A possible quantum internet by 2030, explained

Quantum computers today operate as isolated machines, each running inside a sealed environment with no way to link to another system like classical machines over the internet. IBM and Cisco want to change that. The two companies have set a long-term plan to connect quantum computers over long distances, with the hope of proving that the idea is workable before the end of 2030.

Classical computers share information easily over global networks but a similar foundation does not yet exist for quantum machines. Building a network for them is far more complex, requiring hardware, software, and research, and both companies say universities and federal labs will be needed to move the work forward.

The company said the long-term plan “could lay the groundwork for a future quantum computing internet.” The partnership with Cisco is not new; the companies have been working together since 1999, and Cisco has deep experience in networking research.

Why the network matters

Quantum computers promise to solve problems in fields like chemistry, physics, and security, tasks that could take classical systems thousands of years. But quantum computers are still early-stage machines, often producing errors. IBM, Google, and others are racing to build systems that can operate reliably. IBM hopes to have a fully functional, error-corrected machine by 2029.

The joint plan calls for a proof-of-concept quantum network in the next three to five years. After that, the partners want to build technologies that can connect more machines across longer distances through the early 2030s. The work ties into IBM’s broader quantum roadmap, which lays out steps toward more capable systems.

IBM said this kind of network could one day support problems “requiring up to trillions of quantum operations – allowing industries to pursue tasks like massive optimisation challenges or the design of new medicines and materials.”

Turning stationary qubits into “flying” qubits

Cisco has a lab studying how to link quantum machines, but the work starts with a basic issue: IBM’s quantum computers sit inside large cryogenic systems, cooled to temperatures where atoms barely move. The information inside them is stored in stationary qubits. To connect these machines over a network, that information must be converted into “flying” qubits. Jay Gambetta, director of IBM Research and an IBM fellow, described these as qubits that move through space rather than sit still.

Proving the concept is only the first step. The companies then need to show they can scale the system beyond linking two machines. The goal is to connect many quantum computers located in different places and stretch those connections across far longer distances. Doing this will require new microwave-optical transducers and optical-photon technologies, Cisco said.

Building the hardware and software needed

The project also needs a new type of interface for quantum computers. IBM is building a quantum networking unit, called a QNU, which will act as the link between quantum processors, or QPUs. The unit must transform stationary quantum information into flying quantum information that can be sent to other machines.

On the software side, Cisco is developing a high-speed framework that can reconfigure the network as computations run. The idea is to deliver entanglements to QNUs once each machine finishes its part of a calculation, allowing the network to pass quantum information between systems without breaking the process.

Both companies are also studying how a “network bridge” might work. The bridge would use multiple Cisco quantum network nodes that could connect hundreds of IBM QPUs through the QNU interface. The initial goal is to work inside a single data centre. To move quantum information between distant sites, microwave qubits will have to be converted into optical signals that can travel through fibre-optic cables via switches. The device that enables this is the microwave-optical transducer, but the technology does not exist at the quality or scale needed, so IBM and Cisco will work with research groups, including the Superconducting Quantum Materials and Systems Center at the Fermi National Accelerator Laboratory near Chicago.

As the project progresses, the companies plan to publish open-source software that ties the pieces together and allows researchers to build on their work.

A joint vision for a full quantum system

“We are looking at this end-to-end as a system … rather than two discrete road maps,” said Vijoy Pandey, senior vice-president of Cisco’s Outshift innovation incubator. “We are solving it jointly, which has a much better chance of this thing going in the same direction.”

He added: “IBM is building quantum computers with aggressive road maps for scale-up, and we are bringing quantum networking that enables scale-out. Together, we are solving this as a complete system problem including the hardware to connect quantum computers, the software to run computations in them, and the networking intelligence that makes it work.”

If the architecture comes together, IBM’s quantum computers could one day handle workloads so large that all the classical computers in the world working together could not match. IBM believes the network could support runtimes with trillions of quantum gates, the basic operations that make quantum applications possible. That scale could open the door to new methods for designing medicines, materials, or other advanced systems.

There are still major uncertainties. The companies admit the plan depends on a series of breakthroughs that may or may not arrive on schedule. But they believe a broad quantum internet made up of thousands of distributed systems is possible by the late 2030s. A network of that size could enable quantum sensing, quantum communications, and other tools that share information in global distances.

IBM also sees potential practical uses. For example, quantum sensor networks could pick up tiny shifts in climate or weather that classical tools miss. But none of this can happen until today’s noisy and error-prone quantum computers are refined.

The race toward fault-tolerant quantum machines

IBM has been open about its goal to build a fault-tolerant machine: not one that never encounters errors, but one that keeps working even when errors occur. Last month, scientists showed they could run a quantum error-correction algorithm on AMD classical hardware. This is the kind of work that helps researchers understand how to manage errors until quantum systems grow strong enough to handle them on their own.

Classical computers have long been fault-tolerant because they follow design patterns that allow them to keep running even when something goes wrong, but quantum computers are far more delicate. Their qubits are sensitive to outside interference, which makes them prone to mistakes. The aim is not to build a flawless machine but to build one that can keep operating despite disturbances in the operating environment.

Pure-play quantum companies often get the spotlight, but IBM has been involved in the field for decades. It currently runs the largest fleet of quantum computers in the world, and has more than 25 systems with over 100 logical qubits. Its long-term goal is to bring them together in a network that could one day support computations far beyond the reach of today’s tools.

(Photo by JJ Ying)

Want to discover how IoT is transforming telecoms and connectivity? Join the IoT Tech Expo in Amsterdam, California, and London. Explore how innovations in 5G, edge computing, and IoT are shaping the future of networks and services. The event is part of TechEx and co-located with other leading technology conferences. Click here for more information.

Telecoms News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

Tags: cisco, connectivity, hardware, internet, networks, wireless