Research & learn with Google Quantum AI software tools

The speaker, a Quantum Software PM at Google, introduces two new tools: Cirq 1.0 and the Quantum Virtual Machine (QVM). These tools aim to develop quantum programming skills and support research. Google's goal is to make these tools available to a broad ecosystem, including academics, industry researchers, students, and educators. The speaker emphasizes the mission to find new applications for quantum computers and to foster a diverse and inclusive quantum workforce. Cirq 1.0 is a major update to Google's programming framework, featuring backward compatibility, stability, and a rich set of features. The QVM offers free, unlimited access to quantum hardware simulation, making it easier to prototype and learn about quantum devices.

Dave Bacon, Head of Software Engineering at Google, discusses Cirq 1.0 in detail. Cirq is an open-source Python framework for quantum computing designed for the NISQ era. Since its introduction in 2018, Cirq has undergone numerous updates, adding new features and capabilities. Cirq 1.0 promises a stable API, semantic versioning, and a commitment to backward compatibility. New features include a Transformers API for compiling, classical control integration, and an improved gate set. The release reflects community contributions, with 169 contributors and over 3,000 commits. The talk also highlights the importance of Cirq's stability for education and research, ensuring that code written by students and researchers remains functional across updates.

Abe Asphall, head of education and outreach at Google Quantum, discusses the use of Google's quantum tools in educational settings. The goal is to enable individuals excited about quantum computing to contribute meaningfully to the field. Asphall emphasizes the ease of starting with Cirq 1.0 and the Quantum Virtual Machine (QVM), both accessible through Google Colab. The QVM simulates state-of-the-art quantum processors, allowing for rapid iteration and feedback in coursework. Updates to Cirq's documentation are highlighted, making it more accessible and user-friendly. The talk concludes with an introduction of August Earle, a PhD student and Google Quantum Educator, who discusses the importance of accessibility in quantum programming languages and the improvements made to Cirq for educational purposes.

August Earle, a PhD student at UCLA and a Google Quantum Educator, shares his experience in teaching a quantum programming course at UCLA. The course aimed to be accessible to students without prior knowledge of quantum physics or linear algebra. Earle discusses the evolution of the course, which initially covered three major quantum programming languages but eventually focused solely on Cirq due to its intuitiveness and ease of use. He highlights the improvements made to Cirq's documentation and the impact on student engagement and creativity. Earle emphasizes the importance of making quantum programming accessible to all students to spark interest and motivation in solving complex problems.

Kira Peltz, founder of the Coding School and executive director of Qubit by Qubit, discusses the initiative to introduce quantum computing to K-12 education. Qubit by Qubit aims to train a diverse quantum workforce and ensure that quantum education is accessible. Peltz shares the impact of their programs, which have reached over 15,000 students globally, with a focus on underrepresented backgrounds. The initiative with Google aims to train K-12 educators in quantum computing to integrate it into their classrooms, ensuring that all students are introduced to quantum concepts before graduating high school. Peltz highlights the potential of quantum computing to increase student interest in STEM fields.

Craig Ginley introduces STEM, a library for high-performance analysis and simulation of stabilizer circuits, particularly for error correction. He explains that STEM is designed to handle large-scale circuits that other tools might struggle with. STEM's performance is significantly faster, especially when handling multiple samples. A key feature of STEM is its ability to extract decoder configuration directly from a quantum circuit, simplifying the process of updating codes. Ginley also mentions the flexibility of STEM, allowing users to utilize its building blocks in creative ways. He encourages those interested in quantum error correction to try STEM for its efficiency and ease of use.

The speaker provides an overview of updates to Google's open-source software ecosystem, focusing on tools for quantum chemistry and machine learning. OpenFermion, an electronic structure package for quantum computers, and the Fermionic Quantum Emulator (FQE), which simulates electron behavior efficiently, are highlighted. FQE's ability to reduce simulation time and memory requirements is emphasized. Additionally, TensorFlow Quantum is updated with new features, including matrix product state circuit simulation, which enables more efficient simulations for quantum circuits. The talk concludes with an invitation to learn more about these tools directly from the team and to explore the software packages and resources available.