Depending on who you ask, the advent of large-scale quantum computers may be anywhere from five years to several decades away. Companies like Google and IBM are currently locked in a race to be the first to achieve quantum supremacy—that is, make a quantum computing chip that can outperform the most powerful classical computers—but once these devices come online, the next major challenge will be networking them effectively.
This so-called “quantum internet” already exists in its infancy. A handful of governments, financial and research institutions have had small internal networks for routing information between experimental quantum computers for years, and earlier this year China successfully routed quantum information through a satellite in low Earth orbit. The quantum internet will offer a number of advantages over its classical predecessor, such as perfectly secure data exchange and far more efficient data processing.
In currently existing quantum networks, information is being routed between the same types of quantum nodes, which transmit, store, and process qubits as they are routed through the network. Today, researchers at the Institute of Photonic Sciences (ICFO) in Spain announced that they managed to transfer information between two different types of quantum nodes in their laboratory, overcoming a significant hurdle on the route to a quantum internet.
Unlike normal computers, which store information in binary bits (either as a 1 or a 0), quantum computers traffic in qubits, which store information as a 1, a 0, or a superposition of both at the same time. This information is generally encoded in a particle of light—called a photon—and in the case of the ICFO experiment, the qubit was encoded in the photon using a technique called time-bin encoding.
This requires a photon to pass through an interferometer—a device used for superimposing light waves—from which it’s guided onto either a short or long path of optical fiber. When it emerges on the other side, the photon is encoded with a superposition of the times it takes to traverse the long and short paths—in short, a time qubit. According to the ICFO researchers, this type of encoding makes the qubit particularly robust to decoherence, or the destruction of the information contained in the qubit.
Although photons are the informational medium par excellence in quantum networks, quantum nodes can be achieved using a variety of different materials, each of which has its own strengths. For instance, according to the ICFO researchers, a node made from a cloud of laser-cooled rubidium atoms is an ideal medium for generating qubits and encoding them in photons. A node made from a crystal infused with small amounts of praseodymium ions, on the other hand, is better for storing qubits for “long” periods of time (in the quantum realm, this would be measured in microseconds).
A quantum network would benefit by being able to use different types of nodes depending on the network’s application, but figuring out how to send a photon from one type of node to another has proven difficult for researchers.
“It's like having nodes speaking in two different languages,” Nicolas Maring, a research fellow at ICFO, said in a statement. “In order for them to communicate, it is necessary to convert the single photon's properties so it can efficiently transfer all the information between these different nodes."
In their experiment, the ICFO researchers used laser-cooled rubidium atoms to generate a qubit encoded in a single photon with a very short wavelength (780 nanometers). Their device then converted this photon into the wavelength used to route non-quantum communications through optical fiber (1552 nanometers) and the photon was sent through optical fiber to an adjacent lab. At this other lab, a device developed by the scientists was used to convert the photon to a 606-nanometer wavelength so that it could be received by a crystal infused with small amounts of praseodymium ions.
The qubit was able to be sustained for 2.5 microseconds in the crystal. It’s not much, but it was long enough to retrieve the qubit with hardly any information loss. As ICFO researcher Hugues de Riedmatten told me in an email, this type of crystal could possibly be used to store qubits for as long as a minute in the future.
The ICFO team’s success is a big step forward for quantum networking. Not only did it demonstrate compatibility between two very different types of quantum nodes, it also demonstrated that quantum information can be routed between these two nodes using existing telecom fiber optic cables.
According to the team, the next step is to create larger quantum networks that consist of more than two different nodes, as well as distributing entangled photons between different nodes.
“Finding quantum nodes that can have all the required capabilities (storage, processing, etc) is extremely challenging,” Riedmatten said. “The ability to combine different systems, which could have different capabilities could solve the problem, would be a big step towards a quantum internet.”