Quantum Motion Superinductor Explained: A Quantum Sensor 10,000 Times Smaller, Built in a Standard Silicon Chip

on

Quantum Motion superinductor explained — the spiral inductor versus the tiny titanium nitride superinductor on a 22-nm CMOS chip

Diagram: Quantum Motion — the spiral inductor (left) versus the TiN thin-film superinductor (right) on a 22-nm CMOS chip (official figure from the Nature Electronics paper).

TL;DR: Quantum Motion superinductor explained in one line: researchers have built a superinductor directly into a standard silicon chip, shrinking the quantum sensor it powers by more than 10,000 times and making it over 100 times more sensitive. Published in Nature Electronics on 4 November 2026, it is a step toward quantum computers built in ordinary chip factories — though it has not yet read a real qubit.

📌 KEY HIGHLIGHTS

  • The result: a superinductor built inside a commercial 22-nm CMOS silicon chip, using a titanium nitride film already present in the manufacturing process.
  • The size win: the sensor's footprint shrinks by more than 10,000 times — the new superinductor is a strip just 0.36 micrometres wide, replacing a spiral inductor large enough to see with the naked eye.
  • The speed win: the sensor picks up a clear signal in as little as 1 picosecond, versus 625 picoseconds for the previous best — roughly 625 times faster.
  • The sensitivity win: the integrated sensor is more than 100 times more sensitive than the state of the art, according to the paper.
  • The honest limit: the device has been tested as a charge sensor so far. It has not yet read the state of a real qubit — that is the team's stated next step.

Quantum motion superinductor explained: what the team built

On 4 November 2026, a team led by Quantum Motion in London published a paper in Nature Electronics titled "A nonlinear superinductor in a deep-submicrometre integrated circuit". The team also included researchers from CIC nanoGUNE's Quantum Hardware group in Spain, the London Centre for Nanotechnology, Imperial College London and the University of Cambridge. The lead author is Tom Swift, a staff quantum engineer at Quantum Motion.

The achievement is easy to state and hard to do: the team built a superinductor — a circuit element with extremely high inductance — directly inside a standard silicon chip made with a commercial CMOS process. CMOS (complementary metal-oxide-semiconductor) is the recipe behind nearly every computer chip on the planet. The chip was fabricated in GlobalFoundries' 22-nanometre FDSOI process, the same kind of production line that makes ordinary electronics.

The trick is a material that was already there. The 22-nm process includes a thin film of titanium nitride (TiN) in the transistor gate stack. Cooled below about 1 kelvin — roughly one degree above absolute zero — that film becomes superconducting. Because TiN has very high kinetic inductance (about 1 nanohenry per square), a tiny strip of it can do the job of a bulky coiled inductor. The team paired the superinductor with a silicon quantum dot on the same chip, creating a fully integrated radio-frequency single-electron transistor, or rfSET — the most widely used charge sensor for semiconductor spin qubits.

What is a superinductor, and why do quantum chips need one?

An inductor is a component that resists sudden changes in current — think of it as an electrical flywheel. A superinductor is an inductor whose impedance (opposition to alternating current) exceeds a natural benchmark called the superconducting resistance quantum, about 6,450 ohms. The new device reaches a maximum impedance of 11.5 kilo-ohms, comfortably past that bar.

That high impedance is what makes the device useful. Reading a silicon spin qubit means detecting the movement of a single electron — an unimaginably small signal. The standard tool for the job, the rfSET, needs a high inductance to work, and on a CMOS chip that inductance has traditionally come from a flat spiral inductor. The problem is size: the spiral is visible to the naked eye on a 3×3 mm chip, while the qubits it serves are nanometres across. Conventional spirals store only 50–200 nanohenries per square millimetre, so they eat a large share of every chip. That size mismatch is what caps how many qubits can sit on one chip.

The TiN superinductor removes the spiral altogether. At 0.36 micrometres wide and about 50 micrometres long, it packs roughly 10,000 times more inductance into the same area. And because it is integrated into the chip rather than bolted on, parasitic losses fall — part of the reason the sensor is so much more sensitive. The film's non-linear inductance also boosts the readout signal at high power, another gain that comes free with the material.

FeatureSpiral inductor (old way)TiN superinductor (new)
SizeVisible to the naked eye on a 3×3 mm chipStrip 0.36 µm wide, ~50 µm long
FootprintBaselineMore than 10,000× smaller
Inductance density50–200 nH/mm²~1 nH per square of TiN film
Sensor sensitivityState of the artMore than 100× better
Minimum integration time625 picoseconds1 ± 0.3 picoseconds
Maximum impedanceBelow the superinductor threshold11.5 kΩ (threshold ≈ 6.45 kΩ)
ManufacturingStandard CMOS spiralStandard 22-nm CMOS, no exotic materials

Quantum Motion superinductor explained — a 12-inch silicon wafer, the kind of standard chip manufacturing the new sensor uses

A 12-inch silicon wafer. The new superinductor is built in a standard commercial CMOS process — the same kind of manufacturing used for ordinary chips. Illustrative image: 2x910, Wikimedia Commons (CC BY-SA).

