Noise-stabilized discrete time crystals on digital quantum processors

This paper demonstrates that structured quantum noise, specifically engineered ancilla errors on IBM superconducting processors, can paradoxically stabilize discrete time crystals in a kicked Ising model by acting as spatiotemporal disorder that prevents thermalization, thereby enabling robust subharmonic oscillations that are absent in noiseless simulations.

The Gist

The Big Idea: Turning a Bug into a Feature

Imagine you are trying to keep a spinning top balanced on a table. Usually, if the table is shaky (noisy), the top falls over immediately. In the world of quantum computers, "noise" (random errors) is the ultimate enemy. It usually destroys delicate quantum states before scientists can study them.

This paper presents a surprising twist: The researchers found a way to use the "shaky table" to actually keep the top spinning longer.

They discovered that on specific types of quantum computers, the random errors don't just destroy the system; they accidentally create a "shield" that stabilizes a strange, rhythmic pattern called a Discrete Time Crystal (DTC).

What is a Discrete Time Crystal?

To understand a DTC, imagine a music beat.

• Normal Rhythm: If you tap a drum every second, the sound happens every second.
• Time Crystal Rhythm: If you tap the drum every second, but the sound only echoes back every two seconds, you have broken the rhythm. The system is "resisting" your beat and finding its own, slower rhythm.

In physics, this is a "time crystal." It's a state of matter that repeats itself in time, not just space. The problem is, these crystals are usually very fragile. A little bit of noise usually makes them stop echoing and just fall into a chaotic mess (thermalize).

The Experiment: Building a Lattice on a "Heavy-Hex" Floor

The researchers used IBM's superconducting quantum computers (specifically the "Eagle" and "Heron" processors). These computers have a fixed layout of qubits (the quantum bits) that looks like a honeycomb with extra legs (called a "heavy-hex" lattice).

However, the scientists wanted to study a different shape: a Kagome lattice (which looks like a pattern of interlocking triangles and hexagons, like a woven basket).

The Analogy: Imagine you have a floor made of hexagonal tiles, but you want to build a house with triangular rooms. You can't just move the tiles. Instead, you have to use "helper" tiles (called ancillas) to bridge the gaps between the tiles you actually want to use.

The researchers used these "helper" qubits to connect the main qubits, effectively building their triangular lattice on top of the computer's fixed hexagonal floor.

The Two Magic Regimes

The paper describes two different ways this "noise-stabilized" time crystal works, depending on the shape of the lattice they built.

1. The "Guarded Gate" Regime (Boundary-Assisted)

• The Setup: In some lattice shapes, the edges of the system act like a special "gate" that naturally wants to keep a rhythm.
• The Problem: Without help, the middle of the system gets chaotic and ruins the rhythm.
• The Fix: The researchers found that the noise (the errors from the helper qubits) acts like a chaotic wind. Surprisingly, this wind pushes against the "gate" just enough to lock it in place. The noise and the gate work together like a dance partner, where the chaos of one helps the other maintain a steady beat.
• Result: A strong, rhythmic echo that lasts a long time, localized at the edges.

2. The "Noise-Only" Regime (The Miracle)

• The Setup: In other lattice shapes, there are no special "gates" at the edges. In a perfect, noise-free world, this system would instantly fall apart and become chaotic.
• The Miracle: When the researchers turned on the noise, something magical happened. The system started to echo!
• The Analogy: Imagine a room full of people trying to whisper a secret. In a quiet room, everyone talks over each other and the secret is lost. But if you add a specific kind of background static (noise), it actually forces everyone to pause and listen, creating a synchronized whisper that lasts much longer.
• Result: The noise itself created the rhythm. Without the noise, there was no rhythm at all.

How They Proved It

The researchers didn't just guess; they did a "double-check" using two methods:

1. The Real Hardware: They ran the experiment on the actual IBM quantum computers and measured the "magnetization" (a way of checking if the qubits are spinning in sync).
2. The Digital Twin: They ran a simulation on a supercomputer that included a model of the computer's specific errors.

