[E-Lang] Re: [EROS-Arch] Liquid software requirements.

Mark S. Miller markm@caplet.com
Sun, 31 Dec 2000 23:18:42 -0800

I think Jonathan wrote:
>> In the OS context, this has a further avantage. If the 
>> encryption can be
>> pushed down to the link layer in the communications system, it can
>> frequently be offloaded to hardware, at which point we cease 
>> to care very
>> much about its computatoinal cost.

For single key operations, which is what would apply here, offloading them 
to separate hardware makes very poor use of both pieces of hardware.  If 
you've got separate hardware, use it for something that's either 
computationally intensive or has special needs.  Single key crypto is 
neither.  It would be only a little less plausible to imagine special 
purpose hardware for lowercasing strings of characters.

For PK operations, it would indeed be very cool to have special hardware for 
doing modular exponentiation of large integers.  I remember a chip design 
for this by Rivest et al from about 15 years ago.  Has anything happened 
with this idea?  Will elliptic curves make this issue irrelevant?

AFAIK, the only compelling needs for special purpose hardware are good 
entropy and TEMPEST shielding.  (And Opaque Boxes as well, if you're into 
that sort of thing.)