Why it matters for quantum computing

Quantum computing has a scaling problem, and it is not only about qubits. As Tom Swift put it in Quantum Motion's announcement: "In quantum computing we talk a lot about scalability, but we focus mainly on the qubits themselves. In reality, we need to think about the scalability of the entire system and in particular how we might scale to a million-qubit system. With that goal in mind, we need to consider how we can make all of the different aspects of a quantum computer, the readout, the control, smaller and if possible, as close to the same size as the qubits themselves."

That is the real significance of this paper. Silicon spin qubits are attractive precisely because they can be made in the same factories as ordinary chips — unlike superconducting qubits, which need specialised fabrication. But that advantage only pays off if everything around the qubit shrinks too. A sensor 10,000 times smaller, built with no exotic materials in an existing commercial process, removes one of the biggest obstacles to packing thousands — eventually millions — of qubits onto silicon.

It also strengthens the case for the silicon route at a moment when the field is splitting. Superconducting machines such as IBM's Quantum System Two in Chicago are scaling fast, and software platforms such as Quantinuum's Nexus roadmap for fault-tolerant quantum apps are maturing. Silicon spin qubits have lagged on readout speed and density; this result attacks both at once. The work also connects to other spin-qubit research, including Osaka's triple-quantum-dot spin qubit, which tackled the speed-versus-noise trade-off from the qubit side.

Beyond quantum computing, the team notes the same TiN-based inductors could serve radio astronomy, low-power cryoelectronics, quantum sensing and simulations — anywhere a compact, ultra-sensitive superconducting circuit helps.

What it can't do — yet

The result deserves its caveats, and the team states them plainly. The device was demonstrated as a charge sensor: it detects single electrons. It has not yet read the state of a real qubit. Reading a spin qubit requires converting the qubit's spin into a charge change first, and that full readout chain has not been shown with this sensor.

The stated next step is to integrate the resonator with multi-gate transistors and demonstrate high-speed, high-fidelity spin readout well below the microsecond timescale. Until that happens, this is a sensor result, not a working qubit readout — impressive, peer-reviewed and precisely measured, but one step in a longer chain.

There are practical limits too. The device only works below about 1 kelvin, so it lives inside a dilution refrigerator like every other quantum chip. And a single sensor, however small, does not by itself deliver a million-qubit machine: control wiring, error correction and cryogenic electronics all have to scale alongside it.

What it means for readers in India

No Indian institution or company is involved in this work, and there is no product, price or India launch to report — this is a research result, not something to buy. The relevance for Indian readers is at the trend level. Silicon spin qubits run on the same CMOS production lines as ordinary chips, so progress like this is directly relevant to India's National Quantum Mission (approved in April 2023 with ₹6,003.65 crore of funding through 2030–31) and to India's push to attract semiconductor fabrication. If quantum hardware can be built in commercial fabs, countries building fab capacity — India included — have a clearer path into the field. For students and researchers, the paper itself is openly available as a preprint, and its core idea — using a material already inside the chip — is a useful lesson in how engineering constraints get solved.

Should you care?

✅ Yes, if you follow quantum computing. This is one of the clearest scaling wins of the year for silicon qubits: a 10,000-fold shrink and a 100-fold sensitivity gain, in a commercial process, peer-reviewed in Nature Electronics. ⏳ But keep it in perspective. It is a sensor demonstration, not a quantum computer, and the qubit-readout test is still to come. Watch for the team's next paper — a spin readout below one microsecond would turn this from a promising component into a proven building block.

FAQs

What is a superinductor?
A superinductor is a circuit element whose impedance exceeds about 6,450 ohms, the superconducting resistance quantum. It provides very high inductance in a tiny space, which makes it useful for quantum circuits and ultra-sensitive sensors. The new device reaches 11.5 kilo-ohms using a titanium nitride film inside a standard silicon chip.

How much smaller is the new quantum sensor?
More than 10,000 times smaller than the spiral-inductor version. The superinductor is a strip 0.36 micrometres wide and about 50 micrometres long, while the spiral inductor it replaces is large enough to see with the naked eye on a 3×3 mm chip.

Why does building it in a normal silicon chip matter?
Because it means the sensor can be manufactured in existing commercial chip factories using the standard 22-nm CMOS process, with no exotic materials. That is what makes it plausible to build thousands of sensors — and eventually millions of qubits — on silicon, instead of hand-crafting each one.

Has the sensor read a real qubit yet?
No. It has been tested as a charge sensor, detecting single electrons. The team's next step is to combine it with multi-gate transistors and demonstrate high-speed, high-fidelity readout of a real spin qubit, aiming for well below one microsecond.

Takeaway: Quantum Motion's superinductor shrinks the most widely used quantum sensor by 10,000 times and makes it 100 times more sensitive — built in an ordinary silicon chip factory. The qubit-readout test is still to come, but the scaling argument for silicon quantum computers just got much stronger.

If you want the bigger picture on where quantum hardware is heading, read our explainer on IBM's Quantum System Two in Chicago — the superconducting side of the same race.

Sources

Comments

Post a Comment