[DEMO] Hamiltonian simulation with Generalized Quantum Signal Processing

[Mostafa-Atallah2020]

General information

Name
Mostafa Atallah

Affiliation (optional)
University of Tennessee, Knoxville

Twitter (optional)
https://x.com/M_Ataallah

LinkedIn (optional)
https://www.linkedin.com/in/mostafa-atallah-540334120/

---

Demo information

Title
Hamiltonian simulation with Generalized Quantum Signal Processing

Abstract
This demo shows how to simulate the time evolution exp(-iHt) of a Hamiltonian using Generalized Quantum Signal Processing with qml.GQSP. The idea is to block-encode H as a qubitization walk and then apply a polynomial of that walk, chosen via the Jacobi-Anger expansion so that it equals exp(-iHt); the result is read off the circuit and verified against the exact evolution from scipy.linalg.expm. We also study how the accuracy improves as the polynomial degree grows. It complements the existing GQSP resource-estimation demo, which estimates the cost but never builds a runnable circuit.

Relevant links

• Motlagh & Wiebe, Generalized Quantum Signal Processing, PRX Quantum 5, 020368 (2024): https://journals.aps.org/prxquantum/abstract/10.1103/PRXQuantum.5.020368
• Low & Chuang, Hamiltonian simulation by qubitization, Quantum 3, 163 (2019): https://quantum-journal.org/papers/q-2019-07-12-163/
• PennyLane demo, Resource estimation for Hamiltonian simulation with GQSP: https://pennylane.ai/demos/tutorial_estimator_hamiltonian_simulation_gqsp
• PennyLane demo, Intro to qubitization: https://pennylane.ai/demos/tutorial_qubitization

[CatalinaAlbornoz]

Hi @Mostafa-Atallah2020 ,

Thanks for your interest in submitting a demo! What's the link to your demo? Do you have it in a GitHub repo or somewhere else?

[Mostafa-Atallah2020]

Hi @CatalinaAlbornoz I am going to submit a pr for that I was just checking if its ok to do so

[Mostafa-Atallah2020]

HI @CatalinaAlbornoz I am also sharing with you the colab notebook of the tutorial just in case you wanted to check it out.

https://colab.research.google.com/drive/1kjt6BfaG4XPFHgxBHi4FKSVjZ1ws2IcG?usp=sharing

[CatalinaAlbornoz]

Thanks for clarifying @Mostafa-Atallah2020 !

Before you spend the time making the PR for the demo please give us a couple of weeks to review the content. Sometimes the content fits better as a community demo or as a blog post so we'll assign someone to review your notebook and we'll get back to you with next steps within 2 weeks. If we haven't gotten back to you within 2 weeks please tag me on a comment here!

[CatalinaAlbornoz]

Hi @Mostafa-Atallah2020 ,

After an initial review, we think this would make a great PennyLane demo!

Developing a full demo for the main part of the website usually takes a bit of time, typically requiring 2-3 rounds of review from our team. Would you be available over the next month or so to work on suggested edits?

If you're short on time, no worries at all! We can easily publish your work as a community demo, which is a much faster process. However, if you are interested in the full route, we'd love to guide you through it.

Let us know which path works best for you. We look forward to hearing from you!

[Mostafa-Atallah2020]

Hi @CatalinaAlbornoz I can dedicate time for this and follow the guidlines from the team. would you elaborate more on next steps?

[CatalinaAlbornoz]

That's amazing @Mostafa-Atallah2020 !

Writing a PennyLane demo means that others will discover your work, easily understand it, and build upon it. Unlike papers, PennyLane demos can actually be fun! They still maintain the good scientific content but in a relaxed way that can be engaging and have personality. A demo should feel like a conversation with one of your peers. Here are some tips on how to write demos to maximize the value you and your readers get from it.

