Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

DeepQuantum: A PyTorch-based Software Platform for Quantum Machine Learning and Photonic Quantum Computing

Jun-Jie He, Ke-Ming Hu, Yu-Ze Zhu, Guan-Ju Yan, Shu-Yi Liang, Xiang Zhao, Ding Wang, Fei-Xiang Guo, Ze-Feng Lan, Xiao-Wen Shang, Zi-Ming Yin, Xin-Yang Jiang, Lin Yang, Hao Tang, Xian-Min Jin

TL;DR

DeepQuantum addresses the need for a unified, AI-integrated platform that spans qubit-based, photonic, and measurement-based quantum computing. By embedding quantum components in PyTorch, it enables differentiable, end-to-end hybrid quantum-classical workflows and supports large-scale simulations via tensor networks and distributed architectures. The work demonstrates concrete capabilities across QubitCircuit, QumodeCircuit, and Pattern MBQC models, including QResNet, QCNN, GBS, EPR/TDM, Clements decompositions, and distributed QFT, with benchmark-backed claims of efficiency and scalability. This platform has the potential to accelerate both quantum algorithm development and AI-assisted quantum hardware research, fostering practical quantum utility and paving the way toward fault-tolerant paradigms.

Abstract

We introduce DeepQuantum, an open-source, PyTorch-based software platform for quantum machine learning and photonic quantum computing. This AI-enhanced framework enables efficient design and execution of hybrid quantum-classical models and variational quantum algorithms on both CPUs and GPUs. For photonic quantum computing, DeepQuantum implements Fock, Gaussian, and Bosonic backends, catering to different simulation needs. Notably, it is the first framework to realize closed-loop integration of three paradigms of quantum computing, namely quantum circuits, photonic quantum circuits, and measurement-based quantum computing, thereby enabling robust support for both specialized and universal photonic quantum algorithm design. Furthermore, DeepQuantum supports large-scale simulations based on tensor network techniques and a distributed parallel computing architecture. We demonstrate these capabilities through comprehensive benchmarks and illustrative examples. With its unique features, DeepQuantum is intended to be a powerful platform for both AI for Quantum and Quantum for AI.

DeepQuantum: A PyTorch-based Software Platform for Quantum Machine Learning and Photonic Quantum Computing

TL;DR

DeepQuantum addresses the need for a unified, AI-integrated platform that spans qubit-based, photonic, and measurement-based quantum computing. By embedding quantum components in PyTorch, it enables differentiable, end-to-end hybrid quantum-classical workflows and supports large-scale simulations via tensor networks and distributed architectures. The work demonstrates concrete capabilities across QubitCircuit, QumodeCircuit, and Pattern MBQC models, including QResNet, QCNN, GBS, EPR/TDM, Clements decompositions, and distributed QFT, with benchmark-backed claims of efficiency and scalability. This platform has the potential to accelerate both quantum algorithm development and AI-assisted quantum hardware research, fostering practical quantum utility and paving the way toward fault-tolerant paradigms.

Abstract

We introduce DeepQuantum, an open-source, PyTorch-based software platform for quantum machine learning and photonic quantum computing. This AI-enhanced framework enables efficient design and execution of hybrid quantum-classical models and variational quantum algorithms on both CPUs and GPUs. For photonic quantum computing, DeepQuantum implements Fock, Gaussian, and Bosonic backends, catering to different simulation needs. Notably, it is the first framework to realize closed-loop integration of three paradigms of quantum computing, namely quantum circuits, photonic quantum circuits, and measurement-based quantum computing, thereby enabling robust support for both specialized and universal photonic quantum algorithm design. Furthermore, DeepQuantum supports large-scale simulations based on tensor network techniques and a distributed parallel computing architecture. We demonstrate these capabilities through comprehensive benchmarks and illustrative examples. With its unique features, DeepQuantum is intended to be a powerful platform for both AI for Quantum and Quantum for AI.

Paper Structure

This paper contains 34 sections, 36 equations, 32 figures, 3 tables.

Table of Contents

  1. Introduction
  2. DeepQuantum Overview
  3. Benchmarks
  4. Gradient and Hessian
  5. Hafnian, Torontonian and permanent
  6. Transpilation and forward computation
  7. Quantum Computing with Qubits
  8. Background
  9. API Overview
  10. Applications
  11. QResNet
  12. QCNN
  13. Photonic Quantum Computing
  14. Background
  15. API Overview
  16. ...and 19 more sections

Figures (32)

  • Figure 1: The architecture of DeepQuantum and its ecosystem, including high-level quantum algorithms, toolchain, and low-level hardware.
  • Figure 2: Benchmark results of gradient (a) and Hessian (b) computation among DeepQuantum, MindQuantum, PennyLane, and PyQPanda.
  • Figure 3: Benchmark results of Hafnian, Torontonian, and permanent computation between DeepQuantum, Strawberry Fields, and Piquasso.
  • Figure 4: Benchmark results of pattern transpilation and forward computation between DeepQuantum and Graphix.
  • Figure 5: Visualization of a QubitCircuit instance.
  • ...and 27 more figures