• Corpus ID: 237363361

Bridging the Gap between Deep Learning and Frustrated Quantum Spin System for Extreme-scale Simulations on New Generation of Sunway Supercomputer

  title={Bridging the Gap between Deep Learning and Frustrated Quantum Spin System for Extreme-scale Simulations on New Generation of Sunway Supercomputer},
  author={Mingfan Li and Junshi Chen and Qian Xiao and Qingcai Jiang and Xuncheng Zhao and Rongfen Lin and Fei Wang and Xiao Liang and Lixin He and Hong An},
Efficient numerical methods are promising tools for delivering unique insights into the fascinating properties of physics, such as the highly frustrated quantum many-body systems. However, the computational complexity of obtaining the wave functions for accurately describing the quantum states increases exponentially with respect to particle number. Here we present a novel convolutional neural network (CNN) for simulating the two-dimensional highly frustrated spin-1/2 J1 − J2 Heisenberg model… 
1 Citations
Neural Network Evolution Strategy for Solving Quantum Sign Structures
Feed-forward neural networks are a novel class of variational wave functions for correlated manybody quantum systems. Here, we propose a specific neural network ansatz suitable for systems with


Two-dimensional frustrated J1−J2 model studied with neural network quantum states
This paper uses a fully convolutional neural network model as a variational ansatz to study the frustrated spin-1/2 J1-J2 Heisenberg model on the square lattice and demonstrates that the resulting predictions for both ground-state energies and properties are competitive with, and often improve upon, existing state-of-the-art methods.
Solving frustrated quantum many-particle models with convolutional neural networks
Recently, there has been significant progress in solving quantum many-particle problem via machine learning based on the restricted Boltzmann machine. However, it is still highly challenging to solve
PEPS++: Towards Extreme-Scale Simulations of Strongly Correlated Quantum Many-Particle Models on Sunway TaihuLight
This paper implements PEPS++ on Sunway TaihuLight based on a carefully designed tensor computation library for manipulating high-rank tensors and optimize it by invoking various high-performance matrix and tensor operations.
Generalization properties of neural network approximations to frustrated magnet ground states
The authors show that limited generalization capacity of neural network representations of quantum states is responsible for convergence problems for frustrated systems.
Solving the quantum many-body problem with artificial neural networks
A variational representation of quantum states based on artificial neural networks with a variable number of hidden neurons and a reinforcement-learning scheme that is capable of both finding the ground state and describing the unitary time evolution of complex interacting quantum systems.
Hybrid convolutional neural network and projected entangled pair states wave functions for quantum many-particle states
Neural networks have been used as variational wave functions for quantum many-particle problems. It has been shown that the correct sign structure is crucial to obtain highly-accurate ground state
Optimizing large parameter sets in variational quantum Monte Carlo
We present a technique for optimizing hundreds of thousands of variational parameters in variational quantum Monte Carlo. By introducing iterative Krylov subspace solvers and by multiplying by the
CosmoFlow: Using Deep Learning to Learn the Universe at Scale
  • Amrita Mathuriya, D. Bard, +14 authors Victor Lee
  • Physics, Computer Science
    SC18: International Conference for High Performance Computing, Networking, Storage and Analysis
  • 2018
To the knowledge, this is the first large-scale science application of the TensorFlow framework at supercomputer scale with fully-synchronous training and enhancements enable us to process large 3D dark matter distribution and predict the cosmological parameters ΩsubM/sub, σsub8/sub and nsubs/sub with unprecedented accuracy.
First-principles calculations of electron states of a silicon nanowire with 100,000 atoms on the K computer
Unprecedented simulations on the electron states of silicon nanowires with up to 107,292 atoms carried out during the initial performance evaluation phase of the K computer being developed at RIKEN are reported.
Quantum spin liquids: a review.
This review discusses the nature of such phases and their properties based on paradigmatic models and general arguments, and introduces theoretical technology such as gauge theory and partons, which are conveniently used in the study of quantum spin liquids.