Duke and IonQ entangle three remote quantum processors over fiber

Overview

Updated Jun 22

For years, building a bigger quantum computer has mostly meant cramming more qubits onto one chip. Duke University and the company IonQ just showed a different path. They wired three separate ion processors together with optical fiber and got all three to share one quantum state, with no shared chip and no logic gate at a central hub.

The shared state, called a GHZ state, held at 84 to 88 percent fidelity across modules about two meters apart. A non-locality test beat the classical limit by 27 standard deviations. The result is a working building block for 'modular' quantum machines that grow by networking processors, the way data centers grow by adding servers.

Why it matters

If quantum computers can scale by networking many small processors, useful machines arrive years sooner than waiting for one giant flawless chip.

Questions about this story

Free account needed to ask — your question is kept and asked for you right after sign-up. Answers are public.

No questions yet — be the first to ask.

Key Indicators

3 nodes

Networked processors

Three individually controlled trapped-ion modules shared one entangled state.

84–88%

State fidelity

How closely the shared three-way state matched the ideal GHZ target.

27σ

Beat classical limit by

Standard deviations by which the non-locality test exceeded any classical explanation.

~2 m

Distance between modules

Modules sat about two meters apart, linked by 3-meter optical fibers.

0.095/s

Entanglement rate

Roughly one shared three-way state every ten seconds, the speed still to improve.

Voices

Curated perspectives — historical figures and your fellow readers.

Ever wondered what historical figures would say about today's headlines?

People Involved

Christopher Monroe

Physicist, Duke University; co-founder of IonQ

Leading trapped-ion research at the Duke Quantum Center

Jungsang Kim

Engineer, Duke University; co-founder of IonQ

Working on scalable ion-trap and photonics systems; received South Korea's Changjo Medal, April 2026

Organizations Involved

IonQ

Quantum computing company

Public company pursuing networked, modular quantum systems

A publicly traded quantum computing firm building trapped-ion machines and the photonic links to network them.

Duke Quantum Center

University research center

Co-author of the three-node network result

Duke University's hub for trapped-ion quantum research, closely tied to IonQ's founders.

Timeline

May 2015 June 2026

4 events Latest: June 20th, 2026 · 3 months ago

Tap a bar to jump to that date

June 2026
Three remote ion processors share one state
Latest Research Result

Duke and IonQ entangle three individually controlled trapped-ion nodes into a GHZ state at 84–88% fidelity, with no central logic gate and no post-selection.
June 20th, 2026
April 2026
Kim receives South Korea's Changjo Medal
Award

Prime Minister Kim Min-Seok presented Jungsang Kim with the Changjo Medal in Seoul — South Korea's top national honor for scientists — for securing core technologies behind commercially viable trapped-ion quantum computers.
April 21st, 2026
Two commercial systems linked by light
Milestone

IonQ and the Air Force Research Laboratory entangle two separate commercial quantum systems over a photonic interconnect, a first for networked commercial machines.
April 14th, 2026
May 2015
IonQ founded on trapped-ion bet
Founding

Christopher Monroe and Jungsang Kim launch IonQ to commercialize trapped-ion quantum computing, with networking on the long-term roadmap.
May 1st, 2015

Scenarios

IonQ ships a networked machine made of multiple linked processors

Possible Resolves by End of 2027

Discussed by: IonQ roadmap statements and quantum trade press (Quantum Computing Report, The Quantum Insider)

IonQ has staked its scaling plan on networking. If the three-node result holds up, the next step is a customer-facing system that runs computations across two or more photonically linked processors as one machine. That requires faster entanglement rates and stable, repeatable interconnect hardware.

The three-node result clears peer review

Likely Resolves by Q2 2027

Discussed by: Quantum research community and journal preprint trackers

The demonstration was reported via trade coverage and a preprint. The credibility test is publication in a peer-reviewed journal such as Nature, Science, or Physical Review X. Independent referees checking the fidelity and loophole-closing claims would settle whether the numbers stand.

A rival platform networks three or more nodes at higher fidelity

Uncertain Resolves by End of 2027

Discussed by: Competing quantum groups (neutral-atom, photonic, and other ion-trap teams)

Trapped ions are one of several platforms chasing modular scaling. A neutral-atom, photonic, or competing ion-trap group could demonstrate a three-or-more-node entangled network beating 88 percent fidelity. That would show the modular approach is broader than one company and reset the benchmark.

Entanglement rate jumps tenfold

Possible Resolves by Q2 2028

Discussed by: Trapped-ion researchers citing rate as the main bottleneck

The current rate, about one shared state every ten seconds, is far too slow for real computation. A credible path to modular machines needs that rate up by roughly ten times or more. A follow-up demonstration hitting at least one entanglement event per second would show the bottleneck is movable.

Historical Context

3 moments from history that rhyme with this story — and how they unfolded.

October 1969

ARPANET's first node-to-node link (1969)

Engineers sent the first message between two computers on ARPANET, the network that grew into the internet. The link was crude and the system crashed mid-message, but separate machines had talked to each other.

Then

ARPANET grew from two nodes to a handful within a year as more sites joined.

Now

Networking separate computers, rather than building one giant mainframe, became the dominant way computing scaled.

Why this matters now

Modular quantum computing makes the same wager: link many small processors instead of perfecting one huge one. Moving from a two-node link to a three-node shared state echoes those first network steps.

1995

First trapped-ion quantum logic gate (1995)

At the National Institute of Standards and Technology, a team including Christopher Monroe and David Wineland used a single trapped ion to run a controlled-NOT gate, an early quantum logic operation. It showed that individual atoms could serve as controllable qubits.

Then

The result proved trapped ions could perform basic quantum logic, drawing more researchers to the platform.

Now

It seeded the trapped-ion field and Wineland's later Nobel Prize, and set the technical lineage that produced IonQ.

Why this matters now

The same people and platform behind that first gate now anchor the three-node network. This is the next rung on a thirty-year ladder.

October 2015

Loophole-free Bell test at Delft (2015)

Physicists at Delft University of Technology entangled two electrons in diamond crystals 1.3 kilometers apart and ran a Bell test that closed the main experimental loopholes at once. It was the strongest confirmation yet that entanglement defies classical physics over distance.

Then

The experiment settled decades of debate about whether earlier Bell tests had hidden flaws.

Now

It became a benchmark for rigorous tests of entanglement across remote, separately controlled systems.

Why this matters now

The Duke-IonQ test closes the detection loophole with individually addressable atoms, applying that same rigor to a three-node computing network.