"success"

	Out of curiosity... is there any way to safely embed our
encrypt/decrypt
	functions in that code without it becomming public?
Otherwise I'm not
	sure how we'd be able to plug one of these in to the
network...

The answer is still effectively no. There would have to be a
standard
hardware configuration to freeze it. Or at least a fixed set
of cores
for every usable FPGA, which at least in the beginning would
require
releasing the crypto to a half dozen or better FPGA
developers (most
of the FPGA RC5 user community).

	On Wed, Oct 18, 2006 at 12:55:32PM -0600, Olivier Meyer
wrote:
	> >I just thought I'd let you guys know since this
seems to be a 4 msgs a
	> >year mailing list 

Which is why this list is dead. The proprietary interests
block open
access.

John
_______________________________________________
Hardware mailing list
Hardwarelists.distributed.net
http://lists.distributed.net/mailman/listinfo/hardware

On Wed, Oct 18, 2006 at 01:51:43PM -0600, John L. Bass
wrote:
> 	Out of curiosity... is there any way to safely embed
our encrypt/decrypt
> 	functions in that code without it becomming public?
Otherwise I'm not
> 	sure how we'd be able to plug one of these in to the
network...
> 
> The answer is still effectively no. There would have to
be a standard
> hardware configuration to freeze it. Or at least a
fixed set of cores
> for every usable FPGA, which at least in the beginning
would require
> releasing the crypto to a half dozen or better FPGA
developers (most
> of the FPGA RC5 user community).

Well, anyone who's willing to sign an NDA can become 'staff'
and get
full CVS access, so maybe that's not beyond reason.

> 	On Wed, Oct 18, 2006 at 12:55:32PM -0600, Olivier
Meyer wrote:
> 	> >I just thought I'd let you guys know since
this seems to be a 4 msgs a
> 	> >year mailing list 
> 
> Which is why this list is dead. The proprietary
interests block open
> access.

Ideas for ways to do away with the crypto most welcome. 
-- 
Jim C. Nasby, Database Architect            decibeldistributed.net
Give your computer some brain candy! www.distributed.net
Team #1828

Windows: "Where do you want to go today?"
Linux: "Where do you want to go tomorrow?"
FreeBSD: "Are you guys coming, or what?"
_______________________________________________
Hardware mailing list
Hardwarelists.distributed.net
http://lists.distributed.net/mailman/listinfo/hardware

Well, how is the data verified and sent back? (conceptually,
I'm not 
looking for details)
There could always be some sort of software layer on a host
computer 
with any number of client FPGAs attached.

By the way, I made a simple top module and tried to
synthesize the code 
today. It started to work but I tried to re-run the
synthesis and 
Synplify Pro just sat there for hours running 100%. I'm not
sure what's 
going on with that - maybe it just needs a reboot. I don't
know how much 
space it will take up in an FPGA. It's a 155 stage 32 bit
pipeline which 
seems large by my standards. It is for this reason that I
don't know 
which device it will fit in (I use Xilinx S3-200 right now)
or how fast 
it will run.

I will give out the code once I figure out if there are any
reasons not 
to do so. It's basically the code I posted to this list 10
months ago 
converted to Verilog and printed out with tac. It's also
part of my 
graduate thesis.

IMO the reward for breaking 72 bit encryption should be
$100k
--
Martin K


John L. Bass wrote:
> 	Out of curiosity... is there any way to safely embed
our encrypt/decrypt
> 	functions in that code without it becomming public?
Otherwise I'm not
> 	sure how we'd be able to plug one of these in to the
network...
>
> The answer is still effectively no. There would have to
be a standard
> hardware configuration to freeze it. Or at least a
fixed set of cores
> for every usable FPGA, which at least in the beginning
would require
> releasing the crypto to a half dozen or better FPGA
developers (most
> of the FPGA RC5 user community).
>
> 	On Wed, Oct 18, 2006 at 12:55:32PM -0600, Olivier
Meyer wrote:
> 	> >I just thought I'd let you guys know since
this seems to be a 4 msgs a
> 	> >year mailing list 
>
> Which is why this list is dead. The proprietary
interests block open
> access.
>
> John
> _______________________________________________
> Hardware mailing list
> Hardwarelists.distributed.net
> http://lists.distributed.net/mailman/listinfo/hardware
>   
_______________________________________________
Hardware mailing list
Hardwarelists.distributed.net
http://lists.distributed.net/mailman/listinfo/hardware

On Wed, Oct 18, 2006 at 06:54:02PM -0400, Martin
Klingensmith wrote:
> Well, how is the data verified and sent back?
(conceptually, I'm not 
> looking for details)
> There could always be some sort of software layer on a
host computer 
> with any number of client FPGAs attached.
 
