[Feature Request] Support for DFTHC (Double Factorized Tensor HyperContraction) Block Encoding [arXiv:2502.15882]

[TeiKim9909]

Hi Qualtran team!

I recently read the paper Fast quantum simulation of electronic structure by spectrum amplification (arXiv:2502.15882) and was really impressed by the Double Factorized Tensor HyperContraction (DFTHC) model. It mathematically demonstrates significant improvements for Hamiltonian simulation.

It would be amazing to have DFTHC block encoding supported in Qualtran. My main goal is to use it for end-to-end QPE resource estimation and circuit generation for chemical systems.

Proposed Features

Specifically, it would be great to see the following functionalities added:

Hamiltonian-to-Bloq Generation: A straightforward way to generate a DFTHC bloq directly from a molecular Hamiltonian (e.g., passing in standard 1-electron and 2-electron integrals).

QPE Resource Estimation: Hooking up the DFTHC block encoding to the existing QPE framework so we can easily pull exact logical resource estimates, like Toffoli counts and logical qubit requirements.

Gate Sequence Extraction: Support for decomposing the DFTHC bloq all the way down, allowing users to extract the exact logical gate sequences (like Toffoli or Clifford+T networks) needed to run the QPE circuit.

Reference

Guang Hao Low et al. (2025). arXiv:2502.15882

Thanks so much for building and maintaining such a fantastic library! It’s an incredibly valuable tool for the quantum computing community, and I really appreciate all your hard work.