Tangelo v0.4.2 available

Hello, Tangelo users !

Holiday season is in full swing, and our modest present to you is that Tangelo v0.4.2 is now available. We thank authors who cited this project in their recent publications and provided us with feedback to further improve Tangelo. We strive to develop an engine to accelerate research in our field and help researchers leverage the best of what the quantum community has to offer.

Your feedback, bug reports, feature requests and overall contributions are instrumental in this process, and are the cornerstone of successful open-source software. We encourage you to reach out whenever you cannot find the tools you need, face performance issues or come across unexpected errors in your code. Not only it may steer the development of Tangelo in the direction your research needs, but it may also save you significant time and effort that would be spent writing and maintaining your own code otherwise.

In the meantime, we hope you enjoy the changes brought by this new release. Happy holidays !

Credits

This release was possible thanks to the contributions of @JamesB-1qbit @AlexandreF-1qbit @ValentinS4t1qbit and the feedback from Tangelo users. Thank you !

Highlights of this release:

QM/MM notebook

Our previous release included an implementation of Quantum Mechanics / Molecular Mechanics (QM/MM), a computational technique that enables the treatment of a region of a molecular system at a given level of quantum chemistry, while the remainder is modeled with force fields. One of the key and useful applications of this framework is to consider chemical environmental effects, such as explicit solvent molecules, with a negligible computational overhead. It is now illustrated in this tutorial notebook.

iFCI fragment import

In 2022, the incremental Full Configuration Interaction (iFCI) method was massively scaled on QEMIST Cloud to deliver record-breaking results on the simulation of PFAS molecules, known as “the forever chemicals” since they do not generally biodegrade in the environment. This achievement was reported on the AWS HPC blog and Wired, among others platforms.

Tangelo now supports the import of iFCI fragments resulting from a QEMIST Cloud calculation, so that users can explore the potential of quantum computing in conjunction with methods designed to scale to more industrially-relevant use cases. Find more about it in this tutorial notebook.

Improved support for active space selection

Active space selection is a process in chemical modeling that reduces computational requirements while retaining accuracy, by strategically choosing the molecular orbitals that contribute to calculations. On top of the built-in frozen core method and the ability for users to manually specify a frozen orbital list, Tangelo now supports:

• Frozen Natural Orbitals (FNO) as a built-in method
• Active space computed by tools such as SCINE autoCAS, as shown in this example

Check out our updated tutorial.

iQPE Solver and measurement-controlled gates

Tangelo users are now able to simulate adaptive circuits with measurement-controlled gates. This opens up a class of algorithms that can implement processes that can not be done using only unitary gates. It is another useful building block for those of you researching fault-tolerant approaches, among others we had highlighted in one of our hands-on tutorials. We used this feature to expand our Quantum Phase Estimation (QPE) framework and provide an implementation of iQPE (named IterativeQPESolver).

DMET and VQE performance optimizations (part 1)

Quite a few of you have used DMET and VQE approaches in your recent work. Your feedback allowed us to provide significant performance improvements to these workflows, mostly based on cacheing and reusing previous calculations.

• Significant performance improvements in VQESolver.get_rdm method, used in DMET + VQE.
• In the near future, we'll push performance improvements for VQE to speedup updates in the variational ansatz quantum circuit (way less time spent in circuit translation and UCC-like ansatz updates).

Misc

• Feature: UHF symmetry reference labels
• Feature: IBMConnection now supports target instance
• Feature: trim qubits flag for circuit.split
• Bugfix: adapt to new qiskit version for noisy simulation
• Bugfix: DMET fix for solvers and fragment object initialization with options
• Bugfix: trim_trivial_qubits