Floquet-East Glassiness Turns Noisy Quantum Hardware Into an Energy-Landscape Probe

Most discussions of noisy quantum computers treat noise as the enemy. It flips qubits, corrupts gates, washes out interference and turns elegant circuits into disappointing averages. A new August 2026 preprint asks a subtler question: can the same noisy, mid-circuit-measured hardware become a laboratory for the kind of sluggish, heterogeneous dynamics that physicists associate with glasses?

The paper, “Glassy dynamics with softened kinetic constraints on a noisy quantum computer,” was submitted to arXiv on August 19 and revised on August 28, 2026, by Marcel Cech, Igor Lesanovsky and Federico Carollo.[1] The team implements an instance of the Floquet-East model on a superconducting quantum processor and studies not just averaged observables, but full measurement trajectories: space-time records of what the circuit did at each monitored location and time step.[1] That trajectory view is the story. It lets researchers see dynamical patterns that can disappear when everything is averaged into a single smooth curve.

The energy lesson is not that noisy quantum computers are efficient machines today. It is that periodically driven quantum hardware can expose how energy, inactivity and constraints organize themselves in space and time before ordinary averages notice anything interesting.

For Floquet.ca, this article fits the “Google/industry Floquet code work” and “practical energy applications” rotation. It is not a power plant and not a battery prototype. Instead, it is a measurement advance for driven many-body physics: the same physics that underlies Floquet materials, quantum thermal machines, prethermal energy storage and the broader attempt to control matter by periodic pulses rather than static knobs.

Why an East model matters for energy flow

The East model is a classic example of a kinetically constrained system. In plain language, one part of the system can move only if a neighbor is already in the right state. That simple rule creates a bottleneck. Activity does not spread evenly like dye in water; it advances through facilitation, pauses, gaps and bursts. Such models are useful in glass physics because real glasses also appear stuck in complicated energy landscapes: some regions rearrange quickly, while others remain inactive for long periods.

The Floquet-East version adds periodic driving. Instead of one static rule, the system evolves through a repeating sequence of quantum gates. Each cycle defines a Floquet step, and the repeated drive creates a controlled non-equilibrium experiment. The paper combines three ingredients: kinetically constrained unitary operations, mid-circuit measurements and hardware noise.[1] The result is not a perfectly isolated textbook quantum system. It is closer to the devices researchers actually have: a programmable processor whose imperfections become part of the dynamics.

This matters for quantum energy research because useful driven systems usually live between two failures. If the drive is too weak or too constrained, nothing practical changes. If it is too violent or too noisy, the system heats and loses all structure. A softened constraint is a way to probe that middle ground: how much imperfection can the driven dynamics absorb before the organized response disappears?

The experimental platform: ibm_kingston and mid-circuit records

The authors report using the ibm_kingston quantum processor, described in the paper as one of IBM’s publicly accessible Heron r2 superconducting transmon devices with 156 qubits in a heavy-hex connectivity.[2] IBM’s public system page for ibm_kingston identifies it as an available IBM Quantum computer, while the paper emphasizes that the work was done under the open plan agreement and that the authors’ conclusions are their own rather than an official IBM position.[2][4]

The monitored circuit has two kinds of qubits. System qubits carry the constrained Floquet-East dynamics. Ancilla qubits sit beside them and are measured during the circuit so that the experiment records local activity. Instead of waiting until the end and asking only what final state survived, the researchers collect temporally and spatially resolved mid-circuit measurement outcomes.[1] These outcomes form a black-and-white movie of activity and inactivity across the chain.

That movie is the crucial data object. In the supplemental material, the authors state that the circuits were compiled using Qiskit and submitted to ibm_kingston on June 18, 2026 within 10 free runtime minutes of the open plan agreement.[2] For 40 time steps, they report approximate circuit depths of 360 for one parameter set and 1040 for another; they used 10,000 shots for the trajectory illustration and 200,000 shots per parameter set for the dynamical free-energy analysis.[2] Those numbers underline the practical point: trajectory-level many-body data now come from cloud-accessible quantum hardware, not only from bespoke laboratory apparatus.

What the team saw: heterogeneous dynamics

The headline observation is dynamical heterogeneity, a hallmark of glassy dynamics.[1] In a homogeneous system, every region behaves roughly alike once averaged over enough time. In a heterogeneous glassy system, some space-time patches remain inactive while others rearrange. The arXiv abstract says the team quantified this by studying the probability of inactive space-time regions of a given size.[1] That is a natural observable because it turns the measurement movie into a statistical question: how likely is a quiet island of a certain width and duration?

