Building quantum supercomputers: Scientists connect two quantum processors using existing fiber optic cables for the first time

Scientists in the U.K. have successfully connected two separate quantum processors, paving the way for a quantum internet and, potentially, quantum supercomputers.

Increasing the number of quantum bits (otherwise known as qubits) in a quantum computer has proven challenging, as quantum computers are "noisy" — they are sensitive to any interference from heat, movement or electromagnetism and fail much more often than bits in classical computing.

The more qubits there are in a quantum computer, the more complex the system becomes and the greater the risk of decoherence — the loss of quantum information — and the resources needed to prevent errors. That's why scientists are focusing on building more reliable qubits before scaling systems up to the millions of qubits needed for a genuinely useful quantum computer.

In a study published Feb. 5 in the journal Nature, scientists proposed working around this scalability problem by connecting separate quantum processors together using existing fiber optic cabling, thereby increasing the number of available qubits.

This is an important step in demonstrating the feasibility of distributed quantum computing (DQC), whereby quantum processors are connected together to perform calculations. DQC would enable multiple quantum processors to work together to solve increasingly complex problems in far less time than it would take classical supercomputers.

The scientists described how they connected two quantum processors – called Alice and Bob (not to be confused with the quantum computing company Alice & Bob) using a photonic network interface (optical fibers). Sending quantum algorithms across the photonic network interface allowed the two quantum processors to share resources and operate as a single entity.

Distributed computing of the future

By connecting the two processors like this, the scientists could also transmit photons, together with quantum information and, for the first time, a quantum algorithm. Such algorithms are the computational functions that enable quantum computers to solve problems. These were shared by exploiting the phenomenon of quantum entanglement between photons.

The quantum processors could also work together on the test problem using the Grover search algorithm — a quantum algorithm that is designed to find a “needle in a haystack”; searching for a certain piece of information in a large pool of unsorted data.

This breakthrough is key to cracking the scalability problem in quantum computing. Instead of a single machine containing millions of qubits, which would be massive and unwieldy, the new technique allows for computations distributed across many smaller processors. Using small modules of trapped-ion qubits linked by optical cables, it allows qubits in separate QPUs to be entangled.

An additional benefit of connecting processors in a DQC system is ease of maintenance, as modules can be upgraded or replaced without disrupting the rest of the system.

As there was only a 6.6 feet (2 meters) gap between the two quantum processing units (QPUs), future trials of this technology would need to expand the operating distance to ensure the connection remains stable over much longer distances. Quantum repeaters, which increase the range over which quantum information can be transmitted, may also be incorporated into future systems.

Adding more quantum processors would provide further proof that DQC would be a viable solution for building quantum supercomputers. In much the same way that today’s supercomputers are hundreds of classical processors connected together, it is theoretically possible to create a quantum supercomputer by linking quantum processors together over vast distances.

As a proof of concept, the experiment proved that DQC is viable. It also creates the foundations for a secure quantum internet, which could allow for a more secure method of transmitting information, as quantum processors in different locations could be used to build a secure communications network.

In a statement, David Lucas, the principal investigator of the research team and lead scientist for the UK Quantum Computing and Simulation Hub, said the team’s "experiment demonstrates that network-distributed quantum information processing is feasible with current technology."

However, Lucas admitted there's plenty of work to be done before quantum computers are available for practical applications.

"Scaling up quantum computers remains a formidable technical challenge that will likely require new physics insights as well as intensive engineering effort over the coming years," he said.