Quantum ring all-reduce: communication and privacy advantages for distributed learning

• URL: http://arxiv.org/abs/2606.20344v2
• Date: Tue, 23 Jun 2026 08:32:01 GMT
• Title: Quantum ring all-reduce: communication and privacy advantages for distributed learning
• Abstract summary: Machine learning models have scaled to unprecedented sizes, making training across distributed devices the de facto standard in the field.<n>In this work, we explore how quantum communications can make distributed training both more communication-efficient and information-theoretically private, for both classical and quantum learning models.<n>We present a quantum version that reduces per-link online communication by a provably optimal factor of two using pre-shared entanglement and superdense coding, without requiring the learning model or gradient to change.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Machine learning models have scaled to unprecedented sizes, making training across distributed devices the de facto standard in the field. In this work, we explore how quantum communications can make distributed training both more communication-efficient and information-theoretically private, for both classical and quantum learning models. Ring all-reduce is the foundational communication primitive for large-scale distributed training. We present a quantum version that reduces per-link online communication by a provably optimal factor of two using pre-shared entanglement and superdense coding, without requiring the learning model or gradient computation to change. Beyond bandwidth, the primitive enables privacy guarantees that are information-theoretically impossible for any classical protocol, achieving composable ε-secure aggregation, via verified entanglement, at a 2x overhead in GHZ copies. Our hybrid quantum-classical communication architecture yields simultaneous communication and security advantages for large scale distributed training, regardless of whether the learning itself is quantum or classical. Finally, we characterise quantum advantages in gradient conflict detection for server-to-client communication under bandwidth constraints, a setting that arises after ring all-reduce is completed, when full gradient broadcast to external clients is infeasible. Two variants of the problem admit different separations. For margin-based alignment testing (\textsc{GapIP}_τ), the quantum advantage is quadratic in the margin parameter: \widetilde{O}(τ^{-1}\log P) qubits versus \widetilde{O}(\min(\τ^{-2},P)) bits. For sign-consistency auditing against a private parameter matching (\textsc{TieAudit}_ε), the advantage represents an exponential separation in communication complexity: Ω(\sqrt{P}) bits whereas O(ε^{-2}\log P) qubits suffice.

