How close are we to achieving quantum networks?

“If this is true,” said AT&T Communications’ new CEO John Donovan, “which it is, it will entirely change the way our network computing is done, and it isn’t processing but quantum networking that changes the game for us.”

It was November 2017 and Donovan was speaking to an awed conference audience about the future potential of quantum communication and networking. His company had provided seed funding for INQNET (INtelligent Quantum NEtworks and Technologies) – a research programme launched by the California Institute of Technology as part of the Alliance for Quantum Technologies (AQT) consortium.

One of the purposes of the part-publicly funded Big Science initiative is to bring together academia, national labs and industry to accelerate the development of quantum communications and networks in anticipation of quantum computing.

A quantum network will distribute information encoded into quantum states and systems. This information is held at the sub-atomic level in a physical property of something – a photon or electron – and transmitted. Whatever the entity might be, it is its quantum-mechanical properties that are being used rather than the classical properties, as happens now.

“The quantum network has been theoretically shown – it is not a belief but the task of making something that exists on paper a reality,” says Neil Sinclair, an INQNET postdoctoral fellow in physics at CalTech. “Quantum particles hold information, and if we can distribute them then we can potentially overcome problems that have never been solved before.”

Unlike in classical networks, connecting quantum devices means they become more powerful. For example, a quantum chain of computers, in theory, would have exponentially more power than one working alone. The same principle theoretically applies for quantum sensors. These entangled processors would then be the building blocks for the quantum internet, though what this will be in practice is yet to be formulated.

“Future quantum processors have physical limitations due to size and because they are sensitive to the outside world; some of the leading candidates require millikelvin temperatures or manipulation of individual atoms,” explains Sinclair. “By building a network and connecting quantum devices we expand their usefulness.”

Versions of these processors have already arrived. In November, IBM unveiled a 20-qubit [quantum bit] processor that can be used through its public cloud. Parts of the computer had to be cooled to temperatures colder than space, which conveys the epic engineering expertise required to build it. IBM has also constructed a 50-qubit operational prototype.

The last four to five years have seen a significant increase in quantum research and development, driven largely by investments from the major tech firms like IBM, Microsoft and Google, as well as government-backed initiatives. Previously, work was focused on smaller collaborations or single academics tackling individual quantum problems.

However, governments and companies are now in a race to produce meaningful quantum technologies. Google and Microsoft are said to be on the verge of big announcements. In November, Todd Holmdahl, head of Microsoft’s quantum team, said the company was “imminently close” to a big release.

The promise is that the next generation of these processors will help us model things that are impossible to understand today.  

“For example, we cannot model the fundamental interactions exhaustively for the influenza virus,” says IBM’s Leigh Chase. “We cannot create every potential permutation of some problems using only classical mechanics, whereas quantum computing opens up this possibility.”

These devices could also help physicists simulate other sub-atomic systems, as well as be used for advanced computational chemistry and modelling of some N-body problems that cannot be done today.

However, without the supporting communication infrastructure they will arrive and hit a bottleneck, says Sinclair.

For the average observer, talk of sub-atomic systems is hard to understand, let alone visualise or be convinced by. The quantum network is no different.

“It is hard to imagine,” says Rishi Pravahan, who leads the quantum work at the AT&T Foundry. “Today, we cannot fathom the full possibilities, just as it was difficult to comprehend the different uses of the internet when building computers in the late 1950s.”

However, the three main potential purposes of a quantum network are: for cryptographic functions, to build sensor webs and for distributed computing.

Cryptography, one of the best-known examples of which is quantum key distribution using entanglement, has so far received the most funding and research.

Quantum entanglement is when two quantum states are entwined together so that wherever they are, no matter the distance apart, a measurement of the state of one causes the state of the other to be modified instantaneously.

Key distribution exploits the quantum properties of photons to generate a key, which is a random sequence of 1s and 0s shared only by the sender and receiver, usually referred to as Alice and Bob. Once the key is generated, no one else will have it.

Via entanglement Alice and Bob can communicate in a channel. If someone wants to eavesdrop on them they have to make a measurement on one of the photons. However, doing so will disturb the system and the presence of the eavesdropper will be easily detected.

The challenge is to distribute the photons from one point to another without destroying the ‘quantumness’.

At the end of January, led by physicists at the University of Science and Technology of China (USTC) as part of the Quantum Experiments at Space Scale project, a 75-minute quantum-encrypted video conference call was conducted between Asia and Europe. The call used the Chinese satellite Micius and reached a distance of 7,600km – the longest distance quantum encrypted messages have ever been sent.

In theory, similar principles of entanglement for quantum encryption can be used to entwine quantum computers.

“Using entanglement we can create these keys and use them for quantum cryptography, but if we have quantum computers these same systems can be used to entangle our quantum processors, which allows them to move quantum information from one device to the next,” says Sinclair.  

“That means distributing entanglement through the nodes,” Pravahan elaborates. “If you think of a network as a graph with edges between two nodes, the edges will be the network, fibres, lasers, and the nodes would be quantum devices.”