The Result: The simulation matched the real hardware perfectly. This proved that the "noise" wasn't just random garbage; it was a structured, predictable force that was actively stabilizing the time crystal.

Why This Matters

This is a huge shift in how we think about quantum computing:

• Old View: Noise is the enemy. We must eliminate it at all costs.
• New View: Noise can be a tool. By understanding how the noise behaves, we can engineer it to create new states of matter that wouldn't exist in a perfect world.

The Takeaway:
The researchers showed that on current, imperfect quantum computers, we don't have to wait for "perfect" machines to do cool physics. We can use the imperfections we have right now as a "control knob" to create and stabilize exotic, rhythmic states of matter. They turned the computer's biggest weakness into its strongest feature.

Technical Summary

1. Problem Statement

Discrete Time Crystals (DTCs) are non-equilibrium phases of matter characterized by robust subharmonic oscillations that break discrete time-translation symmetry. While theoretically fascinating, realizing DTCs on current quantum hardware faces two primary challenges:

• Fragility to Noise: DTCs are typically destroyed by decoherence and thermalization before they can be observed over long timescales.
• Connectivity Constraints: Superconducting quantum processors (like IBM's) possess fixed connectivity (e.g., heavy-hex lattices), making it difficult to directly implement complex, frustrated 2D lattice geometries (such as Kagome or Lieb lattices) required for studying higher-dimensional DTC physics.

The central question addressed is whether engineered noise can be transformed from a detrimental factor into a stabilizing resource for DTCs, and how to implement complex 2D lattice dynamics on restricted hardware architectures.

2. Methodology

A. Hardware and Embedding

• Devices: The experiments were conducted on IBM's Eagle (ibm_kyiv) and Heron (ibm_torino, ibm_marrakesh) superconducting processors.
• Lattice Embedding: The researchers implemented 2D Kagome and Lieb lattices on the native heavy-hex connectivity of these devices.
  • Ancilla-Assisted Embedding: System qubits (representing spins) were mapped to coordination-2 sites, while ancilla qubits (coordination-3 sites) were used to mediate interactions between non-adjacent system qubits.
  • Gate Implementation: The two-qubit Ising interaction ($\hat{R}_{Z_i Z_j}$) was realized using a three-qubit phase gadget involving an ancilla. This allowed the simulation of arbitrary 2D geometries despite the hardware's limited connectivity.

B. Model and Dynamics

• Model: A periodically driven Kicked Ising Model with a transverse field.
  • Hamiltonian: $\hat{H}(t)$ alternates between a transverse field $\hat{X}$ and Ising interactions $\hat{Z}_i \hat{Z}_j$.
  • Floquet Unitary: $\hat{U}_F = e^{-i \hat{H}_{int} T/2} e^{-i \hat{H}_{field} T/2}$.
  • Parameters: Fixed interaction angle $\theta_J = -\pi/2$ and variable transverse field angle $\theta_x = \pi - \epsilon$ (weak perturbation).
• Initialization: The system was initialized in a fully polarized product state ($|0\rangle^{\otimes L}$).
• Observables: Stroboscopic local magnetization $\langle \hat{Z}_j(t) \rangle$ was measured to detect period-doubling (oscillations at half the drive frequency).

C. Error Mitigation and Simulation

• Error Mitigation: A global depolarizing noise model was used to renormalize raw experimental data, compensating for signal decay without dynamical decoupling.
• Classical Simulation: To validate results, the authors performed Matrix Product State (MPS) simulations. Crucially, they incorporated an effective ancilla-noise model derived from experimental device data.
  • Noise Mechanism: Ancilla errors were modeled as stochastic sign flips of the Ising coupling angle $\theta_J$ with probability $p$ per cycle.
  • Parameter Extraction: The noise probability $p$ was estimated from the decay of ancilla magnetization ($\eta$) via $p \approx (1-\eta)/2$.

