Current quantum computing architectures face a thermodynamic bottleneck where the energy cost of classical error correction scales superlinearly with system size. Reliance on exogenous classical control imposes a latency and power penalty that renders hybrid paradigms unscalable for exascale applications. An architecture for autonomous dissipative quantum processing is defined to replace active feedback with hardware-intrinsic stability. By encoding logical information into the dark state of a driven-dissipative Liouvillian, error correction functions as a continuous thermodynamic process of entropy export. We explicitly define the implementation of logical Clifford gates via adiabatic code deformation and the physical realization of 4-body stabilizer terms using Fluxonium-based superinductor arrays. Thermal analysis confirms that the heat load ($< 0.5$ W/cm$^2$) remains within the cooling capacity of Stirling engines at 77 K, bounded by the Kapitza resistance of the diamond substrate. This establishes a rigorous pathway to self-contained quantum processors operating at thermodynamic limits.
Current quantum computing architectures face a thermodynamic bottleneck where the energy cost of classical error correction scales superlinearly with system size. Reliance on exogenous classical control imposes a latency and power penalty that renders hybrid paradigms unscalable for exascale applications. An architecture for autonomous dissipative quantum processing is defined to replace active feedback with hardware-intrinsic stability. By encoding logical information into the dark state of a driven-dissipative Liouvillian, error correction functions as a continuous thermodynamic process of entropy export. We explicitly define the implementation of logical Clifford gates via adiabatic code deformation and the physical realization of 4-body stabilizer terms using Fluxonium-based superinductor arrays. Thermal analysis confirms that the heat load (<0.5 W/cm2) remains within the cooling capacity of Stirling engines at 77 K, bounded by the Kapitza resistance of the diamond substrate. This establishes a rigorous pathway to self-contained quantum processors operating at thermodynamic limits.
Practical implementations of quantum information processing will ultimately rely on our ability to control large-scale quantum systems. This control faces two major challenges: the exponential growth of the state space with the number of components, and the ability to control dissipation in these systems. While current small- and medium-scale quantum platforms rely on traditional approaches that reason at the level of individual quantum states and aim to isolate systems from their environment, such strategies become increasingly impractical as system size grows. The complexity of interactions and the presence of noise in realistic implementations demand scalable approaches that go beyond state-by-state control. Such approaches require a shift toward reasoning in terms of collections of states that share symmetries or dynamical properties, and towards dealing with systems being far from their equilibrium state because of their openness. This thesis explores methods to understand and engineer dissipation for stabilizing quantum states and subspaces, enabling autonomous error correction, and facilitating robust logical operations, all through the lens of subspaces of the exponentially large state space. In particular, we develop symmetry-protected and nonlinear stabilization schemes that not only stabilize states and subspaces but also exhibit intrinsic error-correcting features rooted in driven-dissipative dynamics. We introduce a novel error correction strategy based on gradual entropy removal, offering scalable implementations with reduced overhead. Finally, we investigate the interplay between dissipation and logical operations, uncovering both limitations and opportunities in realistic quantum sensing and gate protocols, with a special focus on finite-energy bosonic codes.
We propose a novel measurement-free scheme for stabilizing a spin-oscillator hybrid qubit via autonomous quantum error correction. The engineered Lindbladian renders the code space into an attractive steady-state subspace, realized by coupling the storage mode to a rapidly cooled bath through a controlled beam-splitter and spin-dependent displacement interactions. The continuous variable-discrete variable hybrid approach to autonomous quantum error correction preserves the hardware efficiency of conventional dissipation engineering while simplifying the required system-bath coupling. The construction is compatible with simple logical gates and leverages primitives already demonstrated in experimental platforms, such as trapped-ion systems, suggesting a practical route to hardware-efficient, noise-biased logical qubits without repeated syndrome measurements and feedforward.
