JuliaCon 2025

DeviceLayout.jl is a package developed at the AWS Center for Quantum Computing (CQC) for computer-aided design (CAD) of quantum integrated circuits. At the CQC, we use DeviceLayout.jl to design superconducting quantum devices on our path to building a fault-tolerant quantum computer—devices we’ve used for experiments like Demonstrating a Long-Coherence Dual-Rail Erasure Qubit Using Tunable Transmons and Hardware-efficient error correction using concatenated bosonic qubits. We developed it to allow designers to produce and iterate on device layouts quickly and easily. We’ve paid particular attention to scalability in support of both larger quantum processors and a larger, collaborative team. As examples, we’ll highlight how we can use a schematic-driven workflow for layout of a 17-qubit processor in a modular, reproducible project; we’ll then show how it works together with Palace, an open-source tool for electromagnetic finite-element analysis also developed at the CQC. We have released DeviceLayout.jl on GitHub as an open-source project, where it joins Palace as part of an open-source toolchain for electronic design automation (EDA) of quantum integrated circuits and other electromagnetic devices.

The package supports the generation of 2D layouts and 3D models of complex devices using a low-level geometry interface together with a high-level schematic-driven workflow. At the geometry level, the user writes code to draw 2D geometric entities like polygons and paths, optionally using units with Unitful.jl; position and orient these entities with coordinate transformations; organize them into a hierarchy of local coordinate systems and references; apply operations like rounding, offsetting, or geometric Booleans; and assign each entity metadata that can be used in different ways by different backends. In the schematic-driven workflow, the user specifies a schematic in terms of components and the connectivity between them, followed by automatic placement and routing. The results can be rendered with various backends, including for output in the GDSII format; used to construct and mesh a 3D model; and interfaced with other tools like the open-source electromagnetics solver Palace. Finally, we leverage the Julia package manager for process design kit (PDK) management, allowing users to maintain a library of versioned process technologies and components for portable, reproducible layout scripts.