We’re creating a more connected travel industry, underpinned by sustainability and long-term investor relations.
Technology Evangelist and Architect, Amadeus
Tech giants around the world are competing with a few ambitious start-ups to build the world’s next super computers. What are they, how do they work, and why should you care? Here’s a crash course on quantum computing, the next big thing that could change the world of computing as we know it.
Before we start talking about them, we have to start with what we’re all using today. Contemporary computing is based on binary digits – or bits. Everything you see, type, do, on the computer you’re using right now is decided on by a combination of two digits, either 0 or 1. Binary code is at the heart of everything that goes on inside our computers – it is a system of representing numbers, letters, commands, images and sounds.
If the basic building blocks of traditional computing are bits, for quantum computing they are qubits (quantum bits). And this is precisely what’s unique about quantum computers: they don’t use binary code. They use qubits.
Traditional computing is based on laws of classical physics. Similarly, quantum computers are based on the laws of quantum physics, which says that qubits do not have the exact state of 0 or 1. Qubits are built on the so-called superposition of 0 and 1. Once measured, the result will be either 0 or 1, but until this happens, a qubit exists in both states simultaneously.
A good analogy for illustrating bits and qubits is a sphere. A bit can be at the poles only. A qubit instead can be at any place on that sphere.
Let’s explore this idea further. If we combine two bits then they can be 00, 01, 10 or 11 but only one of these. Qubits instead will represent the superposition of all four possible states. In general, a quantum computer with n qubits will be in a superposition of 2^n states simultaneously. The traditional computer instead will be only in one out of these.
When people talk about quantum computing, they often refer to the famous example of Schroedinger’s cat . This is a thought experiment used by physicist Schroedinger to explain his theory of quantum physics. His experiment involves an imaginary cat trapped in a metal chamber with a radioactive atom that may or may not have emitted radiation.
Until you open the chamber, you do not know if the cat is dead or alive, and so the cat, says Schroedinger, is simultaneously dead and alive, until it is observed or “measured”.
In more technical terms people often say: if you want to know that state of something, you need to measure it, but the act of measuring it will influence the result.
Schrödinger's cat - click for source
The Trekkies among you may prefer the Heisenberg Uncertainty Principle as an example. Heisenberg said that it is not possible to measure the location and speed of a particle with a desirable precision, because it is always moving and changing. One unfortunate consequence of this is that “beaming” like in Star Trek, is not technically possible. In order to achieve tele-transportation, we would need to capture the exact state, location and speed of every single atom, electron and particle of the human body. This is why in Star Trek, the transporter is built with a “Heisenberg Compensator”; its entire purpose is to invalidate the Heisenberg Uncertainty Principle.
We could delve further into the science of tele transportation, but since this YouTube video does it so well, we thought we could go back to the original topic of quantum computers….
OK, so all of this is fascinating, but how will this impact computing?
Consider for an instant that your computer is actually just a very advanced calculator. Things like sending and receiving emails, writing and printing documents, or creating and changing pictures are based on fundamental mathematical operations like addition or subtraction.
Traditional computers use “gates” to perform basic calculations of bits, which means they’re adding or subtracting all sorts of variations of 0s and 1s. A simple case is taking two bits as an input, to combine them into a single bit which is the output. In tech-speak this is called performing a logical operation (AND or OR) to realize a Boolean function.
Below you see the AND gate and its associated Boolean operation table. The technical realization of these logical gates is done with transistors. A modern chip of a traditional computer can contain up to 100 million transistors per square millimeter. These transistors are the foundational bricks of traditional computing.
Logic Gates - click for source
With traditional computing, because all of the inputs are clear and well defined—they are either 0 or 1 – the results are clear and deterministic as well.
However, with quantum computing, there is some uncertainty on the input – remember, we don’t know if the cat is dead or alive until we measure it. As a result, quantum gates are a very different to the traditional ones. Instead of performing logical operations they manipulate the qubits themselves –they perform an “action” on a quantum register. The goal of that manipulation is to increase the probability of a certain outcome.
One consequence is that there are no quantum transistors. Instead, there are devices to execute an action on multiple qubits at the same time.
At the moment there are competing technologies that can perform this action. A qubit can be represented with a little magnet, and influenced with a much bigger outside magnet to change its position. A qubit can also be represented by the energy state of an atom, which will be influenced by an outside an energy impulse.
These are just a couple examples, there are several ways physicists can take an action on a quantum register to measure a qubit. The important point is that there is no singular implementation of a quantum transistor.
Because quantum computing is all about calculating probabilities, the real power of quantum computing lies in dealing with problems where many calculations must be done in parallel to lead to a single result at the end. In math speak people talk about optimization or number theory problems. Studies have shown that some problems that take years to solve with traditional computing, can be addressed with quantum computing in a matter of months or even less. The experts have created the term quantum supremacy which defines the speedup of quantum computer over a classical one in a particular field.
For example: Google published a research paper where quantum computing was 10^8 times faster than the traditional counterpart. In real numbers this means the solution was found in less than a minute rather than taking more than 30 years.
In order to build and use quantum computers a few challenges must be overcome. One is energy – it takes a vast amount of energy to power just one calculation on a quantum computer. Another challenge is how to initialize the qubits with arbitrary values, which means controlling the input to the quantum calculation. The number of qubits is still rather limited – below 100. Hence, only a few selected ones will be able to master challenges. Most of us will use quantum computing provided as a cloud service. Actually, IT giant IBM has already such an offer in its portfolio.
There is no doubt that quantum computing age has started to become real. It will still take a few years until we will manage large scale quantum computers. But for some specific problems the so-called quantum supremacy is already achieved at the order of 50 qubits. In the meantime, Amadeus is preparing itself for this eventuality by reading the latest research, and discussing this topic with our chief technology leaders.
Stay tuned for our next post, where we will talk about how quantum computing could impact the travel industry and Amadeus in particular.