Distributed Quantum Information Processing: A Review of Recent Progress

Johannes Knörzer, Xiaoyu Liu, Benjamin F. Schiffer, Jordi TuraarXiv 2025 · DOI · arXiv

TL;DR

This review explores how interconnecting multiple quantum processors—rather than building one giant chip—overcomes the physical limits of current hardware. By shifting from monolithic to distributed architectures, researchers can not only scale qubit counts but also unlock "multi-copy" algorithms that perform tasks impossible for single-copy quantum devices.

What Problem Does This Paper Address?

Building a single, massive quantum computer (a "monolithic" device) is incredibly difficult. As you add more qubits to a single chip, you run into a "scalability wall": crosstalk between qubits increases, wiring becomes a nightmare, and cooling requirements spiral out of control.

The authors address the shift toward Distributed Quantum Information Processing (DQIP). Instead of one giant processor, DQIP envisions a network of smaller, manageable quantum nodes connected by classical and quantum communication links. The challenge is not just the hardware—how to move a quantum state from point A to point B without destroying it—but also the software: how do we design algorithms that can run across a fragmented network?

Methods

The paper surveys the three pillars of the distributed quantum ecosystem:

  1. Physical Interconnects: The "cables" of the quantum internet. This includes quantum state transfer (moving a qubit), transduction (converting a stationary qubit, like a superconducting circuit, into a "flying" qubit, like a photon), and remote entanglement generation.
  2. Circuit Knitting: A clever software trick to run large programs on small hardware. Researchers use wire cuts (breaking a circuit in time) or gate cuts (breaking it in space) to simulate a large, unified quantum computer using several smaller ones, albeit at the cost of more classical processing.
  3. Multi-Copy Protocols: This is a conceptual leap. In a distributed setup, you can have two nodes hold identical copies of the same quantum state $\rho$. By performing joint measurements on these copies (using tools like the SWAP test or Bell sampling), you can calculate properties like "purity" or "entanglement" much faster than you could with a single-copy device.

Datasets

As this is a comprehensive review paper rather than an empirical study, it does not use a specific training dataset. Instead, it synthesizes results from hundreds of experimental and theoretical papers, ranging from laboratory-scale demonstrations to metropolitan-scale quantum networks.

Key Insights and Ideas

  • The Multi-Copy Advantage: One of the most profound insights is that having access to multiple copies of a state simultaneously changes the fundamental complexity of a task. For example, estimating certain nonlinear properties of a quantum state is exponentially hard with one copy but becomes efficient if you can measure two copies together.
  • Transduction is the Bridge: Superconducting qubits (the current leaders in computing) operate at microwave frequencies, but optical fiber communication requires infrared light. Developing efficient "transducers" to bridge this gap is the "holy grail" of quantum networking.
  • Fidelity Trade-offs: Interestingly, the paper notes that quantum links don't have to be perfect to be useful. Techniques like entanglement distillation allow nodes to take many "noisy" entangled pairs and distill them into a few high-quality pairs.
  • Virtual Distillation: Before we have full error correction, we can use distributed nodes to perform "virtual distillation"—using multiple noisy copies of a state to mathematically "clean" the output of a computation.

Key Results

The authors highlight several major milestones that prove DQIP is moving from theory to reality:

  • Distance Records: Remote entanglement has been demonstrated over 101 km of fiber and even via satellite-based channels.
  • Hardware Scaling: Neutral atom arrays have reached scales of thousands of qubits, while trapped-ion systems have demonstrated high-fidelity "shuttling" of qubits between different zones on a chip.
  • Networked Logic: Experimentalists have successfully performed gate teleportation, where a logic operation is executed between two qubits that never physically touch, using a shared entangled resource.
  • Error Suppression: Real-world applications of virtual distillation have shown the ability to suppress errors by one to two orders of magnitude in superconducting systems.

Limitations and Future Work

Despite the progress, several "bottlenecks" remain:

  • The Rate Problem: Current rates for generating remote entanglement are still too slow for high-speed computing. We need orders-of-magnitude improvements to make distributed gates competitive with local ones.
  • Quantum RAM: We lack robust "quantum memories" that can store a qubit for long periods while waiting for a network signal to arrive.
  • Software Stack: There is no "Windows" or "Linux" for quantum networks yet. Developing compilers that can automatically "knit" circuits across a dynamic, noisy network is a massive open area of research.

The authors conclude that while the road to a "Quantum Internet" is long, the transition to modular, distributed hardware is likely the only viable path to truly large-scale, fault-tolerant quantum computing.

Summary updated

Abstract

Distributed quantum information processing seeks to overcome the scalability limitations of monolithic quantum devices by interconnecting multiple quantum processing nodes via classical and quantum communication. This approach extends the capabilities of individual devices, enabling access to larger problem instances and novel algorithmic techniques. Beyond increasing qubit counts, it also enables qualitatively new capabilities, such as joint measurements on multiple copies of high-dimensional quantum states. The distinction between single-copy and multi-copy access reveals important differences in task complexity and helps identify which computational problems stand to benefit from distributed quantum resources. At the same time, it highlights trade-offs between classical and quantum communication models and the practical challenges involved in realizing them experimentally. In this review, we contextualize recent developments by surveying the theoretical foundations of distributed quantum protocols and examining the experimental platforms and algorithmic applications that realize them in practice.
BibTeX
@misc{knorzer2025distributed,
  author = {Johannes Knörzer and Xiaoyu Liu and Benjamin F. Schiffer and Jordi Tura},
  title = {{Distributed Quantum Information Processing: A Review of Recent Progress}},
  year = {2025},
  eprint = {2510.15630},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2510.15630},
  howpublished = {arXiv},
}
Figure