Re: [Demexp-dev] VIA x86-like processors with cryptographic facilities

[David MENTRE]

Brian Hurt <address@hidden> writes:

> But there is still random information there.  Even if the number is 70%
> 1's and 30% 0's, you're still getting (approximately) 0.6 bits of random
> information per bit (actually, it's a little bit less than this- I'm
> forgetting the actual formula at the moment).  So if you need 256 bits of
> truely random bits, you'd need to collect 256/0.6 or 427 bits of biased
> bits, and then hash the value down to 256 bits, "concentrating" the
> randomness.

Thank you, I wasn't aware of this issue and possible solution to it.

Yours,
d.
-- 
 David MENTRÉ <address@hidden>