What does quantum computing mean to project delivery? How might this nascent technology change the way we do business?
Let’s start with an observation: we have no lack of computing power today. There are few project problems that are constrained by the available processing power. Powerful CPUs and GPUs (such as those used for bitcoin mining) power our machine-learning and natural language systems. They drive the recommendation engines used in online commerce and social media sites. At present we are data-constrained. The search for good data to improve project delivery is an active and current field of research.
However, as the volumes of accessible quality data grow, this will change. Our ability to exploit these enormous volumes of data to improve project delivery may fall behind. How might the quantum computer change the picture? First, we need to appreciate - at a very high level - how quantum computers work, and how they differ from mainstream computers.
A conventional computer is built in silicon. At its heart beats a central processing unit (CPU), executing billions of instructions every second. Billions of tiny transistor circuits etched in silicon manipulate, through logic gates, numbers at high speed: adding, subtracting, multiplying and dividing.
Their speed gives an illusion of a parallel system, propelling our web browsers, email clients, and project scheduling software with ease. Fundamentally, a conventional computer is akin to a high-speed abacus, performing calculations sequentially.
A quantum computer is very different: it performs calculations in parallel. A quantum computer cleverly allows us to peer into, observe and – increasingly - manipulate the quantum realm. This is no mean feat, and is a triumph of modern physics and engineering. As the famous physicist Richard Feynman once said: “If you think you understand quantum mechanics, then you don’t understand quantum mechanics”.
Quantum computers are based upon qubits. These are atomic-scale structures that allow us to bridge the gap between classical and quantum realms. Qubits perform calculations that combine quantum effects across the set of qubits that make up the computer. This ghostly interaction -entanglement and superposition of quantum waves - is analogous to the famous “is it dead or is it alive?” Schrodinger’s cat. Because they can operate in parallel, there are particular types of problem quantum computers can solve far quicker than digital computers.
Initial excitement centred around the potential for quantum algorithms to break modern encryption (Shor, 1994), which takes inordinate amounts of time on a conventional computer. Since then, a wide range of other algorithms Wikipedia have been developed, including extremely fast unstructured database searches, such as Grover’s.
The race is now on to develop practical quantum computers for widescale use.
Many tech companies are entering the quantum business. Companies such as IBM, D-Wave, Google and Honeywell are developing quantum computers. The sector is attracting start-ups and venture capital. The race is on to overcome the engineering challenges of the quantum realm. In many ways this is similar to the race for peaceful nuclear fusion.
Quantum computers are very delicate. They must be cooled to extremely low temperatures to avoid upsetting the delicate quantum mechanics. The particular challenge of decoherence limits the time for which the quantum algorithms can operate. Error correction and fault tolerance are key.
There are many other practical challenges. The one I find strangest of all – and a beautiful irony - is that owing to their fundamental nature and complexity, we cannot efficiently simulate (ie model) in a digital computer the behaviour of a modern quantum computer before building one!
It’s unlikely that our home desktop computers will have quantum processors in the near future. Instead, we will see the large tech and cloud providers take the first step through hybrid systems. These will combine the power of traditional CPUs and GPUs, with quantum processors running the algorithms that yield the greatest performance gain. I believe that classic “digital” computers are here to stay, given the huge range of optimised algorithms and an ever-growing proven and re-usable code base.
Quantum Thinking in Projects
Today it’s hard to make a firm prediction for the specific applications of quantum computing in project delivery. The algorithms show great promise, but the practical implementation needs to develop further before we can see their strengths.
Regardless, I am excited and optimistic about quantum computing, because of the mindset change it will bring. Why is this?
Projects are about change. Making sense of the world, and predicting its evolution (and thereby being able to influence this) is central to our profession. Science not only explains the world around us. It enables us to make predictions using the governing “equations of motion”, whatever shape they take. We as project delivery professionals know just how complex those equations of motion are, not least with our stakeholders. Owing to this uncertainty, we deal in probabilities, and can rarely promise the absolutes.
Fortuitously, probability is the central language of quantum computing. As we develop systems powered by quantum algorithms, we’ll get better at using the language of probability. Perhaps the worlds of project delivery and quantum mechanics have more in common than we realise.
As they say, watch this (quantum) space!