SoylentNews is people
SoylentNews
Arthur T Knackerbracket has found the following story:
Physicists at the University of Bonn have managed to create optical hollows and more complex patterns into which the light of a Bose-Einstein condensate flows. The creation of such highly low-loss structures for light is a prerequisite for complex light circuits, such as for quantum information processing for a new generation of computers. The researchers are now presenting their results in the journal Nature Photonics.
Light particles (photons) occur as tiny, indivisible portions. Many thousands of these light portions can be merged to form a single super-photon if they are sufficiently concentrated and cooled.
The artist's rendering shows how potential wells are created for the light in the microresonator through heating with an external laser beam (green).
The individual particles merge with each other, making them indistinguishable. Researchers call this a photonic Bose-Einstein condensate. It has long been known that normal atoms form such condensates.
Prof. Martin Weitz from the Institute of Applied Physics at the University of Bonn attracted attention among experts in 2010 when he produced a Bose-Einstein condensate from photons for the first time.
In his latest study, Prof. Weitz' team experimented with this kind of super-photon. In the experimental setup, a laser beam was rapidly bounced back and forth between two mirrors. In between was a pigment that cooled the laser light to such an extent that a super-photon was created from the individual light portions. "The special thing is that we have built a kind of optical well in various forms, into which the Bose-Einstein condensate was able to flow," reports Weitz.
Journal Reference: David Dung, Christian Kurtscheid, Tobias Damm, Julian Schmitt, Frank Vewinger, Martin Weitz & Jan Klärs: Variable Potentials for Thermalized Light and Coupled Condensates, Nature Photonics, DOI: 10.1038/nphoton.2017.139
(Score: 0) by Anonymous Coward on Thursday August 17 2017, @04:32AM
Negative. I've been using quantum proof crypto since I first heard of quantum computation in the 90's; It's dead simple to create, but the foolish "don't roll your own" mantra keeps the sheeple from adopting better crypto so fuck them. Anyone who wants quantum proof crypto already has it.
Some quantum encryption works via entangling then transporting particles possibly teleporting them. However a quantum computer is not limited to only that type of processing. Imagine a qubit, a quantum bit which can be a super position of 0 and 1, and it can be entangled through a system representing a traditional block cipher algorithm, such as those of AES. If you put in 1's and 0's you get 1's and 0's out, but if you put super positions in you get super positions out, in all the bits that depend on the qubit in the undetermined state.
If you have cipher text then you just run the qubits through the calculation, entangling them with only superpositions as input. With enough qubits your quantum system then represents every possible input and output to the algorithm. Next you collapse the output qubits into the known ciphered text and the cascade occurs, collapsing the quantum waveform upstream until the plain text input is revealed. Because the cipher is reversible this is vulnerable to quantum attack.
Imagine instead if you have a one way cipher that throws away bits as it is processing each block, so that it is not reversible, but instead requires that each prior block be known before the next block is produced. This is easy to achieve with a keyed hash algorithm. Each successive block of data input is hashed and used to encrypt the next block AND a large portion of internal state (RAM) is thrown away in the process, as seen in MD family or SHA family, and basically all good hashes. This loss of information poses an impossible barrier to the qubit.
One can not simply collapse the quantum waveform based on the algorithm and cipher text because each block is dependent upon all the prior blocks, the key, AND the unknown state of a batch of working memory which has been discarded. In effect the "unknown" has been factored into the cipher. This is also a form of "Authenticated Encryption" which cryptographers are just now getting ironed out, but I've been using for over 2 decades... You can plug any pseudo-random number generator or hashing algo into such a system to upgrade it at will. I consider quantum proof ciphers a solved problem.
All the hype about quantum encryption should only be about the ability to guarantee tamper proof communication and snooping detection. Most of the cryptographic side of things has been solved via pre-shared key. The primary problem then comes from the initial public key exchange which can include no unknowns.
Right now your web browser will pop up a dialog for HTTP Auth digest / password protected directories. This works via proof of knowledge. The server knows your password based on your user name. Server generates a nonce and gives it to client. Client hashes the nonce with the password and gives the result to the server, server looks up the PW hashes it with the nonce and compares it to your supplied hash. Instead of exchanging the hash, all the client and server need do is exchange nonces / user names. Then both the client and server hash the password with the nonces and use the resulting hash to key the cipher suite. BAM, done. No 3rd party required. No public key crypto required. You can go to your bank and set up a password then securely connect with that method without fear of MITM, and no compromised certificate can cause your security to be spoofed. Both ends must know the password to key the cipher streams. This is basic end to end encryption, which we strangely DO NOT HAVE in our browsers, despite it being as easy as dropping the hash from HTTP AUTH into our cipher suite. end to end encryption using hash based ciphers is safe against quantum computing. The problem is in the initial key exchange during account creation. That is a pretty small window and if not observed then MITM attacks would not work on end to end hash based ciphers.
Since we already have pre shared keys with every site we want to visit securely, the only time we need public key crypto is during account creation. For low risk accounts, such as news forums it may be acceptable risk to use public key crypto to set up the account. Account setup is the only weakpoint, not every connection creation. For high risk accounts, the use of a secure channel such as an app with a pre-shared key or a 2 factor authenticator, or a visit to the bank, or exchanging the key in person with a friend, etc. would be used. Current implementations of public key crypto is weak to quantum computation, so if account creation was observed then future connections could be decrypted, but if we threw away public key infrastructure everything would actually be more secure than it is today. Currently we have no option but to insert a potential MITM into every connection via the very broken CA system. [youtube.com] All our connections use a key exchange based on prime number exponentiation or elliptic curve derivation. There is no way to opt out and use a more secure pre-shared key option. There are ways to do public key exchanges and asymmetric encryption that are quantum proof, but I'll not detail them here as export regulations forbid it, but they do exist. There's at least one option in homeomorphic encryption.
TLDR - If security is ever "threatened" by quantum computers it is because governments want it that way. Same reason your browser must rely on a potentially compromised or coerced / compelled CA and you have no option to just use the pre-shared password for authentication+encryption. Note that the quantum industry likes scaring consumers as it makes the tech seem more valuable than it is. Fear not the scaremongering citizen.