cuTensorNet: A High-Performance Library for Tensor Network Computations

Welcome to the cuTensorNet library documentation!

NVIDIA cuTensorNet is a high-performance library for tensor network computations, a component of the NVIDIA cuQuantum SDK. Functionalities of cuTensorNet are described in Overview with installation and usage guide provided in Getting Started.

Key Features

  • Based on NVIDIA’s high-performance tensor algebra library: cuTENSOR

  • Provides APIs for:

    • Creating a tensor or tensor network object

    • Finding a cost-optimal tensor network contraction path for any given tensor network

    • Finding a low-overhead slicing for the tensor network contraction to meet specified memory constraints

    • Tuning the tensor network contraction path finder configuration for better performance

    • Performing tensor network contraction plan generation, auto-tuning, and its subsequent execution

    • Gradually constructing a tensor network state (e.g., a quantum circuit state), followed by computing its properties, including arbitrary slices of amplitudes, expectation values, marginal distributions (reduced density matrices), as well as performing direct sampling

    • Performing backward propagation to compute gradients of the output tensor w.r.t. user-specified input tensors

    • Performing tensor decomposition using QR or SVD

    • Applying a quantum gate operand to a pair of connected (contracted) tensors

    • Enabling automatic distributed parallelization in the contraction path finder and executor

    • Enabling custom memory management

    • Logging


  • Supported GPU Architectures: Volta, Ampere, Hopper

  • Supported OS: Linux

  • Supported CPU Architectures: x86_64, ARM64, ppc64le