I don't buy this. Specifically, the notion that random numbers from physical sources are important. The theory of pseudorandom generators is very well understo

I don't buy this. Specifically, the notion that random numbers from physical sources are important. The theory of pseudorandom generators is very well understood; and cryptographic PRNGs even come with the guarantee that there is no known way to tell the difference between its output and a truly random string. All you need, then, is 80 bits of true randomness as a seed for RC4. Of course there have been instances when random number generators have been broken, such as Netscape's SSL once upon a time, but that was because it used a non-random seed (the timestamp), and not because of a weakness in the algorithm. And this is all w.r.t. a malicious attacker. Simulations don't even come close to testing the limits of random number generators (in a provable way; Michael Mitzenmacher has a paper on this), and the Mersenne twister is in all likelihood sufficient. http://en.wikipedia.org/wiki/Mersenne_twister http://en.wikipedia.org/wiki/Mersenne_twister It's really cool that they broke the true-randomness bandwidth record, but does every discovery in materials science these days have to be justified with potential applications to computing?