There's a residual that's computed as part of the normal
clients, and
the server checks that.

The issue with having a host machine do the encryption is
that the
decrypted data is then available for anyone to muck with.

> By the way, I made a simple top module and tried to
synthesize the code 
> today. It started to work but I tried to re-run the
synthesis and 
> Synplify Pro just sat there for hours running 100%. I'm
not sure what's 
> going on with that - maybe it just needs a reboot. I
don't know how much 
> space it will take up in an FPGA. It's a 155 stage 32
bit pipeline which 
> seems large by my standards. It is for this reason that
I don't know 
> which device it will fit in (I use Xilinx S3-200 right
now) or how fast 
> it will run.
> 
> I will give out the code once I figure out if there are
any reasons not 
> to do so. It's basically the code I posted to this list
10 months ago 
> converted to Verilog and printed out with tac. It's
also part of my 
> graduate thesis.
> 
> IMO the reward for breaking 72 bit encryption should be
$100k
> --
> Martin K
> 
> 
> John L. Bass wrote:
> > 	Out of curiosity... is there any way to safely
embed our encrypt/decrypt
> > 	functions in that code without it becomming
public? Otherwise I'm not
> > 	sure how we'd be able to plug one of these in to
the network...
> >
> > The answer is still effectively no. There would
have to be a standard
> > hardware configuration to freeze it. Or at least a
fixed set of cores
> > for every usable FPGA, which at least in the
beginning would require
> > releasing the crypto to a half dozen or better
FPGA developers (most
> > of the FPGA RC5 user community).
> >
> > 	On Wed, Oct 18, 2006 at 12:55:32PM -0600, Olivier
Meyer wrote:
> > 	> >I just thought I'd let you guys know
since this seems to be a 4 msgs a
> > 	> >year mailing list 
> >
> > Which is why this list is dead. The proprietary
interests block open
> > access.
> >
> > John
> > _______________________________________________
> > Hardware mailing list
> > Hardwarelists.distributed.net
> > http://lists.distributed.net/mailman/listinfo/hardware
> >   
> _______________________________________________
> Hardware mailing list
> Hardwarelists.distributed.net
> http://lists.distributed.net/mailman/listinfo/hardware
> 

-- 
Jim C. Nasby, Database Architect            decibeldistributed.net
Give your computer some brain candy! www.distributed.net
Team #1828

Windows: "Where do you want to go today?"
Linux: "Where do you want to go tomorrow?"
FreeBSD: "Are you guys coming, or what?"
_______________________________________________
Hardware mailing list
Hardwarelists.distributed.net
http://lists.distributed.net/mailman/listinfo/hardware

On Wed, Oct 18, 2006 at 02:56:12PM -0500, Jim C. Nasby
wrote:
> > Which is why this list is dead. The proprietary
interests block open
> > access.
> 
> Ideas for ways to do away with the crypto most welcome.


Damn, accidentally deleted your email.

In any case, I should have been more explicit: what we need
are actual
plans that remove the need for any proprietary code, and
then people who
will put the effort in to make them happen. Some of what you
mentioned
has been talked about (namely a means to positively identify
what user
is submitting a particular work unit), but that's still just
the tip of
the iceberg.
-- 
Jim C. Nasby, Database Architect            decibeldistributed.net
Give your computer some brain candy! www.distributed.net
Team #1828

Windows: "Where do you want to go today?"
Linux: "Where do you want to go tomorrow?"
FreeBSD: "Are you guys coming, or what?"
_______________________________________________
Hardware mailing list
Hardwarelists.distributed.net
http://lists.distributed.net/mailman/listinfo/hardware