Introduction

../_images/666.png

Jet is an open-source tensor network contraction engine which aims to support the following workflows:

  • Simulate discrete and continuous-variable quantum circuits.

  • Estimate the memory and computational requirements of a simulation.

  • Contract tensor networks with customizable precision and GPU acceleration.

Follow the development guide to get Jet up and running and then have a look at the quickstart tutorials to see what you can do with Jet.

Quickstart guides

The following sections outline the key features of Jet:

  1. Tensors offers a gentle introduction to tensors and how they are modeled in Jet. Learn about tensor notation and contractions here.

  1. Tensor networks provides an overview of tensor networks and how Jet leverages task-based parallelism to contract tensor networks concurrently.

  1. Tensor network files shows how to serialize and deserialize tensor networks. This section details how to use the Jet serializer and presents a formal specification of the tensor network file format.