Machine learning method generates circuit synthesis for quantum computing

Researchers from the University of Innsbruck have discovered a new method to prepare quantum operations on a given quantum computer, using a generative machine learning model to find the right sequence of quantum gates to execute a quantum operation.

The study, recently published in Nature’s machine intelligencemarks a significant step forward in realizing the full scope of quantum computing.

Generative models such as diffusion models are one of the most important recent developments in machine learning (ML), with models such as Stable Diffusion and DALL·E revolutionizing the field of image generation. These models are able to produce high quality images based on the text description.

“Our new model for programming quantum computers does the same thing, but instead of generating images, it generates quantum circuits based on the textual description of the quantum operation to be performed,” explains Gorka Muñoz-Gil from the Department of Theoretical Physics of the University. from Innsbruck, Austria.

To prepare a particular quantum state or to execute an algorithm on a quantum computer, one must find the right sequence of quantum gates to perform such operations. While this is fairly easy in classical computing, it is a major challenge in quantum computing, due to the peculiarities of the quantum world.

Recently, many scientists have proposed methods to build multi-quantum circuits based on ML methods. However, training these ML models is often very difficult due to the need to simulate quantum circuits as the machine learns. Diffusion models avoid such problems because of the way they are trained.

“This offers a tremendous advantage,” explains Muñoz-Gil, who developed the new method together with Hans J. Briegel and Florian Fürrutter. “Furthermore, we show that the denoising diffusion models are accurate in their generation and also very flexible, allowing the generation of circuits with different numbers of qubits as well as types and numbers of quantum gates.”

The models can also be adapted to prepare circuits that consider quantum hardware coupling, ie. how the qubits are connected in the quantum computer. “Since producing new circuits is very cheap once the model is trained, you can use it to discover new insights about quantum operations of interest,” says Muñoz-Gil.

The method developed at the University of Innsbruck produces quantum circuits based on user specifications and tailored to the features of the quantum hardware on which the circuit will operate. This marks a significant step forward in unleashing the full scope of quantum computing.