Quantum computing: a leap forward

Opinions expressed whether in general or in both on the performance of individual investments and in a wider economic context represent the views of the contributor at the time of preparation.

Executive summary: The world needs more computing power – of a new and improved sort – since current computers are unable to handle the growing data deluge. Yet a solution is at hand: quantum computing constitutes a leap forward, providing computing power at significantly faster speeds than are currently achievable. As a result, quantum computers could solve the problems with which humans and super-computers struggle today. Such benefits would have applicability for big data analysis, artificial intelligence, cybersecurity, medicine, industry and finance. Although commercial realisation may be some years away, many governments and businesses are actively investing at present. Against this background, by 2025 the industry could be worth $10bn. IBM, Google and Alibaba appear to have established a relative lead within the field. 

 We are drowning in data. What’s more, the deluge is only set to grow. Consider that in 2016 the world produced as much data as in the entire history of mankind, yet the amount of new data we produce is doubling every year (per Scientific American website). 5m new internet-connected devices are coming online every day, implying that by 2030, over 500bn ‘things’ will be connected to the Internet (per Cisco). These will all generate yet more data. However, without a breakthrough in quantum computing, there will be no meaningful application of artificial intelligence, no driverless cars, no autonomous robots, and no solutions to many of the world’s major unresolved challenges (such as demographics and mortality). 

The reason why is simply because we can’t afford the energy to run such applications. Computers today may be more efficient than ever before, but they still compare very poorly relative to the human brain. Our brains perform around 1,000 floating rate operations per second (flops) under 20 watts of energy. By contrast, the most powerful supercomputer in the world (at the Riken Institute in Japan) would perform the same exercise using 750,000 times more energy. Put another way, if our expectations are that computer processing power will reach – or possibly extend beyond – human capabilities, then we need a radical new way to approach computing. 

The good news is that quantum computers provide this very solution. Moreover, they have moved from beyond a position of theoretical research into an engineering phase, including commercial experiments. While classical computers use the laws of mathematics (binary code comprising zeroes and ones), quantum computers use the laws of physics. The concept dates from the 1940s where the rules of quantum mechanics were outlined, defining what happens at an atomic scale. In the 1980s, academics at MIT began applying these principles to computing. Since then, the number of research papers published on the subject has grown rapidly, reaching close to 6,000 in 2015. Similarly, the number of patents issued in the field is currently running at around 600 annually, with a further 900 in quantum cryptography (all statistics per the European Commission). Governments are investing in applications for space exploration, defence and medical research while the major tech companies are all building quantum research labs. Start-ups have begun to proliferate too. 

A quantum computer works by programming atoms to represent all possible input combinations simultaneously and then runs an algorithm that tests these combinations at once. By contrast, a traditional approach would involve serially cycling every possibility by varying inputs to arrive at a solution. Rather than using transistors that switch on and off the zeros and ones that represent binary digits (bits) of information, quantum computers use quantum bits (qubits) that can represent all combinations of zeros and ones simultaneously. In the sub-atomic world, the laws of physics change and quantum mechanics enables things that are hard to conceive of in our everyday world. Put another way, a quantum-mechanical object with two energy levels can occupy either of those two levels, as well as an arbitrary combination (called a superposition) of the two. This results in an infinite number of quantum states that a single qubit can take. 

Quantum systems therefore no longer consider operations (or problems, questions) as binary, but instead ‘recognise’ all the right answers immediately. The corollary is exponential acceleration – computing power at a significantly faster rate than achieved with conventional computers. Each qubit added to a computer effectively doubles its processing power, an exponential gain not possible with classical computing (where adding bits provide negligible improvements). To provide some context: whereas a standard 300-bit chip could power a basic calculator, a 300-qubit chip has the computing power of two novemvigintillion bits (a two followed by 90 zeros), a number that exceeds the atoms in the universe. Against this background, a quantum computer could be used to solve the problems with which humans and super-computers struggle today because they are exponentially complex. There are multiple uses to which this breakthrough could be applied. Four obvious examples spring immediately to mind: general AI and big data; cybersecurity; medicine and healthcare; and, industry. 

