Quantum computing is a rapidly advancing research field that by 2030 promises to deliver a new generation of supercomputers. These may outperform traditional digital technology at certain tasks that are likely to include molecular and material modelling, logistics optimization, financial modelling, cryptography, and pattern matching activities such as deep learning artificial intelligence.
This page provides a non-technical overview of quantum computing prioners and developments, and provides support for my videos Quantum Computing 2017 Update and Quantum Computing 2018 Update, which are also embedded below. Note that the 2018 video focuses more on quantum computing pioneers and applications, while the 2017 video explains more about different quantum computing technologies.
Traditional or 'classical' computers are built from silicon chips that contain millions or billions of miniature transistors. Each of these can be turned "on" or "off" to represent a value of either "1" or "0". Conventional computers subsequently store and process data using "binary digits" or "bits".
In contrast, quantum computers work with "quantum bits" or "qubits". These can be represented in hardware using the quantum mechanical properties of single atoms, sub atomic particles, or superconducting electrical circuits.
Due to the peculiar laws of quantum mechanics, qubits can exist in more than one state -- or ‘superposition’ -- at exactly the same point in time. This allows a qubit to assume a value of ‘1’, or ‘0’, or both of these numbers simultaneously. In turn, this enables a quantum computer to process a far higher number of data possibilities than a classical computer and to perform massively parallel processing. It also means that every qubit added to a quantum computer increases its power exponentially.
The fact that qubits are more 'smears of probability' than definitive, black-and-white certainties is exceptionally weird. Flip a coin, and it cannot come up both heads and tails simultaneously. And yet the quantum state of a qubit can in some senses do just that. It is therefore hardly surprising that renowned nuclear physicist Niels Bohr once stated that 'anyone who is not shocked by quantum theory has not understood it!'
In addition to assuming superpositions, qubits can become ‘entangled’. ‘Entanglement’ is another key quantum mechanical property, and means that the state of one qubit can depend on the state of another. This is useful and powerful, as it means that observing one qubit can reveal the state of its unobserved pair.
Creating and manipulating qubits is very hard indeed. Many of today's experimental quantum processors exploit the quantum phenomenon that occur in superconducting materials, and hence need to cooled to almost absolute zero (around minus 272 degrees celsius). Significant shielding against background noise is also required, and even then performing computation using qubits requires significant error correction. Indeed, a grand challenge in quantum computing is the creation of a truly fault-tolerant machine.
Companies currently developing quantum computers include IBM, Alibaba, Microsoft, Google, Intel, D-Wave Systems, Quantum Circuits, IonQ and Rigetti. Many of these firms work in conjunction with major university research teams, and all continue to accrue significant progress. The following provides an overview of the world of each of these pioneers in turn.
IBM has been working to develop a quantum computer for over 35 years. It is also making significant progress, with several operational machines. According to the IBM-Q website, while 'today quantum computing is a researcher’s playground', in five years 'it will be mainstream'. As IBM goes on to explain 'in five years, the effects of quantum computing will reach beyond the research lab. It will be used extensively by new categories of professionals and developers looking to this emerging method of computing to solve problems once considered unsolvable'.
In 2016, IBM launched a website called the IBM Q Experience that made a 5 qubit quantum computer publicly available over the Internet. Since this time, this has been joined by a second 5 qubit machine and a 16 qubit machine, both of which are available for anybody to experiment with. To help those wishing to learn about and develop quantum computing, IBM offers an open source quantum computing software framework called Qiskit.
In addition to the above, in November 2017 IBM announced that two 20 qubit machines were being added to its quantum cloud. These can be used by clients who are signed-up members of the IBM Q Network. This IBM describes as 'a worldwide community of leading Fortune 500 companies, startups, academic institutions, and national research labs working with IBM to advance quantum computing and explore practical applications for business and science'.
Also in November 2017, IBM announced that it had constructed a 50 qubit quantum processor, which remains its most powerful quantum hardware to date.
In January 2019, IBM announced unveilled its IBM Q System One as the "world's first integrated universal approximate quantum computing system designed for scientific and commercial use. This modular and relatively compact system is intended to be used outside of a laboratory environment. You can learn more about the IBM Q System One in the this press release.>
Another tech giant that is working hard to make quantum computing a reality is Google, which operates its Quantum AI Laboratory. In March 2017, engineers Masoud Mohseni, Peter Read and Hartmut Neven who work at this facility published this article in Nature in which they contented that short-term returns from quantum computing 'are possible with the small devices that will emerge within the next five years', so supporting IBM's view of the timescale for commercial quantum computing to arrive.
Google's early with in quantum computing involved the use of a machine from Canadian pioneer D-Wave Systems. However, the company is now rapidly developing its own hardware, and in March 2018, announced a new 72 qubit quantum processor called 'Bristlecone'. And yes, we have already reached the point where quantum computing processors are given fancy names.
Over in China, the main web giant is Alibala, not Google. And in July 2015, Alibaba teamed up with the Chinese Academy of Sciences to form the 'CAS - Alibaba Quantum Computing Laboratory'. As its Professor Jianwei Pan explained at the time, this has the mission to 'undertake frontier research on systems that appear the most promising in realizing the practical applications of quantum computing . . . so as to break the bottlenecks of Moore's Law and classical computing'. You can visit the website for the lab here.
Like IBM, Alibaba has now made an experimental quantum computer available online. Specifically, in March 2018 the Chinese e-business giant launched its ‘superconducting quantum computing cloud’ to provide access to an 11 qubit quantum computer. This was developed with the Chinese Academy of Sciences, and allows users to run quantum programs and download the results.
As you may anticipate, Microsoft is also keen to get in on the quantum computing action, and is working with some of the world's top academics and universities to try and make this happen. To this end, Microsoft has set up several 'Station Q' labs, such as the one located at the University of California.
A key element of Microsoft's strategy is the development of quantum computers based on 'topological qubits', which it believes will be less prone to errors (hence requring fewer final system resources to be devoted to error correction). Microsoft also believes that topological qubits will be easier to scale to commercial application. Indeed, according to May 2018 article in Computer Weekly, Microsoft’s vice-president in charge of quantum computing believes that it could have commercial quantum computers on its Azure cloud platform just five years from now.
On the software side, in December 2017 Microsoft released a preview of its quantum computing development kit. This is free to download, and includes a programming language called Q#, and a quantum computing simulator.
As you may also expect, as the world's leading producer of microprocessors, Intel is working to develop quantum computing chips. To this end, it is also hedging its bets by taking two different research approaches. One of these strands is being conducted in conjunction with the leading Dutch quantum computing pioneer QuTech. In November 17, Intel announced the delivery of a 17 qubit test chip to its partner in the Netherlands. Then, in January 2018 at CES, it further announced the delivery of a 49 qubit test quantum processor called 'Tangle Lake'.
Intel's second quantum computing reseach strand is taking place entirely inhouse, and involves the creation of processors based on a technology called 'spin qubit'. This is a significant innovation, as spin qubit chips are manufactured using Intel's traditional silicon fabrication methods. In June 2018, Intel reported that it had begun testing a 26 spin qubit chip.
Already, the qubits on Intel’s spin qubit wafers are only about 50 nanometers across, or 1/1500th the width of a human hair. This means that, maybe a decade from now, Intel could be manufacturing tiny quantum processors containing thousands or millions of qubits. Unlike conventional CPUs, these would need to be supercooled to almost absolute zero. But the potential is truely breathtaking.
D-Wave Systems is a pure-play pioneer based in Canada, and way back in 2007 demonstrated a 16 qubit quantum computer. In 2011, it then sold a $10 million dollar, 128 qubit machine called the D-Wave One to Lockheed Martin. In 2013, D-Wave next sold a 512 Qubit D-Wave Two to NASA and Google. By 2015, D-Wave even broke the 1,000 qubit barrier with its D-Wave 2X, and in January 2017 sold its first 2,000 qubit D-Wave 2000Q to cyber security firm Temporal Defense Systems.
Reading the above list of achievements, you may have concluded that D-Wave has to be the world's leading quantum computing pioneer. It is, after all, the only company ever to sell a quantum computer. However, notwithstanding all of the aforementioned milestones, D-Wave’s work remains controversial. This is because their hardware is based on an 'adiabatic' process called ‘quantum annealing’ that other pioneers have dismissed as ‘restrictive’ and ‘a dead end’. IBM, for example, uses a ‘gate-based’ approach to quantum computing that allows it to control qubits in a manner analygous to the manner in which a transistor controls the flow of electrons in a conventional microprocessor. But in a D-Wave system their is no such control.
