Researchers from the University of Innsbruck have developed a new model to program quantum computers using a machine learning generative model to find the exact sequence of quantum gates to execute a quantum operation. This research was recently published in the journal Nature Machine Intelligence.
Image Credit: NicoElNino/Shutterstock.com
One of the most significant advances in Machine Learning (ML) in recent years has been the use of generative models such as diffusion models, like Dall.e and Stable Diffusion have completely changed the area of picture production. Based on a text description, these models are able to generate images of excellent quality.
Our new model for programming quantum computers does the same but, instead of generating images, it generates quantum circuits based on the text description of the quantum operation to be performed.
Gorka Muñoz-Gil, Department of Theoretical Physics, University of Innsbruck
Finding the right quantum gate sequence is necessary to prepare a specific quantum state or carry out an algorithm on a quantum computer. Due to the peculiarities of the quantum environment, this is a significant issue in quantum computing, even though it is very simple in classical computing.
Recently, numerous scientists have put forth strategies for creating quantum circuits, many of which rely on machine learning techniques.
However, training these machine learning models is often challenging due to the need to simulate quantum circuits during the learning process. Diffusion models are trained to avoid these kinds of issues.
This provides a tremendous advantage, moreover, we show that denoising diffusion models are accurate in their generation and also very flexible, allowing to generate circuits with different numbers of qubits, as well as types and numbers of quantum gates.
Gorka Muñoz-Gil, Department of Theoretical Physics, University of Innsbruck
Additionally, the models can be customized to build circuits that account for the quantum hardware's connectivity, that is, the qubits' connections within the quantum computer.
“As producing new circuits is very cheap once the model is trained, one can use it to discover new insights about quantum operations of interest.”
Gorka Muñoz-Gil, Department of Theoretical Physics, University of Innsbruck
Gil developed this method together with Hans J. Briegel and Florian Fürrutter.
The University of Innsbruck's approach creates quantum circuits that are customized to the characteristics of the quantum hardware they will be operating on and based on user specifications. In terms of realizing the full potential of quantum computing, this is a major advancement.
This research was funded by the Austrian Science Fund FWF and the European Union, among others.
Journal Reference:
Fürrutter, F., et al. (2024) Quantum circuit synthesis with diffusion models. Nature Machine Intelligence. doi.org/10.1038/s42256-024-00831-9