Proof: BC Lottery Corp uses a microsoft OS in their machines

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Mark R

Diamond Member
Oct 9, 1999
8,513
16
81
How good is it? I've seen things out there like this that can command well over $1000.
Is it just the throughput of random bits that makes an expensive generator more desirable?

The intel RNG is believed to be excellent, it passes all mathematical tests of randomness, with no detectable bias or correlation over billions of bits.

The design, while not quantum, is very robust and well thought through. Unlike a true quantum RNG, it could theoretically be susceptible to perturbation - but considerable care has been made (e.g. 2 neighboring resistors on the die are used as the sensor element, and the two signals are subtracted, so as to minimize any manufacturing bias, or effect of external interference on the analog sensor circuit).

The limitation of the intel RNG is that it is relatively slow - about 70 kbps (due to the physical processes used for the randomness source, and because of the need for careful processing to remove biases). For some purposes, this isn't fast enough, and alternative (and usually more complex) processes are needed, and as these are niche products, they can often command high prices.
 

Jeff7

Lifer
Jan 4, 2001
41,596
20
81
The intel RNG is believed to be excellent, it passes all mathematical tests of randomness, with no detectable bias or correlation over billions of bits.

The design, while not quantum, is very robust and well thought through. Unlike a true quantum RNG, it could theoretically be susceptible to perturbation - but considerable care has been made (e.g. 2 neighboring resistors on the die are used as the sensor element, and the two signals are subtracted, so as to minimize any manufacturing bias, or effect of external interference on the analog sensor circuit).
Ah. Nifty. :thumbsup:



The limitation of the intel RNG is that it is relatively slow - about 70 kbps (due to the physical processes used for the randomness source, and because of the need for careful processing to remove biases). For some purposes, this isn't fast enough, and alternative (and usually more complex) processes are needed, and as these are niche products, they can often command high prices.
Ok, that explains that then.
That company has a bigger PCI-e card version with 4 of those modules squished onto it, for about 16Mbit/sec throughput.
That thing's neat: It uses a photon sensor and an angled semi-silvered mirror. The photons will either go through the mirror, or they'll reflect. One way's a 1, the other's a 0. And there's some kind of thing in there for biasing, since there'll be manufacturing variations in the silvering of the mirror.

Cool stuff.
 
Apr 20, 2008
10,067
990
126
I saw the same thing at O'hare on an ad television AND flight information TV the night I got to boot camp. It was likely my last laugh of freedom. People looked at me like I was crazy.