Why should you be interested in quantum computing?

The theme shouldn't intimidate anyone since I try to consider it from a practical perspective. Before quantum computing becomes part of the project portfolio, as big data has become, it will probably take more than a decade. Or not. But if you like lifting the hood to understand what we're talking about, let's look at it.

What is quantum computing?

It's been a while since I wanted to explore quantum computing, and finally, Hector Delesse's speech about "Informatique quantique, entre prouesses et promesses" created the occasion.

Beyond the technical explanations, I appreciated that he mentioned a few articles to sort out between true and wrong:

• TRUE: the quantum computer will be able to help fight against Alzheimer's disease or global warming. On this last point, its computing power allows, for example, to optimize the routes taken by cargo ships with a speed of execution that has nothing to do with traditional computing.
• WRONG: Quantum computing will replace traditional and double productivity within two years.

The critical point that I retained is that quantum computing makes it possible to solve complex problems at a time scale acceptable to humans.

In passing, Hector seizes the opportunity to criticize the GAFAMs for their announcement effects. In 2021, Microsoft was caught flat-footed after announcing that it had made a significant discovery, which made it hope that a solution would be available within five years. The story of the fall is told in this article by Wired.

If you have some appetence for scientific concepts (and master French), you would certainly enjoy watching Fanny Bouton and Olivier Ezratty in their video "Le quantique, c'est fantastique."

By the way, I laughed at their aparté, and I hope it translates well:

“We realized that from year to year we were increasing the level of complexity of the subjects we were dealing with and for this reason we invented a scale, which we call the exponential scale of inbitability.”

If you're wondering, a quantum computer looks like a big refrigerated can pinned to the ceiling with many cables.

What I promised myself to remember:

• Quantum: we speak about infinitely small, as it concerns elementary particles like electrons and photons. They move in all directions and can be in two states simultaneously (superposed state). Other principles of quantum mechanics are used in quantum computing, but I don't think it is necessary to remember them.
• In quantum computing, we speak of Qubits and not Bits (the series of 0 and 1 that allow coding information). These Qubits have the particularity to be both 0 AND 1 simultaneously (plus other properties), which allows for much more complex calculations.
• The quantum co-processor will not replace the traditional computer but complement it for different applications. For those who want to see what the architecture of a quantum computer looks like, watch here. On my side, I wasn't able to follow, but I am an economist, not an engineer!
• I reconnected at that point when the authors talked about quantum computing applications. The image below shows simulations to improve traffic jams in Beijing with 400 cabs. On the right, the picture below shows the result with a quantum algorithm, to be compared with a classical algorithm on the left:

When will quantum computing be available?

Not all experts agree, but some think quantum computing will only be available as a technology around 2035. But some states and some big players like IBM already have computers.

The point that may give blockchain and cyber security players cold sweats is that the quantum computer could offer the resolving power needed to "crack" RSA public keys (used in cryptography and for bitcoin). The horizon given by this Global Risk Institute report is also around fifteen years.

Fortunately, advances in post-quantum cryptography are sufficient to make the shield available even before the widespread use of quantum computers.

The CISO's opinion: The good news is that to protect your servers, you only need to increase the length of the encryption keys, which is feasible with traditional computers. Some algorithms may become obsolete, and will need to be replaced. The NSIT has selected algorithms that are likely to withstand quantum decryption. We need to start looking into this and monitor developments.

The question is how companies will implement it and at what cost.

To go further, you can read this article on Entrust's website.

When will the quantum computer be commercially available?

In September 2020, D-Wave announced the general availability of a quantum computer for enterprises. What's interesting about their offering is that it's available in the cloud and comes with startup support.

IBM also proposes a quantum computing offer based on usage.

For the quantum computer, we understand that the road will be long, with the delays impacting Rigetti, one of the best-funded startups in the field. Instead of the 128 qubits, it thought it would be able to make available by 2020, and it delivered only 32 in February 2022! Not easy to keep one's promises in such an innovative field.

If you want to deep dive into the topic: Olivier Ezratty created a free 800-page ebook with no less than fifty pages on applications. So, I leave him the last word :

The feasibility of commercially useful quantum computers remains an open question. There are significant technologic challenges to overcome […]. So, quantum computing is full of scientific and technological uncertainties even before being economical and market ones.

I

‍