鈥淧HYSICS should represent a reality in time and space,鈥� wrote Albert Einstein in 1947, which explains why he didn鈥檛 seriously believe in the 鈥渟pooky at a distance鈥� behaviour of the quantum world. Today, not only do we know that these strange behaviours exist, but physicists have realised that they can be exploited to build quantum computers, sensors and cryptographic systems that would outgun today鈥檚 technologies.
That鈥檚 the theory. In practice, quantum devices are proving extremely difficult to build because they must operate under noise-free conditions. But what if we could create 鈥渘oisy鈥� quantum devices? This may be possible, thanks to an obscure property of the quantum world called discord 鈥� a hot but controversial topic.
The microscopic world of photons, electrons and atoms is very different from the classical world of our everyday experience. Here the strange rules of the quantum world apply, such as superposition, where particles can exist in two places at once, or simultaneously spin clockwise and anticlockwise. It is the very act of observing the particle that fixes it in one state or the other.
Advertisement
In the 80s and early 90s, physicists like Richard Feynman, David Deutsch and Peter Shor showed that devices using quantum behaviour could carry out certain calculations far faster than any digital computers, which process information through on/off or 0/1 electrical signals. Quantum computers use a different approach, where information is in the form of quantum bits, or qubits, and the 0 and 1 exist in superposition. The qubits can share a property unique to the quantum world called entanglement, in which two quantum particles become inextricably linked no matter how far apart they are. This creates a huge capacity for processing data which would leave today鈥檚 computers for dust.
Controlling entanglement, however, is a frustrating goal. Collisions with air molecules, stray electromagnetic signals, heat and many other factors create noise, which rapidly degrades this quantum feature. A quantum algorithm that corrects for noise can preserve entanglement long enough to compute accurate results, but so far no general-purpose quantum computer has been built using more than a few qubits. A large-scale computer based on pure entanglement is unlikely to become a reality in the next decade.
What if we could build quantum devices that tolerate noise 鈥� or even exploit it? Within the last few years, this has become a real possibility. To achieve it, we need to look beyond entanglement to a quantum phenomenon called discord.
Discord was first discovered in the early 2000s by three groups working independently in the UK, US and Poland. At its basis is the realisation that quantum is not a yes/no option. A system can be fully quantum, and hence criss-crossed with entangled connections. But it can also be only partially quantum, lacking entangled connections but possessing other features of quantum theory. In essence, discord measures this quantumness, encompassing both entanglement and what was once thought of as unwanted noise. It is ubiquitous in quantum systems.
For seven years, discord remained a niche topic. It was a mathematical entity and its practical importance wasn鈥檛 clear. Worse, researchers dreamed up competing definitions, and it wasn鈥檛 clear if discord should be measured using one number or several. Similar issues dogged the effort to understand entanglement in the 90s, but for discord the conceptual problems are even greater. However, interest in discord took off five years ago when evidence began to accumulate that it can add 鈥渜uantum power鈥� to a system even when entanglement is absent. Previously it had been assumed that entanglement was a critical requirement.
A key development came in 2008 when researchers at the University of New Mexico took a . They found that as more qubits are added, the computer continues to work efficiently even as the number of data entries grows exponentially 鈥� something impossible to achieve with digital computers. Interestingly, this large improvement was achieved without a large increase in the amount of entanglement, leading to the conclusion that discord had enabled it. Remarkably, the DQC1 model works using only one qubit protected from noise, while all its remaining qubits are fully noisy. It shows us that a whole lot of noise need not be a nuisance, but rather a resource when put together with just a little bit of clean signal.
鈥淎 whole lot of noise need not be a nuisance, but rather a resource鈥�
For our part, we have shown that discord appears to play a useful role in quantum sensors 鈥� a potential way to boost sensor precision while using less energy. These could be used to probe fragile biological specimens that are damaged by light exposure, for example. that in some types of noisy quantum sensors, where entanglement isn鈥檛 protected from noise, a boost in quantum power is possible using techniques that exploit discord.
It has been exciting to be a part of this rapidly evolving field of research, observing at first hand the emergence of a big unifying idea. showed just how widely it is being applied: from testing concepts such as Maxwell鈥檚 demon (a thought experiment about the second law of thermodynamics) to . Last year the took place at the Centre for Quantum Technologies in Singapore, bringing together 70 researchers from all over the world.
Discord, however, is a controversial topic, and some researchers are sceptical of its significance. This is largely because it has taken many years to develop a clear picture of discord as an important and usable physical quantity. There is also a critical weakness in the experiments which indicate that discord might be a useful resource. Although they show that discord is present in substantial quantities, they don鈥檛 reveal a direct role for it in quantum enhancement. The presence of discord could be coincidental. But this view is changing as more applications that utilise discord are being uncovered.
Consuming discord
One of the most exciting discoveries came last year when a team led by Mile Gu of the Centre for Quantum Technologies and Ping Koy Lam of the Australian National University in Canberra showed a . They demonstrated that the improvement in the amount of information about a secret encoded message that can be extracted using a true quantum machine is equal to the discord consumed in the process (Nature Physics, vol 8, p 671).
Some also shed a historical light on discord, showing that it might not be such a novel concept after all. In 1935, Einstein, Boris Podolsky and Nathan Rosen famously published what is known as the , a thought experiment to highlight a supposed incompleteness in quantum theory. Indeed, it was in response to this paper that Erwin Schr枚dinger coined the term entanglement. Fellow physicist Niels Bohr also published a response. While it was expressed in qualitative terms, it . Schr枚dinger was certainly correct in calling entanglement a genuine quantum feature, but we think Bohr also had a point and that discord is a genuine quantum feature too.
The once-obscure line of discord research is rapidly becoming mainstream now that there is clear evidence that noisy quantum devices will provide stepping stones to powerful quantum technologies. Discord might even play a role in our understanding of the quantum-to-classical transition that explains the emergence of our everyday experience of the real world, as well as other foundational issues in physics. We can expect to hear a lot more discord in the future.
This article appeared in print under the headline 鈥淏rave new quantum world鈥�
