GPU-acceleration of tensor renormalization with PyTorch using CUDA

被引：5
|
作者：
Jha, Raghav G. [1 ]
Samlodia, Abhishek [2 ]
机构：
[1] Thomas Jefferson Natl Accelerator Facil, Newport News, VA 23606 USA
[2] Syracuse Univ, Dept Phys, Syracuse, NY 13244 USA
来源：
COMPUTER PHYSICS COMMUNICATIONS | 2024年 / 294卷
关键词：
Tensor; Lattice field theory; Spin models; CUDA; !text type='Python']Python[!/text; Speedup;
D O I：
10.1016/j.cpc.2023.108941
中图分类号：
TP39 [计算机的应用];
学科分类号：
081203 ; 0835 ;
摘要：
We show that numerical computations based on tensor renormalization group (TRG) methods can be significantly accelerated with PYTORCH on graphics processing units (GPUs) by leveraging NVIDIA's Compute Unified Device Architecture (CUDA). We find improvement in the runtime (for a given accuracy) and its scaling with bond dimension for two-dimensional systems. Our results establish that utilization of GPU resources is essential for future precision computations with TRG. Published by Elsevier B.V.
引用
收藏
页数：8
相关论文
共 50 条
  • [21] GPU-acceleration of waveform relaxation methods for large differential systems
    Dajana Conte
    Raffaele D’Ambrosio
    Beatrice Paternoster
    Numerical Algorithms, 2016, 71 : 293 - 310
  • [22] GPU-acceleration of neighborhood-based dimensionality reduction algorithm EmbedSOM
    Smelko, Adam
    Krulis, Martin
    Klepl, Jiri
    16TH WORKSHOP ON GENERAL PURPOSE PROCESSING USING GPU, GPGPU 2024, 2024, : 13 - 18
  • [23] Nonrigid Image Registration Using Spatially Region-Weighted Correlation Ratio and GPU-Acceleration
    Gong, Lun
    Zhang, Cheng
    Duan, Luwen
    Du, Xueying
    Liu, Hanqiu
    Chen, Xinjian
    Zheng, Jian
    IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, 2019, 23 (02) : 766 - 778
  • [24] Computationally efficient viscoelastic flow simulation using a Lagrangian-Eulerian method and GPU-acceleration
    Ingelsten, Simon
    Mark, Andreas
    Jareteg, Klas
    Kadar, Roland
    Edelvik, Fredrik
    JOURNAL OF NON-NEWTONIAN FLUID MECHANICS, 2020, 279
  • [25] Correlation acceleration in GNSS software receivers using a CUDA-enabled GPU
    Liangchun Xu
    Nesreen I. Ziedan
    Xiaoji Niu
    Wenfei Guo
    GPS Solutions, 2017, 21 : 225 - 236
  • [26] Correlation acceleration in GNSS software receivers using a CUDA-enabled GPU
    Xu, Liangchun
    Ziedan, Nesreen I.
    Niu, Xiaoji
    Guo, Wenfei
    GPS SOLUTIONS, 2017, 21 (01) : 225 - 236
  • [27] Smooth Particle Hydrodynamics GPU-Acceleration Tool for Asteroid Fragmentation Simulation
    Buruchenko, Sergey K.
    Schaefer, Christoph M.
    Maindl, Thomas I.
    14TH HYPERVELOCITY IMPACT SYMPOSIUM (HVIS 2017), 2017, 204 : 59 - 66
  • [28] Gaggle: Genetic Algorithms on the GPU using PyTorch
    Fenaux, Lucas
    Humphries, Thomas
    Kerschbaum, Florian
    PROCEEDINGS OF THE 2023 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE COMPANION, GECCO 2023 COMPANION, 2023, : 2358 - 2361
  • [29] GPU-acceleration of stiffness matrix calculation and efficient initialization of EFG meshless methods
    Karatarakis, A.
    Metsis, P.
    Papadrakakis, M.
    COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING, 2013, 258 : 63 - 80
  • [30] Improving GPU Throughput through Parallel Execution Using Tensor Cores and CUDA Cores
    Ho, Khoa
    Zhao, Hui
    Jog, Adwait
    Mohanty, Saraju
    2022 IEEE COMPUTER SOCIETY ANNUAL SYMPOSIUM ON VLSI (ISVLSI 2022), 2022, : 223 - 228
12345