NIST sends entangled photons through 38 miles of suburban fiber in real-world quantum network test

[Taker92]

NIST researchers and collaborators have pulled off a genuinely important real world stress test for quantum networking, successfully sending entangled photons through 62 kilometers of ordinary commercial fiber optic cable strung above the streets and sidewalks of the Maryland suburbs, from the agencys Gaithersburg campus to the University of Maryland in College Park

Quantum networks depend on entanglement, the phenomenon Einstein famously called spooky action at a distance, where two particles share a unified quantum state and measuring one instantly determines the result of measuring the other regardless of the distance separating them, researchers believe networks built on this property could eventually link telescopes to produce sharper combined images, let entangled sensors detect and pinpoint earthquakes or volcanic eruptions, connect quantum computers to tackle problems too complex for any single machine, and enable ultrasecure communications where any hacking attempt would be immediately detectable

The genuinely hard technical problem is that building dedicated fiber networks purely for quantum applications would be prohibitively expensive, so the team from NIST, the Joint Quantum Institute and the New York based company Qunnect deliberately tested whether quantum networking could instead run on the existing aerial fiber infrastructure that already powers everyday internet traffic, and that infrastructure is genuinely rough on fragile quantum states, NIST physicist Oliver Slattery put it bluntly, calling it about as bad a connection as you can possibly have, since the fibers expand and contract with daily temperature swings, sway in the wind, and get landed on by birds, all of which can twist and distort the polarization that quantum information gets encoded in

To fix this in real time, the team deployed Qunnect built stabilization devices that sent reference light beams through the fiber alongside the actual entangled photons, measured how those reference beams polarization had been distorted by the fiber, then applied the exact inverse correction to keep the experimental photons properly entangled, the results were genuinely solid, they transmitted 1,500 entangled photon pairs per second and successfully distributed entangled photons 92.8 percent of the time over a full 24 hour period, needing only 7.2 percent of the time for active polarization correction

This wasnt actually a distance record, a European team already distributed entangled photons over 248 kilometers of underground fiber back in 2022, but lead author Yicheng Shi framed the real significance well, calling this a genuine stress test of quantum networking systems, and the fact that it held up in an environment this exposed and noisy is exactly the kind of validation the field needs before quantum networks can realistically piggyback on existing internet infrastructure rather than requiring an entirely new dedicated buildout

'Spooky' Particles Transit DC Suburbs, a Step Toward a Quantum Networkwww.nist.gov

[Storm]

About as bad a connection as you can possibly have is such a great quote, using genuinely hostile real world infrastructure specifically to stress test the system rather than a pristine underground lab fiber is exactly the right approach if the eventual goal is practical deployment on existing networks

[Tel92]

92.8 percent successful distribution over a full 24 hour cycle with only 7.2 percent needed for correction is a genuinely strong result given how much thermal expansion, wind and physical disturbance that aerial fiber goes through between day and night

[ShadowPilot]

Not being a distance record but still being the more important result is a good distinction, the 2022 European 248 kilometer underground result proved distance was possible under controlled conditions, this proves resilience under genuinely uncontrolled real world conditions, which matters more for actual deployment

[StringTheory95]

The potential applications list, linked telescopes, earthquake detection sensor networks, distributed quantum computing, ultrasecure communications, shows just how broadly useful a functioning quantum network could eventually be well beyond just cryptography, which tends to dominate most public conversation about quantum technology

[TheGame_Real]

Birds landing on fiber lines disrupting quantum entanglement is such a wonderfully mundane physical detail buried in genuinely cutting edge physics research, real world engineering constraints rarely get more delightfully specific than that

[Skibidi]

Reusing existing internet fiber infrastructure rather than needing a dedicated quantum only network buildout is the practical detail that actually makes this economically viable, a purpose built quantum network would be prohibitively expensive at any real scale

[BlackSunLynx]

The real time correction approach using reference light beams to measure and then invert the fibers distortion is a clever engineering solution, rather than trying to prevent the distortion entirely they just measure it continuously and cancel it out on the fly

[Puma37]

Collaborating with a private company like Qunnect for the actual stabilization hardware shows how quantum networking research is genuinely maturing from pure academic physics into applied engineering with real commercial partners building the necessary supporting technology

[Keane]

This is a great example of unglamorous but genuinely necessary infrastructure validation work, everyone gets excited about qubit counts and algorithm breakthroughs but proving the actual physical transmission layer works reliably in messy real conditions is just as critical

[Optimiser Ruby]

1,500 entangled photon pairs per second is described as needing improvement for practical use, would be curious what the actual target rate is for genuinely useful applications like secure communications or distributed quantum computing, that context would help gauge how far this still needs to scale

[Sophie92]

38 miles through actual suburban fiber is the bit that matters to me. Getting entangled photons to behave in a controlled lab is one thing; sending them through infrastructure full of bends, connectors, temperature changes and whatever mystery someone did to the cabinet last Tuesday is another. Quantum networking needs to survive the real world, not just a very expensive optical table.

[Bright Hermit]

I would be careful about reading too much into the word entangled here without looking at the measured fidelity. Producing correlated photons is one thing; demonstrating that the entanglement remains useful after traveling through noisy, lossy infrastructure is another. The quality of the shared state will determine what applications are practical. A million rubbish pairs aren't automatically better than a thousand excellent ones.