3. Key Contributions

1. Noise as a Stabilizing Resource: The paper demonstrates that structured quantum noise (specifically ancilla-induced spatiotemporal disorder) can stabilize DTCs rather than destroy them. This challenges the conventional view of noise as purely detrimental.
2. Unified Mechanism for DTCs: The authors identify a unified mechanism where ancilla-induced noise acts as a source of spatiotemporal disorder that induces stochastic sign flips in Ising couplings. This mechanism stabilizes DTCs in two distinct regimes:
   • Boundary-Assisted Regime: Where intrinsic boundary $\pi$-modes exist (due to symmetry-charge pumping).
   • Noise-Only Regime: Where no boundary modes exist, and the DTC is sustained solely by the noise.
3. Scalable 2D Embedding: The successful embedding of Kagome and Lieb lattices into heavy-hex architectures using ancilla qubits provides a scalable pathway for simulating frustrated magnets and complex geometries on near-term hardware.

4. Key Results

A. Two Complementary Regimes

The study reveals two distinct dynamical behaviors depending on the lattice geometry and the presence of symmetry-charge pumping (which creates boundary-localized $\pi$-modes):

1. Boundary-Assisted DTC ("Boundary + Noise"):

   • Context: Observed in Kagome82 and Kagome53-I lattices where the noiseless dynamics support symmetry-charge pumping at boundary sites.
   • Mechanism: In the noiseless limit, boundary sites host robust $\pi$-modes while the bulk thermalizes. The addition of ancilla noise cooperates with these boundary modes to suppress scrambling, resulting in a DTC where the subharmonic response is sharply localized at the boundary but extended into the bulk.
   • Result: Experimental data matched noisy MPS simulations ($p \approx 0.02$) over 40 Floquet cycles.

2. Noise-Induced DTC ("Noise-Only"):

   • Context: Observed in Kagome53-II and Lieb lattices where the geometry is designed such that no symmetry-charge pumping occurs ($P = \emptyset$).
   • Mechanism: In the noiseless limit, these systems rapidly thermalize with no subharmonic order. However, the introduction of ancilla-induced noise (stochastic sign flips) slows down thermalization and generates a long-lived DTC response.
   • Result: On the noisier ibm_kyiv device (higher $p \approx 0.1$), the noise-induced DTC was more pronounced than on the quieter ibm_marrakesh ($p \approx 0.02$). This confirms that the noise itself is the stabilizing agent.

B. Quantitative Agreement

• The error-mitigated experimental data showed quantitative agreement with noisy MPS simulations incorporating the extracted ancilla noise parameters.
• The simulations successfully reproduced both the amplitude and lifetime of the subharmonic oscillations, validating the effective noise model.

C. Information Blockade and Scrambling

• OTOC Analysis: Out-of-Time-Ordered Correlators (OTOCs) were used to measure quantum scrambling.
  • In the Boundary-Assisted regime, pumped sites acted as an "information blockade," preventing the spread of quantum information and protecting the boundary modes.
  • In the Noise-Only regime, the stochastic sign flips introduced by noise effectively suppressed scrambling, creating a prethermal plateau that would not exist in a noiseless system.

5. Significance

• Paradigm Shift: This work fundamentally shifts the perspective on quantum noise in NISQ (Noisy Intermediate-Scale Quantum) devices. It demonstrates that engineered noise can be a tunable control knob to induce and stabilize non-equilibrium dynamical phases.
• Robustness in Higher Dimensions: It provides evidence that DTCs can be stabilized in 2D systems without relying on many-body localization (which is unstable in higher dimensions), using noise-induced disorder instead.
• Hardware Utility: The ancilla-assisted embedding technique allows for the simulation of complex, frustrated 2D lattices (Kagome, Lieb) on current heavy-hex hardware, expanding the scope of digital quantum simulation.
• Prethermalization: The results highlight the role of prethermalization in protecting quantum order, showing that noise can extend the lifetime of these prethermal states significantly beyond what is possible in idealized, noiseless simulations for certain geometries.

In conclusion, the paper establishes that on scalable superconducting quantum processors, ancilla-induced noise is not merely a limitation but a functional resource capable of stabilizing discrete time crystals through a unified mechanism of spatiotemporal disorder, applicable across different lattice geometries and boundary conditions.