Related papers

• Quantum Incremental Learning with Mixed State Prototypes [53.319426778720434]
  We introduce a novel quantum incremental learning framework based on trainable mixed-state prototypes.<n>Our model achieves high-dimensional feature concentration using a minimal number of qubits.
  arXiv  Detail & Related papers  (2026-08-11T04:22:42Z)
• Exponential Advantage of Multipartite Entanglement over Quantum Communication with Applications to Bounded-Storage Cryptography [0.5862480696321741]
  We show that a shared Greenberger-Horne-Zeilinger state enables completion of a task using only logarithmically many bits of classical communication from each sender.<n>Without preshared entanglement, any protocol achieving high success probability requires communication from at least one sender.<n>As a cryptographic application, we construct a seeded two-source extractor and establish an exponential separation between entangled and unentangled quantum side-information.
  arXiv  Detail & Related papers  (2026-07-30T10:04:53Z)
• Nonlocal continuous-variable quantum nondemolition gates by optical connections [0.5729426778193398]
  Nonlocal quantum gates, coupling quantum systems located at distance, are crucial for distributed quantum computing.<n>We propose a library of feasible protocols to implement a necessary nonlocal continuous-variable (CV) quantum nondemolition (QND) gate.
  arXiv  Detail & Related papers  (2026-03-13T10:13:37Z)
• Scalable & Noise-Robust Communication Advantage of Multipartite Quantum Entanglement [0.0]
  Quantum resources offer advantages over classical methods in addressing this challenge. We show that when the receiver and the senders share a multi-qubit Greenberger-Horne-Zeilinger (GHZ) state, certain global functions of the distributed inputs can be computed with only one bit of classical communication from each sender. We also show that the entanglement-based protocol exhibits significant robustness under white noise.
  arXiv  Detail & Related papers  (2024-09-20T05:17:09Z)
• Effects of Different Q-swaps Modes on Percolation Threshold in Small-world Quantum Networks [2.228665992307317]
  We show that Quantum Entanglement Percolation (QEP) has a better percolation performance than Classical Entanglement Percolation (CEP) We also introduce quantum walk as a new scheme to have an extra reduction in the percolation threshold.
  arXiv  Detail & Related papers  (2024-01-22T11:30:22Z)
• Quantum Federated Learning With Quantum Networks [7.842152902652214]
  We present a quantum-classical transfer learning scheme for classical data and communication with a hub-spoke topology. While quantum communication is secure from eavesdrop attacks and no measurements from quantum to classical translation, due to no cloning theorem, hub-spoke topology is not ideal for quantum communication without quantum memory. We also demonstrate the first successful use of quantum weights for quantum federated learning, which allows us to perform our training entirely in quantum.
  arXiv  Detail & Related papers  (2023-10-23T16:45:29Z)
• Exponential Quantum Communication Advantage in Distributed Inference and Learning [19.827903766111987]
  We present a framework for distributed computation over a quantum network. We show that for models within this framework, inference and training using gradient descent can be performed with exponentially less communication. We also show that models in this class can encode highly nonlinear features of their inputs, and their expressivity increases exponentially with model depth.
  arXiv  Detail & Related papers  (2023-10-11T02:19:50Z)
• Sparse Decentralized Federated Learning [35.32297764027417]
  Decentralized Federated Learning (DFL) enables collaborative model training without a central server but faces challenges in efficiency, stability, and trustworthiness.<n>We introduce a sparsity constraint on the shared model, leading to Sparse DFL (SDFL), and propose a novel algorithm, CEPS.<n> Numerical experiments validate the effectiveness of the proposed algorithm in improving communication and efficiency while maintaining a high level of trustworthiness.
  arXiv  Detail & Related papers  (2023-08-31T12:22:40Z)
• Quantum Federated Learning for Distributed Quantum Networks [9.766446130011706]
  We propose a quantum federated learning for distributed quantum networks by utilizing interesting characteristics of quantum mechanics. A quantum gradient descent algorithm is provided to help clients in the distributed quantum networks to train local models. A quantum secure multi-party computation protocol is designed, which utilizes the Chinese residual theorem.
  arXiv  Detail & Related papers  (2022-12-25T14:37:23Z)
• lo-fi: distributed fine-tuning without communication [61.87390899015787]
  We investigate completely local fine-tuning, which we refer to as lo-fi. During lo-fi, each node is fine-tuned independently without any communication. By removing the communication requirement, lo-fi reduces resource barriers for fine-tuning large models.
  arXiv  Detail & Related papers  (2022-10-19T20:15:18Z)
• Conference key agreement in a quantum network [67.410870290301]
  Quantum conference key agreement (QCKA) allows multiple users to establish a secure key from a shared multi-partite entangled state. In a quantum network, this protocol can be efficiently implemented using a single copy of a N-qubit Greenberger-Horne-Zeilinger (GHZ) state to distil a secure N-user conference key bit.
  arXiv  Detail & Related papers  (2022-07-04T18:00:07Z)
• Quantum communication complexity beyond Bell nonlocality [87.70068711362255]
  Efficient distributed computing offers a scalable strategy for solving resource-demanding tasks. Quantum resources are well-suited to this task, offering clear strategies that can outperform classical counterparts. We prove that a new class of communication complexity tasks can be associated to Bell-like inequalities.
  arXiv  Detail & Related papers  (2021-06-11T18:00:09Z)
• Computation-aided classical-quantum multiple access to boost network communication speeds [61.12008553173672]
  We quantify achievable quantum communication rates of codes with computation property for a two-sender cq-MAC. We show that it achieves the maximum possible communication rate (the single-user capacity), which cannot be achieved with conventional design.
  arXiv  Detail & Related papers  (2021-05-30T11:19:47Z)
• Composably secure data processing for Gaussian-modulated continuous variable quantum key distribution [58.720142291102135]
  Continuous-variable quantum key distribution (QKD) employs the quadratures of a bosonic mode to establish a secret key between two remote parties. We consider a protocol with homodyne detection in the general setting of composable finite-size security. In particular, we analyze the high signal-to-noise regime which requires the use of high-rate (non-binary) low-density parity check codes.
  arXiv  Detail & Related papers  (2021-03-30T18:02:55Z)
• Fault-tolerant Coding for Quantum Communication [71.206200318454]
  encode and decode circuits to reliably send messages over many uses of a noisy channel. For every quantum channel $T$ and every $eps>0$ there exists a threshold $p(epsilon,T)$ for the gate error probability below which rates larger than $C-epsilon$ are fault-tolerantly achievable. Our results are relevant in communication over large distances, and also on-chip, where distant parts of a quantum computer might need to communicate under higher levels of noise.
  arXiv  Detail & Related papers  (2020-09-15T15:10:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.