+1 for adoption, since the work is relevant to wg.

Few remarks:
A section on availability of hardware cryptos and sample comparison to
software crypto (on the same platform) might make the document more
relevant. Most of the boards support some form of hardware crypto (mostly i
ve used TI and STM boards). Also i guess Arduino Uno (example board used in
the draft) also has an addon shield for hardware based crypto. Apart from
improvements in execution time and memory, such hardware cryptos are
usually FIPS certified and may be a requirement for some product lines
before they enter a specific market. From implementors perspective, such
information will help in formulating hardware requirements for a
product/solution.
A key take-away for me from the current document is that software crypto is
horribly slow and unusable! Its a given that software crpyto will be slow
but the data puts it into perspective on how slow it could actually be!

Also implications on multi layered security (section 13) needs more
attention and I feel alternatives could be elaborated.

Regards,
Rahul
_______________________________________________
Lwip mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lwip

Reply via email to