On Sunday, April 21, 2002, at 11:09  PM, Eugen Leitl wrote:

> On Sun, 21 Apr 2002 [EMAIL PROTECTED] wrote:
>
>> Why would one want to implement a PRNG in silicon, when one can
>> easily implement a real RNG in silicon?
>
> Both applications are orthogonal. PRNG != entropy.
>
>> And if one is implementing a PRNG in software, it is trivial to
>> have lots of internal state (asymptotically approaching one-time
>> pad properties).
>
> Yes, but software is too slow to be able to handle >GBit data rates. 
> It's
> inefficient use of CPU silicon real estate.
>

What real-life examples can you name where Gbit rates of random digits 
are actually needed?

Even high-bandwidth transfers of MPEGs, for example, will be done with 
conventional ciphers using only a tiny fraction of this bandwidth for 
the random number parts of the ciphers.

Speaking of real world issues, it's been half a dozen years since 
Goldberg and Wagner broke the Netscape "time of day random number 
generator." I've heard of no serious attacks when a PRNG is used 
properly. Not to say it can't happen, or won't happen, or hasn't already 
happened.

In any case, if someone wants Gbits per second of random numbers, it'll 
cost 'em, as it should. Not something I think we need to worry much 
about.

--Tim May
"As my father told me long ago, the objective is not to convince someone
  with your arguments but to provide the arguments with which he later
  convinces himself." -- David Friedman

Reply via email to