The answer shows a crossover from area-dominated to perimeter-dominated scaling.[1] For non-specialists, area scaling means the cost of an inactive patch grows mainly with the amount of space-time it occupies. Perimeter scaling means the boundary of the patch becomes the important object. In glass theory, this is not a decorative detail. It is associated with proximity to a dynamical first-order phase transition, where active and inactive histories compete in trajectory space.[1]

The experiment is not merely watching qubits decay. It is using repeated driving and measurement to ask whether “quiet” regions of a quantum many-body history behave like the inactive bubbles known from glass physics.

There is an important caution here. Noise is not magically turned into a virtue in all circumstances. The authors explicitly use noise as part of a softened-constraint interpretation, and they compare the observed behavior to effective bit-flip processes rather than claiming perfect microscopic control over every error channel.[2] That makes the result more, not less, useful. Energy technologies will not be built from ideal Hamiltonians alone; they will be built from driven systems with losses, measurement back-action, disorder and imperfect control.

Why this belongs on a quantum-energy site

At first glance, glassy dynamics may sound far from quantum heat engines or beyond-Carnot thermodynamics. The connection is energy accounting under constraints. Heat engines, quantum batteries and light-driven materials all rely on steering energy through a system while preventing it from simply thermalizing away. A Floquet protocol is a work source. A noisy processor is an open quantum system. A measured trajectory is an energy-and-information record. Put those together, and the experiment becomes a small but concrete example of how modern quantum thermodynamics is moving from ensemble averages to histories.

Traditional thermodynamics asks for average heat, work and entropy. Stochastic thermodynamics asks about fluctuating trajectories. Quantum trajectory methods push that idea into systems where measurement itself shapes the evolution. The Floquet-East experiment sits in that family. It uses a periodically driven circuit to create non-equilibrium histories and then analyzes rare inactive regions, not just mean activity.[1] That is exactly the kind of thinking needed when future devices must diagnose whether a driven material is storing useful order, leaking it as heat or hiding it in fluctuations.

The authors also make the code and data for the work available through Zenodo.[3] That matters. Trajectory-level claims depend on analysis choices: how inactive clusters are defined, how finite-size data are compared, how noise is modeled and how rare events are sampled. Public data and code make it easier for other groups to test whether the same signatures appear in different processors, different constrained circuits or different noise environments.

A realistic benchmark for Floquet engineering

One attractive feature of this paper is that it does not oversell the result as a finished technology. It shows a near-term quantum device doing near-term quantum science. The processor is noisy. The dynamics are deliberately monitored. The model is simplified. But the observable is sophisticated: a glass-style trajectory statistic in a periodically driven many-body system.

That combination is useful for Floquet engineering because many proposed applications depend on the same three-way balance. First, periodic control must create a response that is not available in the static system. Second, the response must survive realistic dissipation and noise. Third, researchers need diagnostics that can tell whether the response is robust or just an average hiding rapid microscopic disorder. The Floquet-East experiment gives a template for that diagnostic style.

For materials, an analogous question might be whether a laser-driven phase remains organized across domains or flickers in local patches. For quantum batteries, it might be whether stored ergotropy is globally available or trapped behind local constraints. For heat-control devices, it might be whether energy flow is genuinely rectified or only appears so after averaging over noisy cycles. In all three cases, trajectory-resolved thinking can prevent researchers from mistaking a smooth mean curve for a stable mechanism.

What to watch next

The first watch item is reproduction on other hardware. The paper’s result is strongest if related Floquet-East or kinetically constrained circuits can be run on different superconducting processors, trapped-ion systems or neutral-atom platforms, and if the same inactive-cluster scaling appears after appropriate calibration. The second watch item is energy-resolved measurement. Today’s circuit records are binary measurement outcomes; future experiments could connect those histories more directly to heat, work, entropy production or ergotropy.

The third watch item is design. Once researchers can measure glassy trajectory structure in a controlled Floquet circuit, they can try to engineer it. That could mean tuning noise as a resource, choosing drive sequences that stabilize useful inactive regions, or deliberately avoiding parameter regimes where a device becomes dynamically jammed. In practical energy language, the goal is not glassiness for its own sake. The goal is to know when constraints protect stored order and when they merely block useful power flow.

The bottom line

The new Floquet-East result is not a claim of beyond-Carnot efficiency, and it is not a shortcut to practical quantum energy storage. Its value is methodological. It shows how a noisy, periodically driven quantum processor can become a microscope for non-equilibrium histories: the local starts, stops and inactive bubbles that determine whether a driven system keeps structure or dissolves into heat.

That is a serious step for Floquet energy research. The field needs more than beautiful effective Hamiltonians. It needs ways to read the messy record of driven quantum matter as it actually evolves. Glassy trajectory analysis on quantum hardware is one promising way to do that.