Well designed entropy sources are not slow. I don't know where this idea comes from. The entropy available from realistic circuits is in excess of 100Gbps per 100um squared. The limit is extraction which is algorithmic, but they can be pretty small too. Adding a PRNG is purely down to what SP800-90 requires. The PRNG slows things down in a bps/unit area sense. But it's there because the spec mandates it. Covering post quantum extractors would be something worth doing. I've yet to see a quantum RNG that's faster than modern macroscopic entropy sources.
would it be feasible to XOR slow quantum signal with padding with fast pseudorandom signal to have something that is both pseudo and truly random, and the question from mathematical point of view - is such stream a truly random or still pseudorandom?
Does it sound too bad to build a global unique random number generator that would distribute unique random number in secure way to every server that want to create a further keypair for their actual purpose, just like how let's encrypt doing in certificate front??
Randomness beacons are already out there and are useful for a bunch of things, but you don't want to use a public randomness source to generate your cryptographic keys. Otherwise, all I have to do is know roughly when you made your key, and I can record the randomness from that time window and figure out your key. The NIST randomness beacon comes with a big warning in fact, as you can see here csrc.nist.gov/projects/interoperable-randomness-beacons
Hi Graham, I made a video just recently but related more to quantum gravity and deformed systems. Would interesting to hear your insight on the random number generator I put in model thanks
Well designed entropy sources are not slow. I don't know where this idea comes from. The entropy available from realistic circuits is in excess of 100Gbps per 100um squared. The limit is extraction which is algorithmic, but they can be pretty small too. Adding a PRNG is purely down to what SP800-90 requires. The PRNG slows things down in a bps/unit area sense. But it's there because the spec mandates it. Covering post quantum extractors would be something worth doing. I've yet to see a quantum RNG that's faster than modern macroscopic entropy sources.
would it be feasible to XOR slow quantum signal with padding with fast pseudorandom signal to have something that is both pseudo and truly random, and the question from mathematical point of view - is such stream a truly random or still pseudorandom?
Does it sound too bad to build a global unique random number generator that would distribute unique random number in secure way to every server that want to create a further keypair for their actual purpose, just like how let's encrypt doing in certificate front??
Randomness beacons are already out there and are useful for a bunch of things, but you don't want to use a public randomness source to generate your cryptographic keys. Otherwise, all I have to do is know roughly when you made your key, and I can record the randomness from that time window and figure out your key. The NIST randomness beacon comes with a big warning in fact, as you can see here csrc.nist.gov/projects/interoperable-randomness-beacons
Graham, as always this is a great explanation!
Hi Graham, I made a video just recently but related more to quantum gravity and deformed systems. Would interesting to hear your insight on the random number generator I put in model thanks
That's impossible unless we have unlimited space for bits
"cheap easy to use random number generator" @ around 4:40...Quantum emotion has one, maybe?
Quantum blah blah blah...