What Is the Quantum Internet?

How the Quantum Internet Differs from the Classical Internet

The classical internet transmits information by copying data from node to node. When you send an email, your message is duplicated at every router along the path, with each copy faithfully reproducing the original bits. This copying is essential to how classical networks function: error correction, routing, buffering, and caching all rely on the ability to copy data freely. The trade-off is that any node along the path can read the data, which is why encryption is needed to protect sensitive information.

The quantum internet cannot copy quantum information. The no-cloning theorem guarantees that an unknown quantum state cannot be duplicated, which means quantum data cannot be copied at routers, buffered in memory by making copies, or intercepted without disturbing the original. This limitation is actually the quantum internet's greatest strength for security: any attempt to tap a quantum communication channel inevitably disturbs the transmitted quantum states, alerting the legitimate users. But it also makes building a quantum network fundamentally harder than building a classical one, because the standard networking techniques of copying, retransmitting, and error-correcting classical bits do not apply.

The quantum internet will not replace the classical internet. It will operate alongside it, handling tasks that require quantum resources: distributing entanglement for quantum key distribution, connecting quantum computers for distributed computation, linking quantum sensors for enhanced precision measurements, and enabling quantum communication protocols that have no classical equivalent. Classical channels will continue to carry the bulk of internet traffic (video, web pages, email, file transfers) because these applications have no need for quantum capabilities and would not benefit from them.

Entanglement Distribution: The Core Function

The fundamental operation of the quantum internet is distributing entangled qubit pairs between distant nodes. Once two nodes share an entangled pair, they can use it for quantum key distribution (generating a shared secret key), quantum teleportation (transferring an arbitrary quantum state), or as a resource for distributed quantum computation. The entangled pair is consumed by these applications (measurement destroys entanglement), so the network must continuously generate and distribute fresh entangled pairs.

Over short distances (up to about 100 kilometers in fiber), entanglement distribution is relatively straightforward. A source generates entangled photon pairs and sends one photon to each node through optical fiber. The photons travel at the speed of light through the fiber, and the nodes detect them using single-photon detectors. The success rate is limited by fiber losses: standard telecommunications fiber absorbs roughly 0.2 dB per kilometer at 1550 nm wavelength, meaning that after 100 kilometers, only about 1% of photons arrive. After 200 kilometers, only 0.01% arrive, making direct entanglement distribution impractical beyond this distance.

Free-space optical links, including satellite-based links, offer an alternative for long distances. Photon loss in the atmosphere is lower than in fiber for vertical paths (most of the atmosphere's thickness is in the first 10 kilometers), and satellite-to-ground links can distribute entanglement over thousands of kilometers. China's Micius satellite demonstrated entanglement distribution between ground stations separated by 1,200 kilometers in 2017. However, satellite links are intermittent (dependent on orbital passes and weather), have limited bandwidth, and require large ground station telescopes, making them complementary to rather than a replacement for fiber-based networks.

Quantum Repeaters: Extending the Range

Classical optical networks overcome fiber loss using amplifiers that boost the signal strength at regular intervals. Quantum networks cannot use amplifiers because the no-cloning theorem prevents copying (and therefore amplifying) quantum states. Instead, quantum networks use quantum repeaters, devices that extend entanglement across long distances without copying the quantum information.

A quantum repeater works through entanglement swapping. Consider two intermediate nodes, A and B, each connected to a central repeater station R by short fiber links. The source generates entangled pairs and distributes one to A-R and another to R-B. Node R now holds one qubit from each pair. R performs a Bell state measurement on its two qubits, projecting the qubits at A and B into an entangled state, even though A and B never directly interacted. This process extends entanglement from two short links (A-R and R-B) to one long link (A-B). By chaining multiple repeater stations, entanglement can be extended across arbitrary distances.

The challenge is that entanglement swapping requires quantum memory at the repeater station. The repeater must store one qubit of each pair while waiting for the other pair to be generated and distributed, a process that takes time proportional to the link length divided by the speed of light. For a 100-kilometer link, the round-trip time is about 1 millisecond, meaning the quantum memory must maintain coherence for at least this long. Current quantum memory technologies based on trapped atoms, nitrogen-vacancy centers in diamond, and rare-earth ion ensembles have demonstrated storage times from milliseconds to hours, but integrating these memories into practical repeater systems with high efficiency and fidelity remains an active engineering challenge.

First-generation quantum repeaters use entanglement swapping and distillation (purifying noisy entangled pairs into fewer, higher-quality pairs) but require two-way classical communication for each repeater stage, limiting throughput. Second-generation repeaters use quantum error correction to protect quantum states during transmission, enabling one-way operation with much higher rates. Third-generation repeaters use full quantum error correction for both transmission and storage, approaching the theoretical limits of quantum communication. The progression from first to third generation parallels the maturity of quantum error correction technology, with first-generation repeaters feasible with current or near-term technology and third-generation repeaters requiring fault-tolerant quantum processing at each node.

Applications Beyond Secure Communication

Distributed quantum computing connects multiple smaller quantum processors into a single larger system through entanglement links. If two quantum processors with 100 qubits each are connected by entanglement, they can jointly perform computations that would require a single 200-qubit processor. This modular approach to scaling quantum computing avoids the engineering challenge of building ever-larger monolithic chips. The trade-off is that operations between qubits on different processors (remote gates implemented via teleportation) are much slower and less reliable than local gates, so the algorithms and compilation must be adapted to minimize inter-processor communication.

Quantum sensor networks connect distributed quantum sensors through entanglement to achieve measurement precision beyond what independent sensors can reach. Entangled sensor networks can perform tasks like detecting gravitational waves, mapping underground structures, synchronizing distant clocks with femtosecond precision, and imaging through opaque media. A network of entangled atomic clocks, for example, could improve GPS accuracy from meters to centimeters, and entangled magnetometers could detect submarine movements or mineral deposits from the surface.

Blind quantum computing allows a user with minimal quantum capabilities (just the ability to prepare and send single qubits) to delegate a quantum computation to a remote quantum server without the server learning what computation is being performed or what the input data is. The user's input and the computation are hidden from the server through quantum cryptographic protocols enabled by the entanglement link. This would allow organizations to use cloud quantum computing services for sensitive computations without trusting the cloud provider with their data or algorithms.

Current Status and Roadmap

Researchers have proposed a staged development roadmap for the quantum internet, progressing from simple quantum key distribution networks (already deployed) through entanglement distribution networks and quantum memory networks to a full fault-tolerant quantum internet. Each stage enables new applications while building on the infrastructure of the previous stage.

Stage 1 (current): trusted-node QKD networks, where quantum keys are generated between adjacent nodes and relayed through trusted intermediate nodes. Deployed in China (2,000 km backbone), Europe (multiple metropolitan networks), and Asia (South Korea, Japan, Singapore). Security depends on trusting the intermediate nodes.

Stage 2 (near-term): entanglement distribution networks without quantum memory, enabling device-independent QKD and basic quantum teleportation over metropolitan distances. Several research groups have demonstrated entanglement distribution over 50 to 100 km fiber links.

Stage 3 (medium-term): quantum repeater networks with quantum memory, extending entanglement to hundreds or thousands of kilometers without trusted nodes. Prototype quantum memories have been demonstrated, but integrating them into functioning repeater systems is ongoing work.

Stage 4 (long-term): a full quantum internet with fault-tolerant quantum processing at each node, supporting distributed quantum computing, blind quantum computing, and arbitrary quantum communication protocols. This stage requires the same fault-tolerant quantum computing technology that is the goal of the broader quantum computing field.