Quantum Networking: Building the Internet's Next Layer

A quantum internet would not be a faster internet. It would be a different one — a network that carries quantum states between machines, enabling things classical links cannot: encryption keys whose interception is physically detectable, quantum computers pooled into one larger machine, and sensors synchronised with precision beyond any classical clock. The pieces exist in laboratories and a few metro-scale testbeds. Connecting them at scale runs into a problem that has no classical equivalent.

Why you can't just amplify a qubit

Classical networks work because signals can be copied: a repeater reads a weak signal and retransmits a strong one. Quantum information can't be copied — the no-cloning theorem forbids it — and it can't be measured without disturbing it. So a photon carrying a qubit down an optical fibre simply gets lost with distance (roughly half are gone every 10–20 km), and there's no amplifier to save it. Direct quantum links top out at a few hundred kilometres in fibre. Everything interesting about quantum networking is about getting past that limit.

Quantum key distribution: the application that ships

QKD lets two parties generate a shared secret key by sending qubits (usually photon polarisations or phases) and comparing measurement results. Its selling point is that an eavesdropper intercepting the photons necessarily disturbs them, which shows up as an error rate the parties can detect. Security rests on physics rather than on the assumed hardness of a maths problem — which is what makes it attractive against future quantum computers.

QKD has been commercially available for two decades and is deployed on metropolitan fibre networks in China, Europe, Korea and elsewhere; China's Micius satellite demonstrated intercontinental QKD in 2017. But it has real limitations. It needs dedicated fibre or line-of-sight; the "trusted node" relays used to extend range are themselves points of vulnerability; the equipment is expensive and has had implementation flaws; and it only distributes keys — it doesn't authenticate the parties, so it still needs classical cryptography. For these reasons the NSA, NCSC and other national security agencies have said they don't endorse QKD for government use and recommend post-quantum cryptography instead. QKD is real technology with a narrow niche, not the future of encryption.

Entanglement distribution: the real goal

The more powerful vision is a network that distributes entangled pairs between distant nodes. With shared entanglement, two nodes can teleport a qubit state from one to the other (using a classical message alongside), run distributed quantum algorithms, and perform "blind" computation on a remote quantum computer without revealing the input. This is what "quantum internet" properly means.

Extending entanglement beyond direct-link range needs a quantum repeater: a node that holds one half of an entangled pair in a quantum memory long enough for the other segment to succeed, then performs an entanglement swap that links the outer ends. Each piece — memories that hold states for milliseconds to seconds, swapping with high fidelity, error correction across links — has been demonstrated individually. Chaining them reliably over real distances is the frontier. Results in the past few years include entanglement between quantum memories over tens of kilometres of deployed fibre in the Netherlands, Germany, China and the US, and the first small multi-node networks with entanglement swapping.

The hardware

Quantum network nodes are built from the same platforms as quantum computers — nitrogen-vacancy centres in diamond, trapped ions, neutral atoms, superconducting circuits — coupled to photons that travel down fibre. Converting each platform's native photon wavelength to the telecom band (around 1550 nm) where fibre loss is lowest is a key engineering step. For satellite links, free-space optical channels avoid fibre loss altogether, and several countries are launching quantum communication satellites for both QKD and entanglement distribution.

What it would enable

• Modular quantum computing: linking several modest quantum processors into one large fault-tolerant machine, which most vendors' roadmaps assume.
• Secure delegated computation: using a remote quantum computer without it learning what you computed.
• Quantum sensor networks: entangled clocks and telescopes with sensitivity beyond classical limits — arrays that could improve GPS-free navigation and astronomical imaging.
• Device-independent QKD whose security doesn't depend on trusting the hardware.

Timeline, honestly

Metropolitan QKD networks: deployed now, niche use. Entanglement between a few nodes over tens of kilometres: demonstrated. Reliable quantum repeaters enabling continental entanglement: a research goal for the 2030s. A general-purpose quantum internet: further. National programmes — the EU's Quantum Internet Alliance, the US Quantum Internet Blueprint, China's national backbone — are funding the intermediate steps. The technology is real and progressing; the timelines in vendor press releases are not.