Distributing quantum information between remote quantum processors has never been done before. Transferring entangled quantum states of light and interacting them with superconducting quantum computers or their early forms is very challenging because the processors are isolated from the environment; if there is any outside light or other magnetic fields they will fail.

In the coming year INQNET will start by distributing quantum entangled states over a lab bench, scaling up to remote locations within the Chicago area over fibre optics to multiple users.

These signals and states are highly sensitive, corruptible and experience loss and die out over long distances.

That is partly why to complete the lengthy video call successfully, USTC had to convert quantum information back into classical information (voltages and currents) at several points in the network, making it hackable, and then back again.

To transfer information over very large distances, a device akin to a quantum repeater is needed, which would be more or less the quantum version of what is required for analogue phones.

“A repeater is a small quantum computer that has real quantum resources, memories and logic elements for computational tasks,” says Jungsang Kim, professor in the Department of Electrical and Computer Engineering at Duke University, in the USA.

Although this is a very active area of research no one has ever demonstrated the technology due to this loss.

Unlike a regular signal, that if diminished can be amplified, quantum states vanish. A repeater should eventually allow systems to overcome losses.

Furthermore, to create a network that routes by signal, something akin to a packet router – a large processor that reads packets of information at high speed in the classical world – is required. Only very advanced quantum computers could do this, says Kim.

Eventually, the researchers at INQNET want to work with those who are developing quantum technologies and other elements of necessary infrastructure, says Sinclair, to try to interface with the network to build and scale in a coherent and cohesive way. The first network experiments they are doing will be at Fermilab, in Illinois. They will build a network first the size of a table and then of a room, to finally distribute information over distant locations.

A known concern about quantum technology is that it could render classical computer encryption useless. Today’s communication infrastructure passes digital information through fibre-optic pathways and wireless airwaves, using encryption to prevent eavesdroppers from reading the content of the message. The encryption is mathematically complex to unravel, but in a quantum-enabled world the codes would be considerably easier to crack.

“The countries that have quantum networks will have more secure communications than those that don’t; some might consider that to be a destabilising factor in society,” says Rodney van Meter from the Joint Centre for Quantum Information and Computer Science at Keio University, Japan, and author of ‘Quantum Networking’.

However, this is not an immediate issue. “The idea that quantum computers are going to break encryption in minutes or seconds is wrong,” he adds. “Building a quantum device to do this encryption is actually really hard, and even when we succeed it is not something that is going to happen within minutes; it is going to take months, years or decades depending on different factors.”

Donovan, in his talk at Web Summit about the future promises of the network, said: “If I want to take a mobile call with you I don’t need a phone – if we are quantumly entangled I can just think it and we can have a conversation... we will have a channel that will always be available to send data back and forth at super-high speeds, transformatively.”

He was referring to the somewhat sci-fi-esque concept of mixing the quantum with the biological. Nobel Prize-winning physicist Erwin Schrödinger first floated this idea around 75 years ago and there has been some research to suggest it could be theoretically possible. However, Pravahan says Donovan was being “futuristic” when he said this.

Sinclair is also doubtful, pointing out that “it is really challenging to interface with these quantum systems, let alone in the hot, living, biological environment”.

‘We are at a point where we cannot shrink chips any more, so sooner or later we need to move to the quantum realm. Any company not doing this is going to be left behind.’

A powerful quantum network of computers, however, could one day simulate the dynamics and environment of complicated biological systems at the quantum level to further medical research and biological and quantum entanglement.

Sinclair adds: “Some people are thinking about this and [eventually] we could have dedicated groups of quantum biologists that are pioneering in this area – but that is 50-plus years away.”

A topic once confined to obscure corners of academia is now the obsession of tech giants, conglomerates and China, so does this mean quantum technology is now inevitable?

“Absolutely,” say Pravahan, “We are at a point where we cannot shrink chips any more, so sooner or later we need to move to the quantum realm. Any company not doing this is going to be left behind.”

Others are far more modest, though not completely sceptical. When asked about his thoughts on quantum networking, Chintan Patel, a senior strategist at Cisco, said: “It is a long, long-term play and I don’t think we will see anything develop in decades, but it is important for industry to keep an eye on these things.”

Setting a timeline is difficult because there are too many factors and unknowns and many technologies that need to be discovered, says Pravahan. But, pushed for an answer, he says it’s conceivable the quantum internet could be here in 10-20 years, though it will never fully replace the classical systems we have today. “I am confident in saying that quantum devices are really good at solving certain problems, but we will need classical communication infrastructure to serve as the backbone for the quantum network,” he adds.

However, progress is still uncertain. “There is a lot of potential for quantum to understand the nature of reality and how we work, but the truth is, right now, there are finite resources and help that we have to explore these very basic features of nature, as well as the possibility of using them to improve lives,” says Sinclair.

“We hope to demonstrate quantum networking to show we are at the forefront of something potentially transformative and cumulative and if we do this we can continue to find new applications.”