Instead, a D-Wave quantum computer takes advantage of the fact that all physical systems tend toward minimum energy states. So, for example, if you make a cup of tea and leave it standing, when you come back it will be cold as it will have declined to a more minimal energy state. The qubits in a D-Wave system also do this, and so what D-Wave does is to use its hardware to solve optimization problems that can be expressed as ‘energy minimization problems’. This is indeed restrictive, but still allows the hardware to run certain algorithms far faster than a classical computer. You can view a great video in which D-Wave explain their approach to quantum computing here.
In August 2016, this paper in Physical Review X reported that certain algorithms ran up to one hundred million times faster on a D-Wave 2X than on a single-core classical processor. One of the authors of this research also happened to be Google’s Director of Engineering. This all said, the jury remains out on the value of D-Wave's work to the general development of our quantum computing future.
Another quantum computing pure-play is a start-up called Rigetti. The company already has over 120 employees, and has made a 19 qubit quantum computer available online through its developer environment called Forest.
Another quantum computing start-up is Quantum Circuits, which was established by leading quantum computing professor Robert Schoelkopf and other colleages from Yale University. The company has raised $18 million of venture capital, and plans to beat the computing industry giants in the race to make a viable quantum computer.
Finally on this list of quantum computing pioneers, we have pure-play IonQ. The company is developing quantum computing based on a 'trapped ions' approach, which it argues 'combines unmatched physical performance, perfect qubit replication, optical networkability, and highly-optimized algorithms' in order to 'create a quantum computer that is as scalable as it is powerful and that will support a broad array of applications across a variety of industries'.
QUANTUM COMPUTING USERS AND APPLICATIONS
As noted at the start of this article, anticipated applications for quantum computing include molecular and material modelling, logistics optimization, financial modelling, cryptography, and pattern matching activities such as deep learning artificial intelligence. Already some large businesses are also actively reseaching exactly what quantum computing may do for their research and development, their products and services, and thier bottom line. As so below I just a few examples.
Daimler is working with both IBM and with Google to investigate how quantum computers may be used in logistics to help optimize vehicle delivery routes, or the flow of parts through factories. The company is also researching how quantum computers could be used to simulate chemical structures and reactions inside batteries, and so assist in the improvement of electric vehicles.
Also in the automotive sector, Volkswagen has been working with both Google and D-Wave Systems to see how quantum computers may assist with traffic flow optimization problems, as well as to help it develop better batteries.
Over in the financial sector, JPMorgan is working with IBM to explore how quantum computers may assist with trading strategies, portfolio optimization, asset pricing and risk analysis. Similarly Barclays is also participating in the IBM Q Network to investigate if quantum computers could be used to optimize the settlement of large batches of financial transactions.
As already noted, in 2011 aerospace giant Lockheed Martin was the first purchaser of a quantum computer manufactured by D-Wave Systems, and has continued to investigate use the technology for applications including air traffic management and system verification. Airbus is similarly investigating how quantum computers could speed up its research activities, and has invested in the quantum computing software company QC Ware.
Meanwhile, Accenture Labs, biotech innovator Biogen and quantum software company 1QBit are researching how drug discovery could be accelerated by using quantum computiers to make molecular comparisons. Highlighting the possibilities, in September 2017, IBM used its 7 qubit hardware to simulate the structure of a three-atom beryllium hydride molecule. In October 2017, Google and Rigetti also announced OpenFermion, which is software for running chemical simulations on a quantum computer.
OUR QUANTUM FUTURE
As this article has hopefully demonstrated, quantum computing is fairly rapidly morphing from fantasy to reality. Indeed, it is now reasonable to suggest that sometime in the late 2020s there will be quantum supercomputers available from the cloud that will find very practical and cost-effective application. Indeed, it is perfectly possible that, ten years from now, major web search and cloud AI services will be leveraging the power of quantum computers, and that most users will none the wiser to this fact.
Before the above comes to pass, quantum computers will have to pass the point of 'quantum supremacy'. Or, in other words, quantum computers will need to be proven to be better than classical computers at completing at least some computing tasks.
The concept of quantum supremacy was first coined in 2012 by John Preskill in an excellent paper that you can access here. There is also a great follow-up by John available here that discussess how it will be 'noisy intermediate-scale quantum (NISQ)' quantum hardware that will drive the next phase of the computing revolution. There are also some good, shorter articles on the subject to found in the March 2018 edition of the MIT Technology Review, and in a November 2017 edition of Nature.
Finally, for those wishing to read more, but not wishing to click on everything hyperlinked above(!), here is selected top ten sources for more information: