Turning Quantum States into Music at Australia’s First Ever Qiskit Hackathon
By Robert Davis and Desiree Vogt-Lee
Quantum computing is notoriously counterintuitive; it challenges us to grapple with concepts that can be difficult to imagine. We often rely on our sense of sight to make those concepts a little easier to grasp, by representing quantum information with visualization models like the Q-sphere or the circuit diagram, and even creative visual arts projects like the recent Quantum Circuit Disks series. But what happens when we represent quantum using not only imagery, but also sound?
One team of Australian researchers is showing the world exactly what that looks like with a project that turns quantum circuits into music videos. That project, which the creators have named “qMuVi” (“quantum Music Video”), earned the titles of both 1st place winner and Community Choice winner at the recent Qiskit Hackathon Melbourne, a hybrid in-person and virtual event held in early July that marked the first ever Qiskit Hackathon in Australia. The event brought together 35 participants over four days to learn about quantum computing and Qiskit, and to use their new knowledge to hack together a diverse array of novel quantum computing projects. The event as a whole was a tremendous success. But before we talk about that, let’s take a closer look at that winning quantum music videos project.
Quantum music videos take home the gold
At the most basic level, qMuVi is a Python library that anyone can use to transform quantum circuits into music video files. Created by researchers Gary Mooney, Harish Vallury, Yang Yang, and Gan Yu Pin, the tool allows you to “hear” how a quantum state evolves as it is processed by a quantum circuit. qMuVi does this by sampling the density matrix, the mathematical object that represents the physical quantum state at different points in time while the circuit runs.
A density matrix lets us represent quantum states as a collection of values that we can analyze to understand all the physical properties of the state — convenient for performing calculations on and extracting information about the quantum state. We can use density matrices for both “pure” quantum states, which are quantum states that give us information about an exact quantum system, and for “mixed” quantum states, which are statistical ensembles of multiple pure states.
Pure states are made up of superpositions of basis states — e.g., quantum states written in the form of 0s and 1s, or written in other basis states used in quantum computing (like + and — or i and -i). To decide which notes will represent a pure state, qMuVi translates each of the states in the superposition to an integer value, and then maps those values onto a musical scale. Each state in the superposition is assigned a note, and all together, the pure state represents a musical chord.
For example, if qMuVi samples a pure state that is equal to (|00⟩ + |11⟩)/sqrt(2), it will translate the two states in the superposition to intergers — (|0⟩ + |3⟩)/sqrt(2). This tells qMuvi to play notes 0 and 3 on the musical scale simultaneously. qMuVi uses the same approach for mixed states, but because mixed states represent ensembles of multiple pure states, qMuVi represents them as multiple chords. A similar method determines the volume at which the notes will play, and the instruments that will play them. Impressively, the qMuVi team was also able to use QUI — a quantum programming and simulation environment developed by the Hollenberg group at the University of Melbourne — to manipulate the evolution of the quantum state such that it reproduces specific songs.
To see how this all works in action, check out the showcase video created by the qMuVi team:
To learn more about the quantum music video tool, and experiment with it for yourself, be sure to visit the project Github.
Qiskit Hackathon Melbourne
Quantum music videos were far from the only project to be developed at Qiskit Hackathon Melbourne. The event officially kicked off on July 4 at the University of Melbourne’s Melbourne Connect facility. The first two days of the Hackathon centered on a series of lectures presented by speakers from The University of Melbourne, RMIT, Monash, UNSW, Quantum Brilliance and IBM Quantum. Lectures were punctuated with hands-on learning activities like the Qiskit Blocks Social Challenge — a Minecraft style escape room game where the objective is to solve quantum circuit puzzles.
The event concluded with two days of hacking, where students either joined projects pitched to them by mentors from universities across the city of Melbourne, or formed independent groups to develop their own original project idea. At the end of the event, teams presented their projects in a series of three-minute pitches to a panel of judges, who evaluated the projects on Originality, Usefulness, Community Benefit, and Presentation. (See the end of this post for the full list of Qiskit Hackathon Melbourne projects.)
Qiskit Down Under
The recent Qiskit Hackathon Melbourne event may have marked the first ever Qiskit Hackathon to take place in Australia, but the Australian Qiskit community goes back much further than that. In 2017, the University of Melbourne became one of the very first hubs in the IBM Quantum Network. Last year, Australian researchers used Qiskit and IBM Quantum hardware to to simulate a discrete time crystal, and to generate some of the largest entangled states ever created on a quantum device.
Australia has long been an enthusiastic supporter of quantum computing research, and one of its largest cities — Sydney, Australia — has earned a reputation as being home to one of the world’s highest concentrations of quantum expertise. Quantum researchers, students, hobbyists and enthusiasts in Australia have proven themselves to be a vital part of the global Qiskit community. If the talent, creativity and enthusiasm we saw at Qiskit Hackathon Melbourne are any indication, Australia will continue to be an important part of our community for many years to come. We have no doubt that quantum music videos will be just one of many amazing quantum computing projects it produces.
Qiskit Melbourne Hackathon 2022 projects
1st Place and Community Choice Winner
Expressing Quantum Computation Through Generated Art: Created a Python module that can transform Qiskit quantum circuits into music videos.
2nd Place Winner
Calibrating Error Extrapolation for Variational Quantum Algorithms: Explored the feasibility of calibrating error extrapolation on an 8 qubit water molecule model.
3rd Place Winner
Quantum Adversarial Machine Learning: Explored if QML architectures are more robust to adversarial attacks than their classical counterparts.
4th Place Winners
Quantum Procedural Generation: Used Grover’s algorithm to generate the growth of a flower and stems.
Using Quantum Machine Learning to Model Financial Wellbeing: Used quantum machine learning to predict the financial wellbeing of individuals.
A Visual Representation of Quantum Noise: Created a quantum wheel of fortune based on the noise of IBM quantum computers.
Approximating Large-Depth QNNs with a Quantum Tangent Kernel: Analysed parameter evolution in variational algorithms with large depth circuits.
Investigating the Performance of Distance 2 Heavy Hexagon Codes on IBM Devices: Investigated the properties of the smallest heavy-hexagon code capable of detecting a single error and the distance 2 code on IBM devices.
Quantum Chemistry Simulations using Entanglement Forging: Explored the technique of entanglement forging to decrease the number of qubits required for quantum chemistry simulations.
Quantum Twitter Sentiment Analysis: Developed a quantum machine learning model to classify the sentiments expressed in a dataset of Tweets.
