Re: Good enough for crypto?

Scott Wilber wrote:
> 

> QNG is a recognized trademark from our years on the market.  The PC
> part just stands for Personal Computer.  "QNG" does indeed stand for
> Quantum Number Generator, but this does not mean that our generators
> derive their entropy entirely from pure quantum sources.  This would
> have been prohibitively expensive, and hard to manufacture.  All our
> generators do have a quantum component in their entropy source.  This
> comes primarily from shot noise in semiconductor junctions, and to a
> much smaller extent from thermal noise, which, as I mentioned earlier
> in this thread, provides mostly chaotic noise.
> 
> As our generators advance, the portion of entropy derived from quantum
> sources increases.  Our latest hardware generator, the ComScire QNG
> Model J1000KU, (1Mbps) derives most of its entropy from shot noise.
> Hence its entropy source is almost entirely quantum.
> 
> The quantum component in the PCQNG is smaller, since we were
> constrained to use only the hardware components programatically
> available in a typical personal computer.  This compromise was made in
> order to provide a high quality, very low cost TRNG for convenient use
> by anyone with a PC running Windows.  The PCQNG is a true random
> number generator in the sense that is commonly in use today.  

I still think that the use of the term 'quantum' tends
unfortunately to suggest that the technology is in the 
direction of what is currently done in 'quantum 
cryptography'. Anyway, would other hardware generation 
of randomness, e.g. Lavalamp etc, be also appropriately 
qualified with 'quantum'? (I don't think that's good.)
Another difficulty I have is with the term 'mathematical
analysis'. In my humble view, that would mean in the
present context building a (purely) mathematical model
to interpret/evaluate the data (i.e. the bit sequences
obtained), like employing the ARIMA model in time
series analysis, which apparently isn't the case, if
I don't err.

M. K. Shen