AI as a success factor – smart technologies for the economy

TREND REPORT wants to actively help make information, experience and knowledge about digitalisation freely accessible to all. The handbook AI deals with the new technologies and their impact on the most important industries. In the open-content work, journalists and guest authors from science and business present innovative developments and their application potential in the B2B and B2C sectors.

https://handbuch-ki.net/ (German Version)

Chapter 3.8: Quantum Computing 

Author: Georg Gesek

Quantum Computing, today! 

What is Quantum Computing?

The computers in our everyday lives are so-called Turing machines, named after the English mathematician and computer pioneer Alan Mathison Turing. These calculating machines are characterised by understanding numbers, which are represented in binary in our current systems, as commands for manipulating data, which in turn represent other numbers.

This makes it possible today to build very powerful computers that can manipulate binary data, i.e. data composed of only 2 characters, e.g. 0 & 1, very quickly by means of a programme code. We call these manipulations logical operations, or gates, when we speak of the physical construction of these computing units. A very often used example is the NAND gate, which generates an output state from two input states. Since these are binary states, so-called bits (a made-up word from “binary digit”), the result is an input pattern with 4 possibilities (0/0; 0/1; 1/0; 1/1), whereas the output has only 2 possible states (i.e. either 0 or 1). In the case of the NAND gate, the output state is always 1, unless both input states are 1, in which case the output becomes 0 (“Not AND” operation). Therefore, computational operations with logical gates in Turing machines can usually no longer be reversed, since information is lost. In the same way, the data memory can be deleted.

Thus, entities in Turing machines, such as in the case of a virtual reality, are of a different nature than those in our physical universe. Or have you ever experienced that an object suddenly disappears from the universe because the information about it has been deleted?

In fact, we know today that the information that makes up our reality cannot be deleted in principle. It cannot even be copied! The information about an object or its elementary particles that make it up is therefore of a different quality than the bits in a classical computer: we call this quality quantum information.

The findings about quantum information since the beginning of the 21st century are remarkable. Humans have now developed technologies with which we can specifically manipulate this quantum information: We group these together under the term quantum technology.

A quantum computer is nothing other than a machine for the controlled manipulation of quantum information, with the aim of running a previously defined programme, i.e. an algorithm. The algorithm then serves to process a computational task, which solves a concrete problem in the sense of the programme, for which it was written.

Therefore, the purpose of a quantum computer is no different from that of a Turing machine. Why then are IBM, Google, Microsoft but also start-ups like Rigetti or my own company NOVARION making such gigantic efforts to build a universally applicable quantum computer?

Without having to know now how a quantum computer works in detail, one thing is already clear, and another outrageous insight of the 21st century: our reality is obviously nothing but the computational result of a gigantic quantum computer that we call the universe!

Against this background, we can now also understand that even a small part of the universe, which we can control technologically and understand as a human-built quantum computer, must have tremendous computing power. And more than that, quantum computing must be the most efficient way to compute a physical system. If there were a simpler algorithm, our universe would have used it.

This is why building a universal quantum computer is so appealing: because of the sheer potential computing power. Incidentally, the term “universal” refers to the type of quantum gates, i.e. the possible logical links of quantum information. This is supposed to be possible in such a machine without limits in principle, just as Alan Turing formally proved that his computer can also calculate all possible algorithms in the universe in principle.

But wait: That also means that a Turing machine must be able to calculate quantum algorithms. That is also the case! But as the Nobel Prize winner for the development of quantum electrodynamics, Richard Feynman, remarked in the early 1980s: “Why do all my computer simulations of quantum physics take so long? Why can’t we use native quantum computers?” As far as we know, this was the first time Feynman ever uttered the term quantum computer.

This means that quantum computers, because of their different way of calculating, are also potentially drastically more powerful than our computers today. This is why billions are now being invested in this promising technology every year. This different way of computing is characterised by two specific quantum mechanical effects, which we call superposition and entanglement. Superposition allows a quantum machine to superimpose two or more input states in a single memory cell. We call these memory cells qubits (quantum bits) in analogy to conventional computers. Entanglement, in turn, allows two or more quantum bits to share the same quantum information without having to be in physical contact with each other, no matter how far apart they are spatially.

Because of these two remarkable properties of the qubits, one can already imagine that the computing speed can be considerably greater than with the Turing machines because of the parallel information storage as well as processing.

Where do we stand today?

If you knew, you could get any prize from the big IT companies for this information! At the moment, bets are being placed on whether the qubits will be built with extremely deep-cooled (millionths of a degree above absolute zero) solids or perhaps better with ions (electrically charged atoms) trapped in magnetic fields and isolated in a vacuum. The first technology has so far proved difficult to control and the second equally difficult to scale.

What we do know is that, in any case, we need a well-controllable and highly integrable technology for the manipulation of quantum information in order to build a commercially successful quantum computer, which for concrete applications in industry must therefore have a better price-performance ratio than our current and also future computers based on conventional integrated circuits in semiconductors.

Therefore, we also know that there are at least two technological milestones between today’s quantum chips and ion traps, which control a few dozen qubits more or less poorly, and the quantum computer used industrially on a broad scale.

My company NOVARION is therefore in the process of setting up a quantum electronics laboratory in Vienna in order to master these technological challenges ahead of us within the next 5 years!

How can companies prepare for the quantum revolution?

It should be mentioned that quantum technology will not “merely” produce new incredibly fast computer systems for the advancing digitalisation and artificial intelligence, but will also bring about gigantic advances, e.g. in medicine or telecommunications. In general, it can be said that quantum technology will revolutionise all systems that have to do with information, and later also those that are associated with energy generation or storage.

Therefore, I recommend that every company with an annual turnover of around € 100 million or more should already start looking at applications of quantum technology in their core business. The first companies to successfully implement quantum technology in their field will revolutionise their market technologically.

As far as the area of my own business is concerned, namely information technology, we remember Alan Turing’s elegant proof that our computers today can already calculate quantum algorithms. Although they take a relatively long time to do so, we can use them to prepare our software, in whatever field, for the next but one generation of universal quantum computers, so that when they become available, we can immediately realise the great competitive advantage that lies in this still sleeping giant. Because its awakening is coming for sure!

What is NOVARION working on?

The goal of my own company is the hybrid universal quantum computer. We have already explained the term universal in relation to all types of quantum algorithms, so what is meant by the extension “hybrid”? It turns out that quantum computing is far from opening up a more advantageous computational path for all types of problems.  In fact, to multiply two numbers directly, for example, we do not need a quantum computer. This task is so trivial that quantum algorithms do not give us any advantage. If, on the other hand, we want to decompose a large number into its prime factors, a Turing machine always scales the number of calculation steps with the input quantity (in this case the length of the number) in the exponent of a function. In this way, one very quickly encounters the so-called computer infinity, thus not leaving the basic computability of the problem, but will by no means experience the result of the calculation of the Turing machine.

Therefore, I see the development of a computer chip as a practicable approach, which simultaneously contains a Turing machine and a quantum machine in integrated circuits and thus represents a hybrid universal quantum computer on the chip.

Translated with www.DeepL.com/Translator (free version)