On Sat, Dec 25, 2010 at 2:12 PM, <cpol...@surewest.net> wrote: > ... > Check out Markus Jacobsson et al, "A Practical Secure Physical Random > Bit Generator", 1998, using the turbulence of airflow inside the drive > as the source of randomness. Can't do much better than that.
how much turbulence does my SLC FDE make? the reason i prefer on die is that pre-boot operations and/or host init can make use of these sources via built-in facilities without need for additional drivers to external devices that may in turn require bus initialization and interrupt allocation, and so on, etc. likewise, if bootstrapping a secure network requires strong random numbers a network based entropy distribution setup to hosts without their own physical sources is not so useful for that task. there are many other considerations weighting toward on-die implementations, like clock and sample rates, but proper hardware entropy engineering is a verbose tangent way too long for this already meandering discussion... [0] :) 0. if you're really curious, check out Cryptographic Hardware and Embedded Systems proceedings, any hw design texts by authors of these proceedings, and then you'll know what your known unknowns are and can brazenly blaze forward into the esoteric or halt early satisfyingly convincing yourself that you could give two shits about what it takes to build proper kit. _______________________________________________ Full-Disclosure - We believe in it. Charter: http://lists.grok.org.uk/full-disclosure-charter.html Hosted and sponsored by Secunia - http://secunia.com/