Re: [Cfrg] 3DES diediedie

[Peter Gutmann <[e-mail]>]

David McGrew (mcgrew) <[e-mail]> writes:

>See for instance slides 8 and 9 of Daniel Shumow's talk at NIST’s LWC
>workshop last year:
>http://csrc.nist.gov/groups/ST/lwc-workshop2015/presentations/session4-shumow.pdf

So looking at slide 6 from that, the first four systems he lists are desktop
PCs (in all but form factor), it's only the last two that are down at the
resource levels of IoT.  I'm not sure why he picked the Arduinos there because
I wouldn't really consider them terribly representative of IoT devices, was it
to get something that people are familiar with?  Even if you're wanting to
restrict yourself to well-known complete systems I think at least an ESP8266
(80Mhz SoC with 96K RAM, 64K flash, no multiply or divide by default) should
get a mention.

Slide 9 is even further removed from IoT practicality, that stuff may be fine
on the PC-equivalents but won't work on real IoT gear.

I'm currently working with some embedded systems guys to come up with a list
of requirements for IoT crypto (as with the TLS-LTS stuff, various IP/legal
issues means many contributors don't want to say anything in public), I'll
post it to the list when we've finished arguing :-).

Peter.