Development of the ITensor software
The ITensor library supports the productive development of robust and efficient software based on tensor networks. The innovative design of the ITensor library lets users focus on the connectivity of tensor networks instead of lower-level details and is modeled on tensor diagram notation.
ITensor has a multi-layered design including basic dense tensors; sparse tensor types; sophisticated handling of quantum number symmetries; and high-level algorithms for matrix product states and general tensor networks.
In addition to providing ongoing support for users, there is a major effort at CCQ to expand ITensor’s capabilities to make handling complex quantum systems easier, to support new application domains, and to incorporate the latest algorithmic developments and tensor network formats.
Project Leaders: Miles Stoudenmire, Matt Fishman
Project Scientists: Joseph Tindall, Olivier Gauthé, and Lukas Devos