Their simulations show a spin-1/2 chain model experienced approximately 3.5 times faster thermalization in the presence of noise. This work establishes a new paradigm where noise, traditionally seen as an obstacle, can be harnessed for practical advantages on quantum computers even before fully fault-tolerant machines are available.
Noise Accelerates Gibbs State Preparation in Quantum Systems
In the current Noisy Intermediate-Scale Quantum era, noise is widely regarded as the primary obstacle to achieving fault-tolerant quantum computation. However, certain stages of the quantum computing pipeline can, in fact, benefit from this noise. In the study, the researchers exploit the Eigenstate Thermalization Hypothesis to show that noise generically accelerates a fundamental task in quantum computing, the preparation of Gibbs states. The researchers demonstrate this behavior using classical and quantum simulations with Haar-random and phase-flip noise, respectively, on a spin-1/2 chain with a local Hamiltonian.
Their non-integrable model sees approximately 3.5 times faster thermalization in the presence of noise, while their integrable model, which would not otherwise thermalize, reaches a thermal state due to noise. Since certifying a local Gibbs state is relatively easy on a quantum computer, their approach provides a new practical solution to a key problem in quantum computing.
Simulations reveal that introducing phase-flip noise can smooth the convergence toward localized Gibbs states, an important step in preparing data for quantum computations, and potentially accelerating the process on near-term devices. While noise is typically viewed as a hindrance to quantum computing, this work demonstrates that specific noise models can actively aid in achieving desired quantum states, particularly in strongly correlated quantum systems exhibiting chaotic dynamics.
The research builds on a growing body of work exploring constructive uses for noise within the NISQ era, moving beyond simply mitigating its effects. The team’s approach uses the Eigenstate Thermalization Hypothesis, observing that noise suppresses recurrences typically seen in noiseless quantum circuits, leading to a more efficient path toward thermalization. This is particularly notable in integrable models, systems that would not normally reach a thermal state, which, when subjected to this noise, demonstrably do thermalize.
Eigenstate Thermalization Hypothesis & Localized Gibbs States
Simulations reveal a non-integrable spin-1/2 chain experienced approximately 3.5 times faster thermalization when noise was introduced, a finding that challenges conventional wisdom regarding the role of noise in quantum computing. The work builds on the Eigenstate Thermalization Hypothesis, which explores thermalization dynamics in closed quantum systems and describes how subsystems exchange information with their environment to reach a Gibbs state. This approach allows for benchmarking thermalization by measuring local observables in subregions of a quantum processor, provided the coupling map doesn’t isolate qubits.
The versatility of this protocol stems from the Eigenstate Thermalization Hypothesis’s independence from the specific architecture of Noisy Intermediate-Scale Quantum (NISQ) devices. This means the technique is applicable across various modalities and coupling maps, subject to the constraints of the simulations. The researchers found that naturally occurring phase-flip noise was sufficient to realize the observed advantage, suggesting a practical implementation within the limitations of current hardware.
Haar-Random and Phase-Flip Noise Simulation Methods
Simulations reveal a spin-1/2 chain model exhibited approximately 3.5 times faster thermalization when subjected to noise, a critical step in preparing data for quantum computation. Researchers explored how strategically applied noise could accelerate the preparation of Gibbs states, fundamental to many quantum algorithms, using both classical and quantum simulations. The work proposes a specific noise protocol involving the interleaving of Hamiltonian dynamics with noise, inducing non-integrability and chaos to expedite Gibbs state preparation. This protocol uses Haar-random unitary operations interspersed with Hamiltonian evolution. The simulations demonstrated that even integrable systems, those conventionally expected to remain non-thermal, reached a thermal state when noise was introduced.
Simulations reveal a spin-1/2 chain model achieved 3. Their non-integrable model sees approximately 3.5 times faster thermalization in the presence of noise, while their integrable model, which would not otherwise thermalize, reaches a thermal state due to noise.
Noise Protocol: Interleaving Unitaries with Haar-Randomness
This noise protocol introduces controlled “shocks” between unitary evolutions, effectively enhancing thermalization rates. The approach uses a specific construction of the Haar-random unitary, denoted as = QΛ, to introduce these noise elements. Simulations on a spin-1/2 chain with 24 spins and 3 spins pointing up demonstrated the protocol’s efficacy. The study considered an initial many-body product state, acknowledging the intractability of exact diagonalization due to the Hilbert space dimension. Here, = e^(-iH(t₂-t₁))Me^(-iHt₁) represents the application of noise between time evolutions governed by the Hamiltonian. The researchers then examined the reduced density matrices over a subspace for both noiseless and noisy protocols, revealing the impact of the non-commuting noise operator.
See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
