This is a misleading headline! This 'hard drive' needs to be in a superconducting magnet and held at 2K. I'm not aware of any desktop chassis that comes with a a cryostat...
Can someone here give a quick explanation on how this would work?
Assuming that everything is stored as qubits and that their internal state is used as a way to increase storage (and they found a way to make the internal state not change over time), wouldn't reading from disk destroy this internal state (not to mention give a random output)?
Have they found a way to measure a qubit in some way without possible loss of information?
BQP doesn't really get you closer to EXPTIME at all. We don't even know yet whether it gets you closer to NP. In general, quantum computers only have a quadratic speed-up as far as we know.
I had to use EXPTIME rather than NP because of Savitch's Theorem.
> In general, quantum computers only have a quadratic speed-up as far as we know.
Not really true? In the time setting, there's the case of sampling problems, hidden subgroup problems, the evaluation of linear systems, etc. There are also many other settings (Merlin-Arthur-like round complexity) and communication complexity where exponential and even superexponential increases can be had.
This is super interesting. Right now we are limited by communication due to latency from speed of light. With quantum entanglement, does that mean no latency over great distances? Or is there a latency involved, even in quantum entanglement?
With one exception, you could use them for a synchronised random number generator.
It's not as useful as you'd hope, since a cryptographic PRNG theoretically does the same thing (and you need to individually transport the bits ahead-of-time), but you can at least be sure the numbers are random and no-one is spying. Whether or not that counts as communication is up to you.
Having quantum entangled particles stored in some crystal sounds better than requiring direct optical cable connection between peers for some kinds of quantum computer resistant communication.
You still need direct connection and you still need to do key exchange. Whether you're exchanging keys or quantum particles for the purpose of generating keys is immaterial for the purposes of security.
You mean you would want to entangle two particles, move them apart over great distance and then use them to signal information, more or less instanteniously? That would not work.
This isn't hype though. No one is pretending that you can buy this tech in "5 years". Every lab that wants this tech is expected to build it themselves because you can't just buy it anywhere - it's almost anti-hype!
This is great news for a capture everything NSA. Maybe light up every web cam and every microphone on the planet. What I would do is form a backdoor deal with HD manufacturers so that I got a generation ahead of regular consumer tech. The NSA would need to keep HD shipments from manufacturers off their books.
With its ~20 Billion dollar budget, the NSA could easily hoover (ha) up a large percentage of the entire hard drive market. It could spend 5% of its budget to capture %5 of the HD stream.