[AlexaBliss_Fan]

This also highlights why quantum repeaters matter so much. Classical networks use amplifiers to compensate for signal loss, but quantum information cannot be copied arbitrarily because of the no-cloning theorem. Repeaters that can distribute entanglement between successive links could eventually extend network reach, but building reliable ones remains a major challenge. The 38-mile result is a solid piece of the puzzle rather than the completed jigsaw.

[Sentinel68]

What I like is that this seems to be moving beyond the usual laboratory demonstration cycle. If researchers can repeatedly establish entanglement across deployed fiber under changing environmental conditions, that tells us something important about operational reliability. The next milestone should be sustained operation over days or weeks, ideally with automatic calibration and recovery. Nobody wants a quantum network that needs a physicist to visit every time it rains.

[Jenny75]

I'd like to see a comparison between this result and previous field demonstrations using the same fiber length, wavelength and detection criteria. That would show whether the team has improved the usable entanglement rate, the fidelity, the stability or some combination of all three. A record distance is nice for a press release, but reproducible performance under ordinary operating conditions is what eventually turns an experiment into infrastructure.

[Dean83]

People are going to ask when we can download quantum internet, aren't they? This is infrastructure research, not a replacement for your broadband. The point is potentially connecting quantum devices in ways that preserve entanglement, enabling applications such as distributed quantum computing and certain forms of quantum communication. Your Netflix stream will remain stubbornly classical for the foreseeable future.

[Mia86]

On the question of target rates, there probably isn't one universal number. A quantum key distribution demonstration, a network connecting quantum processors and a system intended to distribute entanglement for advanced sensing can have very different requirements. The useful rate depends on the application, the fidelity required and how long the receiving devices can preserve the quantum state. More pairs per second is great, but quality and stability count too.

[SwiftQuarry]

There's a funny mismatch between the headline and the actual achievement. Sending light down a fiber is something telecommunications engineers mastered ages ago; sending entanglement through a real deployed route while preserving enough quality to be useful is the difficult part. The quantum word doesn't make the distance itself magical. It changes what needs to survive the journey.

[Henry78]

The suburban part is genuinely encouraging. Existing fiber networks are already everywhere, so reusing them could be much cheaper than digging up streets to install an entirely separate quantum network. Of course, sharing fiber with ordinary traffic introduces complications, particularly from noise caused by classical signals. The engineering challenge is making the two coexist without the quantum channel throwing a tantrum.

[SpinState22]

A lot of the discussion focuses on photon rates, but latency and memory are going to be equally awkward. Even if you generate entangled pairs quickly, the endpoints need to use them before the relevant quantum states decohere. If the network has to wait for acknowledgements, coordinate multiple links or retry failed operations, the effective throughput could drop considerably. The whole system needs benchmarking, not just the optical source.

[Violet Tiger]

Thirty-eight miles sounds impressive until somebody points out that ordinary telecom fiber carries data thousands of miles. That's comparing apples with quantum oranges, though. Classical optical communication can amplify signals and regenerate them using established techniques; unknown quantum states can't simply be copied and amplified in the same way. Extending quantum links requires different tricks, which is why this result deserves attention.

[NightHarbour30]

Could existing telecom operators eventually run quantum services over their current networks? Potentially, yes, but probably not by plugging a quantum transmitter into a random fiber port and calling it a day. Classical signals can create noise through effects such as Raman scattering, and specialized wavelengths, filtering or dedicated fibers may be necessary. The commercial appeal is real, but the deployment details will decide whether it's economical.

[Layla17]

The exciting possibility is distributed quantum computing. Instead of forcing every qubit into one giant machine, you could eventually connect smaller quantum processors and share quantum resources across a network. That could help with modular architectures, although it introduces demanding requirements for entanglement generation, storage and error correction. This experiment is a useful step toward that vision, not proof that networked quantum computers are around the corner.

[Zatanna]

My main question is whether they tested this over a continuous stretch of fiber or through multiple connected segments with intermediate equipment. Both can be valuable, but they tell us different things about scalability. A metropolitan quantum network will need to deal with junctions, switches and eventually ways of extending entanglement beyond the reach of a single link. The details of the setup matter almost as much as the mileage.

[Karen73]

The 1,500 entangled pairs per second figure needs context before we call it too slow. What's the detection efficiency, how stable is the rate over time, and how many pairs are actually usable after filtering and quality checks? Raw pair production and usable entanglement aren't necessarily the same thing. I'd love to see those numbers alongside the headline distance.

[RonaldoNazario68]

For a practical target, I'd want the team to define the intended application first. A system that needs occasional high-fidelity entanglement for a specialized sensor might tolerate a modest rate, whereas a fault-tolerant distributed computer could require much higher rates to overcome probabilistic operations and losses. Without that context, declaring 1,500 pairs per second either fantastic or inadequate is mostly guesswork.

[Myles_56]

The long-term networking question is whether these links can be stitched together reliably. Imagine connecting three buildings across a city, then adding another link, then another. Losses and imperfect operations compound, so a network that works over one 38-mile route may need substantial improvements before it can support a larger topology. Still, real-world fiber testing is exactly how you discover the problems before deploying at scale.

[Faded Owen]

For everyone worried about the 1,500-pair rate: distance is expensive in photon land. Fiber attenuation means fewer photons survive as the path gets longer, and real-world losses can make an already difficult job harder. You can improve source brightness and detector performance, but those aren't free upgrades. Quantum networking is basically the art of getting incredibly delicate parcels delivered when the postal service is made of glass.