Quantum supremacy is there, I get it, Google tells us. IBM disputes. The battle is on. What about the quantum computer in supermarkets, is it coming soon? Is the quantum claims to meet the IT challenge legitimate?
This horrific quantum mechanics, it explains everything, except relativity; it allows everything or almost everything from the internet to medicine and lasers. It is undeniable: it is quantum mechanics. The counterpart is that we must ignore his intuitions and leave his common sense far from home in the face of a nature that Feynman announces to us absurd.
Table of Contents
Superimposed and entangled
First counter-intuitive behavior: a particle “only” exists when it is observed. Between two interactions its state is indeterminate. The notion of trajectory disappears. A particle is seen in A than in B but between the two it has no reality. All possible states coexist, are superimposed. At the slightest interaction, we leave the quantum state by taking one of the states superimposed at random. It’s a decoherence. Randomly, certainly, but with a deterministic probability.
The interference patterns created by a light beam and two slits to reach a screen testify to this. Between the emission and the arrival on the screen, the photon is not materialized, there is no trajectory. The position of the impact on the screen is random but follows probabilities dictated by the two possible paths. The result is superb clear bands where the probability of impact is maximum and dark when it is low. We can see spectacularly these figures being built photon after photon.
The second test for common sense: two interacting particles remain linked regardless of their distance. This entanglement was the object on the part of Albert Einstein of a fundamental thought experiment ( EPR experiment ) put into an equation by John Bell then highlighted by Alain Aspect with Philippe Grangier, Gérard Roger and Jean Dalibard. The entanglement is experimentally confirmed. It is used in computers and in quantum teleportation.
The quantum state is a singular but precarious world. Nothing is enough to provoke a return to the classical world. Decoherence is a strong limit for computers which must be carefully isolated from the outside world to carry out their calculation: shielding and operating temperature close to absolute zero are required, quantum machines will not stand on a desk. Clever error calculation algorithms compensate for certain disturbing effects that cause decoherence at the cost of significant swelling of quantum computing resources. Quantum machines are making great strides. Two qubits in 2003, fourteen in 2011, twenty in 2018 and fifty announced in 2019. It is not huge but we would like to see the dawn of a new Moore’s law. Fantasy?
The quantum world being counter-intuitive, the quantum calculator will be without a strange surprise. Known programming on conventional computers no longer works: the results are probabilistic, the “copy” function does not exist. Designed for a dedicated type of algorithm, the quantum calculator does not have, like the conventional computer, the predisposition to accommodate any type of program.
Quantum bits, qubits, are in an indeterminate state where the values 0 and 1 are superimposed. The sixteen values superimposed on a four-qubit register can be tested by a single measurement where sixteen measurements would be necessary on a conventional computer. The parallelism of quantum computers opens the way to tenfold power.
Expectations are high to see the power of quantum prevail against the classic. The proud and thunderous designation “quantum supremacy” is less demanding: identifying ONE quantum application more efficient than classic.
It should be noted, however, that the application is generally specific and the evaluation of the treatment in classic questionable. The recent controversy between Google and IBM is an illustration and even hides the technological exploit of Google which reaches fifty qubits.
Security encryption in the quantum era
In 1994, Peter Shor shakes the banking world by proposing an algorithm for quantum calculator likely to break RSA cryptographic encryption. The difficulty in decomposing a large number into prime factors is exponential on a conventional machine and polynomial on a quantum machine, which would greatly speed up the case of RSA security keys. That said, the press gives us a break of a few decades. Indeed, Shor’s algorithm requires twice as many qubits as the size of the key to breaking and several billion quantum gates. The transition from some fifty qubits to several thousand may be made, but are the technologies to achieve this only identified? For my part, this is the question I ask myself as soon as a period of forty years is announced: “forty” in the forty thieves of Ali Baba meant “a lot”.
Close by words, the concepts are different. “Post-quantum” refers to conventional encryption algorithms capable of withstanding the dreaded quantum power. The name quantum cryptography is flattering but misleading: encryption owes nothing to quantum computers, it remains classic. Quantum is used to transmit encryption keys: it detects any intrusion attempt. It is the quantum distribution of keys, a function now provided by the RSA protocol.
The range of the distribution is limited by the attenuation in the fiber and the conventional repeaters of the transmission networks are ineffective: they would destroy the quantum state. Fortunately, teleportation is there.
Quantum entanglement binds two particles regardless of their distance. The fantasy is to teleport objects instantly, but the reality is more prosaic: only the quantum state of a particle can be teleported at a speed limited to that of light. Whatever, we don’t ask for more to extend the transport of a quantum key. China is moving forward on this subject.
So, is the quantum computer for tomorrow?
Alas no, you will not have a quantum computer on your desk soon. Office automation is not on the program and temperatures close to absolute zero are crippling.
Science will likely have a lot to gain from simulations, cryptography will see a revival, will the Internet have its revolution? The gains from research around the quantum computer could be invisible because indirect but tangible.
To deepen, I would like to suggest the essential lessons and lectures full of humor by Richard Feynman (in English), as well as the very complete, very well written and educational blog of Olivier Ezratti.