Tips for writing a PennyLane demo

1. Make sure to check out other PennyLane demos to get a sense of their general structure and style. Note that they all interleave text and code.
2. Write your demo like you're talking to a peer. Make sure all language, code and math are clear and easy to understand for an informed reader.
3. Don't go on tangents, feel free to link to other PennyLane or external content instead of explaining every single concept.
4. Always favour clear code over smart code. While writing docstring for functions is good practice, it can visually overload a demo. Try to minimize doing this by explaining what you’re doing in text and writing clear code. Think of code as paragraphs – what makes it easier to read? Typically, breaking code up into separate easy-to-understand steps, rather than trying to have a clever function that does everything, is preferred.
5. As you write the demo please keep in mind the demo review checklist below.

Demo review checklist:

• The language and style are clear and precise
• The introduction hooks the reader and clearly outlines the learning objectives
• The flow of ideas is uninterrupted, avoiding tangents
• The conclusion highlights the main learning outcomes and invites the reader to learn more
• The code can be easily understood
• The meaning of the inputs and outputs is clear
• The code runs properly/as claimed
• Graphics and plots are clear and not overloaded
• All images have descriptive captions
• Plots look professional (axes are labelled, there’s a legend, frame, and title)
• Links are clear and respond to a need
• References and links work
• All content is original or free to reuse (all contributions must be made under the Apache 2.0 license.)

Submission process

In order to submit your demo you must:

• Create a PennyLane profile at pennylane.ai (see an example here). Note that the author bio for your demo will be taken from the Headline section in your profile, so please fill this out.
• Create a fork of the PennyLane demo repository.
• Create and activate a new virtual environment.
• Run python -m pip install . at the root of the repo, and use the CLI tool command demo new. See the contributing guide in the repo for more details about how to use the tool. Running this command will create: a folder named after your demo within the demonstrations_v2 folder, a Python file (demo.py), a metadata file (metadata.json), and a requirements file (requirements.in).
• The ‘demo.py’ file contains a template using reStructured Text (reST). It can be tricky to write a demo in this format so instead we recommend you write your demo as a Jupyter notebook and convert it to reST by using our Notebook to Demo Converter.
• Once you have your demo as a Python file, rename it to demo.py and replace the original demo.py file.
• Follow the metadata guidelines to complete the metadata file.
• Include any images required for the demo (except for the thumbnails) in the ‘_static/demonstration_assets’ folder, under a folder with the name of your demo.
• Describe or sketch the demo thumbnails. You can upload the sketch in the _static/demonstration_assets folder. Xanadu’s designer will make the final version.
• Create a Pull Request to the master branch of the PennyLane demo repository.
• Let the PennyLane team know once you've made the pull request so that we can start reviewing.

Review

Our content review process consists of a few iterations on GitHub, where we will suggest changes for you to make in your demo. We will mainly look for clarity and consistency. Depending on the demo we may have one or more PennyLane reviewers. If you disagree with any of the suggestions that come up during review, please feel free to let us know and help us understand why.

We encourage you to make the Pull Request early on, even if you haven’t finished the first draft of your demo. This will allow us to provide important feedback early on, when it’s easier to change things if needed.

Marketing

Once the demo is ready, Xanadu will write one or more marketing posts to promote your demo on X/Twitter, LinkedIn, and Slack. We will tag you using the social media handles you shared. Please make sure to provide any feedback about the text for the marketing posts in a timely manner.

As always, should you have any questions just reach out to the team!

By submitting your demo, you consent to our Privacy Policy, Code of conduct, and Apache-2.0 licence.

[Mostafa-Atallah2020]

Hi @CatalinaAlbornoz, I've opened the PR for the demo: #1815 (linked to this issue). Could you please assign me on this and have the team start the initial review whenever convenient? My PennyLane profile, which the author bio is taken from, is https://pennylane.ai/profile/IpsiXboiI. I'm available to iterate over the review rounds.

[CatalinaAlbornoz]

Awesome @Mostafa-Atallah2020 ! That was super quick.

We'll assign someone to review over the next couple of weeks. If you haven't heard from us in the next ~2 weeks please tag me in a comment and I'll check on the status of the review.

Thanks again for making this demo!