO SuperSPARC
chips...

The amusing thing is that the chips themselves are not export-controlled.
Foreign clone-makers buy the chips and sell to the rest of the world,
while U.S. companies are fried by endless red tape.  If you can get 
delivery next week or in six months, which supplier do you pick?

The whole concept of export controls on computers and communications gear
(including cryptography) has got to be demolished.  Smashed to the ground
like the Berlin Wall, a mere memory of decades of totalitarian bureacracy
that ruined real lives and real products.

	John





Thread