Roadmap

Google Quantum AI's roadmap, as its published roadmap figure states it: six technical milestones toward a large-scale, error-corrected quantum computer. The code is src/quantum/roadmap.ts, served by the roadmap tool.

Milestone	Title	Year	Physical qubits	Status
M1	Beyond classical	2019	54	achieved
M2	Logical qubit prototype	2023	10²	achieved
M3	1 long-lived logical qubit	2025+	10³	in progress
M4	Tileable module (logical gate)		10⁴	planned
M5	Engineering scale up		10⁵	planned
M6	Error-corrected quantum computer		10⁶	planned

The figure's caption says Google is currently working towards Milestone 3, a long-lived logical qubit, capable of living for 1 million computational steps with less than 1 error.

Applications by qubit scale

Google's roadmap figure predicts which applications each scale of physical qubits, with quantum error correction, would unlock. The same bands are served by the roadmap tool.

Physical qubits	Predicted applications
50k–100k	Fundamental physics (quantum gravity, many-body states); certified randomness
100k–500k	Quantum magnets; spectroscopy (NMR, ESR); quantum machine learning with "quantum data"; novel phases of matter
500k–5M	Molecular modeling (drug targets and anti-targets, homogeneous catalysis, solar cell dye molecules); topological data analysis; some differential equations; quantum machine learning of "classical data"
5M–50M	Molecular modeling (materials for qubits, battery cathodes, heterogeneous catalysis, fusion reactor design); factoring

Progress, as Google states it

Google has been working on quantum computing since 2014. It announced the first beyond-classical performance in 2019 (Milestone 1), and the first experimental demonstration of scaling quantum error correction in 2023 (Milestone 2). Google plans to tile thousands of surface-encoded logical qubits into a fully fault-tolerant quantum computer.

What quantum computing is, as Google states it

Quantum computers use quantum physics to access different computational abilities from those of classical computers. A qubit can take the value of a complex combination of 0 and 1, while a classical bit stores either 0 or 1. Quantum algorithms rely on interference to amplify correct answers and cancel incorrect ones, and Google expects quantum computing to tackle certain problems that are difficult or impossible for the largest classical supercomputers.