Quantum software and algorithms research

About

The QSAR group works on designing and implementing the software that enables us to write and run algorithms on quantum computers. It was established in 2022 in the Department of Electrical and Computer Engineering at UBC, and is a part of Quantum BC.

News

(2023-09-18) Gabe presents our work on using qutrits for QAOA at IEEE Quantum Week in Bellevue, WA.
(2023-09-01) Jacky, Nishanth, and Ritu join the group, and Gabe begins his new role as a graduate student.
(2023-07-23) Abhishek presented the talk Towards automated quantum circuit optimization with graph-based deep reinforcement learning at the 5th International Workshop on Quantum Compilation.
(2023-06-10) Our collaborative work with IIT Roorkee scientists about predicting the neutron drip line with quantum computing was posted on the arXiv.
(2023-04-18) The QSAR group is excited to announce the award of an NSERC Alliance Quantum grant to work on differentiable quantum transforms with our industrial partner Xanadu.
(2023-01-26) Abhishek presented the talk Quantum-assisted generative models for high-energy physics at the Canadian Graduate Quantum Conference.
(2023-01-01) Mushahid joins the group as its first PhD student.
(2022-11-28) Quantum computing fidelity susceptibility using automatic differentiation was published in Physical Review A.
(2022-11-13) Gideon Uchehara presented work on rotation-inspired circuit cutting for the SC22 Third International Workshop on Quantum Computing Software.
(2022-09-01) Abhishek joins the group as its first MASc student.

Research

Group members work on a broad array of research projects which typically fall into one of the following categories: quantum software and compilation, noise characterization and mitigation, and quantum algorithm development.

Software and compilation

Quantum compilation is the process of translating a high-level description of a quantum algorithm into a set of instructions that is executed on hardware. It's a process with many moving parts that typically involve solving computationally hard problems.

Areas of investigation include:

Developing automated compilation tools that scale into the 100s of qubits or more
Optimizing techniques for cutting and partitioning large quantum circuits
Exploring how machine learning and can help us select the best compilation techniques and devices for an algorithm
Leveraging differentiable quantum programming to learn how to compile quantum circuits, and do so in a way that preserves differentiability

Ultimately, we are interested in doing the "hard part" (such as the optimization process in the gif below), so that users of quantum software don't need to worry about what happens under the hood, and can focus instead on writing quantum algorithms.

Graphic of the quantum compilation stack.

Gif of optimizing a quantum circuit.

Noise characterization and mitigation

In order to improve the operation of our quantum computers, we need good tools to learn about and characterize their behaviour, and quantify how well they work. We are interested in:

Designing adaptive methods for quantum tomography that can learn to characterize and mimic the behaviour of processors
Applying noise characterization and calibration data to improve compilation and algorithm execution
Optimizing the application of error mitigation processes and exploring how they interact with algorithmic components such as compilation and gradient computation.

Gif of optimizing to learn
overrotation angles.

Quantum algorithms

Differentiable quantum transforms being used in an error mitigation pipeline.

Compared to classical computing, there are relatively few known quantum algorithms, and even fewer that will one day achieve the substantial speedups needed to solve life-sized problems.

Our group works on both co-designing and implementing software that facilitates our reasoning about algorithms, and exploring how it can be used to develop new algorithms and applications.

Projects our members are working on include:

Developing qutrit and qudit simulation tools for the open-source framework PennyLane
Exploring the use of higher-level quantum systems like qutrits and qudits for variational algorithms
Applying quantum computing and quantum machine learning to problems in nuclear, particle, and condensed-matter physics
Development and implementation of new differentiable quantum transforms

People

Current members

Abhishek Abhishek (MASc student)
Gabe Bottrill (MASc student)
Gideon Uchehara (co-supervised PhD student)
Jacky Jiang (MASc student)
Mushahid Khan (co-supervised PhD student)
Nishanth Baskaran (PhD student)
Ritu Thombre (MASc student)

Positions

For prospective graduate students

I am not currently accepting applications from prospective graduate students for a 2024 start. Please check back at a later date.

For UBC undergraduate students

The QSAR Lab has two positions available for undergraduate researchers in summer 2024:

Benchmarking reinforcement learning models for quantum circuit optimization (requires proficiency in programming, and experience with machine learning)
Building an interactive quantum debugger (requires proficiency in programming, and experience with React JS and front-end web development)

If you are interested, please contact Olivia directly at olivia @ ece.ubc.ca.

Group code of conduct

All QSAR group members are expected to uphold the following codes of professional and scientific conduct. The text below is partially based on CoCs of the Tropini, Avasthi, and Willis labs.

The CoC was last reviewed in September 2023. As a group, we will review and update it together on at least a yearly basis (in particular, following an influx of new members) so that everyone has the opportunity to contribute their ideas.

Professional conduct

You work with the group, not for the group. All group members, regardless of their academic status, are independent researchers in training. Everyone comes with their own skill set and experience, and contributes their unique perspective to the work that we do. We work together and do not compete with each other.
Be professional with everyone. Treat other group members and your shared space with respect. Ensure everyone has the chance to speak during group meetings, and be mindful of interrupting or talking over others. Our work hours are flexible, so be considerate of the time and schedules of your colleagues. When you are collaborating with others, giving a talk, or at a conference, you are out representing our group. Conduct yourself in the same professional manner as you would within the group, regardless of where you are and who you are with, and follow the CoCs in place at events.
Harassment will not be tolerated. There is zero tolerance for exclusionary comments or jokes, threats, or violent behaviour or language. Offensive behaviour or comments of any kind relating to gender, gender identity and expression, sexual orientation, disability, mental illness, neuro(a)typicality, physical appearance, body size, age, race, ethnicity, religion, lifestyle choices, etc. are not welcome in our group. Please do not take photographs of others, share their work, or post personally-identifying information online without their explicit permission.
Ask for help when needed, and provide it willingly. Everyone is here to learn, so ask lots of questions! When you receive questions, respond to them respectfully without putting the other person down; focus on helping the other person learn how to do something, rather than simply doing it for them. Be mindful of people's time: if your question is something a quick online search will resolve, do so before reaching out.
Provide constructive feedback. Feedback is meant to help someone improve. Identify errors when they occur, and make direct and honest suggestions where appropriate (including to the group lead!). Make feedback about the work, not the person. This includes discussions about the work of others.

Scientific conduct

We share our data and code. We post preprints on the arXiv so that our work is available to everyone. We publish our source code and data under permissive licenses on GitHub or other code-sharing platforms. We also prioritize the use of open-source libraries in our work, and contribute back to them when we can.
We conduct our work with integrity. Never falsify, make omissions, or otherwise manipulate your code or data. Our work should be reproducible by others in the field, and we should provide all that is needed to enable that.
We collaborate. This includes with other research groups at UBC and beyond, companies, and online communities. We treat confidential information obtained during collaborations with utmost care. If you wish to establish collaborations with an external party, great! Please initiate a discussion with the group lead first.
We provide proper attribution. Always cite your sources and images in your written work and presentations, even if that source is yourself. When using external libraries in your project, make sure to check the licenses.
It's okay to make mistakes. Everyone makes them, this is how we learn. It is important to be open and honest about them, accept responsibility, and we will work together to resolve them.