Hands-On Quantum Error Correction with Google Quantum AI

Released by Google Quantum AI in late 2024, this course teaches quantum error correction from the ground up using the same software tools Google’s researchers use in the lab. Instructor Austin Fowler is a pioneer of the surface code and one of the architects of Google’s quantum error correction program.

The course is free to audit on Coursera and is notable for its direct connection to Google’s landmark 2024 result demonstrating below-threshold error correction with the surface code. You learn the theory behind exactly the error correction scheme that was experimentally validated on Google’s Willow processor.

What you’ll learn

• Quantum computing fundamentals: an overview of qubit technologies, the quantum circuit model, and why error correction is the central challenge for scaling quantum computers
• Quantum states and circuits: superposition, entanglement, quantum gates, and measurement reviewed with an eye toward error correction requirements
• Quantum errors: the types of errors that occur on real hardware (bit-flip, phase-flip, depolarizing noise), why classical error correction ideas do not directly apply, and the no-cloning theorem as a constraint
• Error detection and syndrome measurement: ancilla qubits, parity checks, and how to extract error information without collapsing the logical state
• Stabilizer formalism: the Pauli group, stabilizer groups, stabilizer states, and how stabilizer codes encode logical qubits into physical ones
• The surface code: the geometry of the surface code, its X and Z checks, logical operators, the threshold theorem, and why the surface code is the leading candidate for fault-tolerant quantum computing
• Stim: Google’s open-source stabilizer circuit simulator, designed for the scale of error correction experiments, with hands-on coding exercises
• Crumble: Google’s interactive tool for designing and visualising error correction circuits, used directly in the lab by Google researchers

Course structure

The course is organized into modules that build from quantum basics through error detection to the full surface code and its software implementation. Coding labs are integrated throughout using Stim and Crumble, which run in the browser via Google Colab.

No prior knowledge of error correction is assumed. Some familiarity with quantum circuits is helpful, and the course includes a review of the necessary background.

Who is this for?

• Quantum computing students and researchers who want to understand the path to fault-tolerant quantum computing
• Software engineers interested in quantum error correction software and simulation
• Anyone who followed Google’s 2024 error correction milestone and wants to understand the underlying technique in depth
• Developers wanting hands-on experience with Stim, the standard tool for error correction circuit simulation

Prerequisites

Basic familiarity with quantum states, circuits, and linear algebra is helpful but not strictly required. The course includes a self-contained review of quantum computing fundamentals. High school mathematics and some programming experience are sufficient to start, though comfort with Python helps for the lab exercises.

Hands-on practice

Labs are conducted in Google Colab notebooks using Stim and Crumble:

• Simulate error correction circuits in Stim and measure logical error rates
• Use Crumble to design and inspect surface code layouts interactively
• Implement a repetition code and verify that it corrects bit-flip errors
• Build the syndrome measurement circuit for the surface code from scratch
• Decode syndromes using a minimum-weight perfect matching decoder
• Observe the threshold behaviour: logical error rate decreasing as code distance increases

Why take this course?

Quantum error correction is the critical technology between today’s NISQ devices and the fault-tolerant quantum computers that will deliver transformative results. This course teaches the specific error correction scheme that Google has demonstrated experimentally, using the exact software tools the research team uses.

Austin Fowler is the inventor of the topological surface code threshold estimate and has spent his career at the frontier of this field. His explanations carry the authority of someone who built the technology being taught. The connection to Google’s 2024 milestone makes the material feel live and relevant rather than theoretical.