[Cryptech Tech] Seeking comments on a proposal for changes to the Cryptech RNG design.

[Peter Gutmann]

Bernd Paysan <bernd at net2o.de> writes:

>It is highly unlikely that you can continue to operate when your in-FPGA
>jitter-based entropy source fails.  

You can do that, that's what the Capstone design (and stuff I've worked on)
allows for.  Even if the noise source fails, or goes into a still-operational-
but-faulty mode, the various backups ensure that you're still getting usable
random data.

>How could that happen? 

Environmental conditions is the most obvious answer, either natural or
attacker-caused.

>But in any such failure mode, don't continue to operate without alarm.

This assumes that the device is running with a human operator standing by, and
that they can do something to remedy the fault, e.g. swap in a second HSM that
doesn't have the problem.  Often one or both of these are not possible.

Peter.