This work demonstrates real-time predictive classical-quantum feedback control on the IBM Kingston superconducting quantum processor, characterizing the hardware latency bottleneck of dynamic Hamiltonian shaping. Decoherence in quantum systems is traditionally viewed as a technical obstacle arising from environmental noise; this paper interprets decoherence as the coupling of a localized system to a dissipative, high-entropy environmental field. An ancilla-based modular sensor architecture is introduced to facilitate real-time error syndrome extraction without collapsing the target state. A software-defined controller, implemented via a quantized Variational Autoencoder-Feed-Forward Network (VAE-FFN) pipeline and optimized via empirical threshold calibration under strict 4-bit latency constraints, provides predictive Hamiltonian control to stabilize the system boundary. The integration of a 3D volumetric topological complex outlines a geometric architecture designed to mitigate spatial error cascades. Through temporal delay sweeps, empirical telemetry demonstrates that the volumetric lattice achieves localized containment of non-commuting phase defects, dynamically exploiting coherent Larmor precession to constrain spatial error propagation. Furthermore, rigorous statistical repetition mapping reveals a critical Noisy Intermediate-Scale Quantum (NISQ) latency bottleneck. The empirical data establishes that the temporal cost of classical-quantum feed-forward routing incurs a marginal macroscopic fidelity reduction (-1.01%). Crucially, this study reframes this differential not as a systemic regression, but as an explicit architectural trade-off: prioritizing rigorous variance suppression and operational stability over unmitigated peak fidelity. By establishing a highly predictable, predictively controlled baseline, this framework explores practical control strategies consistent with Approximate Quantum Error Correction (AQEC) approaches, bounding the operational consequences of the Eastin-Knill constraint within an addressable temporal control protocol.
Quantum error correction (QEC) normally requires active maintenance using measurements, classical processing and externally controlled correction schedules. Passive QEC, by contrast, was previously established only in unphysical spatial dimensions. Here, we give an explicit scheme for autonomous quantum error correction and computation in two dimensions, formulated as a dissipative quantum cellular automaton with a fixed, local and translation-invariant update rule. The construction uses hierarchical, self-simulating control elements based on ideas from the seminal classical results of Gács (1986, 1989) together with a measurement-free concatenated quantum code. We prove the existence of a nonzero noise threshold under a local noise model. Below this threshold, logical errors on encoded initial states are suppressed exponentially with increasing system size and the memory lifetime diverges in the thermodynamic limit. As an extension, we outline how this mechanism can also be implemented in continuous time by a time-independent, translation-invariant local Lindbladian. Quantum circuits can moreover be encoded in the automaton's initial state and executed fault-tolerantly with polylogarithmic overhead under the same circuit-independent rule. Our scheme thus realizes a self-correcting quantum computer capable of universal computation.
Autonomous quantum thermal machines have recently been proposed as physics-based computing substrates where logical inputs and outputs are encoded in temperature gradients. While such ``thermodynamic neurons'' exhibit a clear trade-off between computational fidelity and heat dissipation, the fundamental information-theoretic limits of temperature-encoded computation remain uncharacterized. Here, we derive rigorous bounds linking average error probability, channel capacity, and entropy production for finite-capacity thermal reservoirs operating far from equilibrium. We prove that the minimal dissipation required to achieve a target average error probability ⟨ξ⟩ diverges as ⟨ξ⟩ approaches a fundamental minimum error floor $\epsMin$ imposed by finite-reservoir thermal fluctuations. We further establish a thermodynamic channel capacity that saturates at high dissipation, and quantify the minimal dissipation required for cascaded networks to maintain target fidelity. We demonstrate that the required dissipation grows linearly O(L) with network depth for shallow networks, but diverges as the network depth L approaches a fundamental maximum limit $L_{\max} = \epsNet/\epsMin$ imposed by the minimum error floor. Furthermore, under strong noise amplification conditions, the required dissipation can grow up to O(L3). Our framework bridges stochastic thermodynamics, finite-time information theory, and autonomous computation, providing rigorous design principles for energy-efficient analog thermodynamic hardware.
Quantum computers hold immense promise for solving problems intractable for classical machines, yet their practical realization is hindered by the fragility of quantum states. Quantum Error Correction (QEC) is indispensable for protecting quantum information from environmental noise and operational imperfections. Traditional QEC schemes often rely on measurement-based feedback, which involves performing measurements, processing the classical outcomes, and then applying corrective operations. This process introduces latency, consumes significant resources, and can itself be a source of errors due to the invasive nature of quantum measurement and the classical-quantum interface. Coherent feedback control offers a compelling alternative by allowing the controller to directly interact with the quantum system using quantum fields, without intermediate measurement and classical processing. This maintains the coherence of the quantum information throughout the feedback loop, potentially leading to more robust and faster error correction. This paper explores the theoretical framework and potential advantages of employing coherent feedback control for autonomous QEC. We review existing QEC paradigms, delve into the principles of coherent feedback, and propose how this approach can be integrated into autonomous QEC architectures. Specifically, we investigate its application to bosonic codes and stabilizer codes, discussing the engineering challenges and the prospects for achieving truly fault-tolerant quantum computation by minimizing the classical overhead and enhancing real-time error suppression.