Begin with machine learning. Quantum computing systems will process exponentially more data in parallel compared to the algorithms used presently for analysing datasets, thereby radically improving classification and analysis, and hence constituting a major leap forward for AI. Next, the problem of cyberattacks and breaches in security could potentially be solved using quantum key distribution. Such a system enables two parties to produce a shared random secret key known to only one of them. In this state, the two communicating users would always be able to detect the presence of any third party trying to gain knowledge of the key. Turning to medicine, and quantum computers will have the ability to map out trillions of molecular combinations, quickly identifying promising candidates, and significantly reducing the cost and time of drug development. Quantum computing could also sequence and analyse a person’s genome much faster than at present to create personalised drugs and healthcare available on a mass-market basis. The possibility of applying quantum computing to industrial problems is equally tantalising. Scientists would be able to find new materials with superior properties which could, for example, create much higher energy-dense batteries or capture carbon from the atmosphere. 

Many businesses have already begun to experiment with the potential afforded by quantum computing. China has launched a quantum satellite which offers ‘unbreakable’ (per Xinhua state news agency) encryption over distances of up to 1,200km. Similar trials are underway in the US, UK, Switzerland and Japan. Meanwhile, Biogen is using a basic version of quantum computing to design new molecules, just as BASF is for polymer design. Likewise, Airbus, Lockheed Martin and Raytheon are reviewing the possibility that quantum computing would permit in terms of plane design, as are BP and ENI in the field of optimised oil extraction. NASA is applying quantum learnings in its considerations about future space travel. Within the field of finance, Goldman Sachs and Morgan Stanley have both stated that they see how quantum computing would have benefits in terms of portfolio optimisation. 

The major caveat, however, is that while the technology works, there are significant engineering challenges to be overcome before quantum computing can be scaled for general usage. Quantum computing systems can only operate in very controlled conditions as qubits are extremely susceptible to noise, vibrations, fluctuating electrical fields and general interference. This is because they are so fragile and hence the state of a qubit can change rapidly. The only situation in which they will not do this is in a pure state of absolute zero – i.e. -273 degrees centigrade. This requires a costly outlay. Furthermore, quantum computers need rare elements such as Niobium. Scaling these systems (in the absence of viable alternatives) may, therefore, be problematic. Even when large qubit processing becomes possible, its success may be measured by the ability to interconnect such systems with classic computers. Various hurdles need to be overcome in this respect. 

Progress is being made to address these challenges, and the emerging consensus is that commercial quantum computing systems should be available within the next five years. McKinsey estimates that around 7,000 people are directly employed in the field at present, with a combined budget of $1.5bn for quantum-technology research. National and supranational funding bodies are increasingly backing quantum efforts with the UK Government operating a £270m ($350m) programme and the EU setting aside a budget of €1bn ($1.2bn) for research across Europe over the next ten years. Against this background, consultants (including Homeland Security Research, Markets & Markets and Persistence Market Research) estimate that the quantum computing market could be worth up to $10bn by 2025. 

From an investment perspective, the quantum computing landscape has much in common with the state of the internet a generation ago. In other words, laboratory-led research is now being commercialised, with blue-chip firms buying into scientific ventures and/or developing their own research efforts. Governments are similarly making a variety of ‘strategic’ investments. Each business at the forefront of the technology currently seems to be taking a slightly different approach and there does not seem to be agreement at present on common hardware/ software/ interconnection standards. 

IBM and Google appear to be leading the field at present, with Alibaba also establishing a strong presence in China. IBM was the first business to gain quantum computing patents and has been active in the field (partnering with MIT) for over thirty years. In 2016, it launched ‘quantum experience,’ an open-source quantum computing platform available via its public cloud. Users can trial algorithms and develop programmes. As of March 2017, IBM said 40,000 users had performed 275,000 experiments. Google’s and Alibaba’s efforts are much-lower profile. Google says its quantum computing knowledge is being applied internally (it has partnered with private Canadian company D-Wave), but believes a possible commercial launch may be achievable by 2019. Alibaba has been working with the Chinese Academy of Sciences since at least 2015, but little information on its progress is publicly available. Other companies currently involved in the commercialisation of quantum computing include Fujitsu, Intel, Microsoft, Mitsubishi Electric and NEC. Telecoms operators such as BT, KPN, NTT and SK Telecom are experimenting with integrating elements of quantum computing (and key distribution) into their networks. Looking ahead, for all business, quantum computing will likely become increasingly integral, constituting a palpable leap forward.1 

Alexander Gunz, Fund Manager, Heptagon Capital