🚀 go-pugleaf

In article <1992Aug19.864.126@ALMAC>, keith.willis@almac.co.uk
writes:

> You mean that one could implement RSA in hardware for a
> reasonably low price?  And would such a device still infringe
> the patents (in the US), or do they only cover software
> implementations?

Chips that "do RSA" are in fact no more than modular
multiplication chips (they perform the calculation C=(M**E) mod N,
for C M E and N very large numbers, typically 512 or 1024 bits).

This can be done quite cheaply. Philips designed but never sold a
smart card to do it, I've been involved in another project to do
it, and some companies integrate such chips into their machines.
They are cheap but slow, faster than a S/W implementation though.

I can't see how Rivest Shamir and Adleman can have a patent on
performing exponentiation (patents cannot cover maths). They can
patent a way of doing it on a chip (I think, under US law at
least). Basically I think what I mean is that you could sell it,
but if you used it as part of the RSA algorithm then you'd have to
cough up the licence fee for the patent.

None of the abopve applies in the UK of course because the patent
only applies to the US.

If any of the above is inaccurate, let me know!

If anyone has a US patent Number for RSA can they let me know and
I'll go and have a read. I'm too poor to pay for the search, and
too busy to do it myself.

Simon Read.
-------

Sorry about the long line lengths I've been sending out. I hope
I've fixed the problem.

In article <1992Aug20.160528.2155@nessie.mcc.ac.uk> segr@nessie.mcc.ac.uk (Simon Read) writes:
>In article <1992Aug19.864.126@ALMAC>, keith.willis@almac.co.uk
>writes:
>
>> You mean that one could implement RSA in hardware for a
>> reasonably low price?  And would such a device still infringe
>> the patents (in the US), or do they only cover software
>> implementations?
>
>Chips that "do RSA" are in fact no more than modular
>multiplication chips (they perform the calculation C=(M**E) mod N,
>for C M E and N very large numbers, typically 512 or 1024 bits).
>
>This can be done quite cheaply. Philips designed but never sold a
>smart card to do it,

Not yet, but Philips will probably sell it in the future.  At least
they are still developing the card (and it does need development,
for instance, it's not yet ISO 7816 compliant, i.e. many smart card
readers cannot handle the Philips RSA card).



-- 
---
Paul Schlyter, Swedish Amateur Astronomer's Society (SAAF)
Nybrogatan 75 A, 6 tr,  S-114 40 Stockholm,  Sweden
InterNet:  pausch@saaf.se

In article <1992Aug20.160528.2155@nessie.mcc.ac.uk>, segr@nessie.mcc.ac.uk (Simon Read) writes:
> I can't see how Rivest Shamir and Adleman can have a patent on
> performing exponentiation (patents cannot cover maths).

They patented a ``cryptographic communications system'' based on the
RSA equations.  They didn't claim either exponentiation or any particular
way of doing it.  As a consequence, if you found a new use for those
exponentiations that's neither claimed by the patent, nor a derivative
of any of its claims, you wouldn't infringe, even though you were
doing the exact same things.

> If anyone has a US patent Number for RSA can they let me know and
> I'll go and have a read. I'm too poor to pay for the search, and
> too busy to do it myself.

4,405,829.