Q
QAtlas laboratory / local compiler

Circuit X-Ray

Inspect a circuit’s actual routing, native-gate conversion, topology pressure, and transparent independent-error estimate before it meets a device.

Shared circuit workspace

Circuit input

OpenQASM 2 static-circuit subset. Files stay in this browser; code is never executed.

Limits: 64 KB, 16 logical qubits, 512 operations. OpenQASM 3 and dynamic control are explicitly unsupported.
Target model

Compile settings

Load an example or paste OpenQASM 2, then validate it locally.

About Circuit X-Ray

See what a constrained topology changes

Circuit X-Ray is for developers, researchers, educators, and advanced students who need to understand the difference between a logical circuit and a compiled physical mapping. It parses a documented OpenQASM 2 static-circuit subset, routes nonadjacent interactions with real SWAP insertion, and shows its assumptions instead of claiming vendor calibration or hardware execution.

What is supported?

Built-in single- and two-qubit gates, controlled phase, barriers, and terminal measurements. OpenQASM 3, custom gates, dynamic control, reset, timing directives, and arbitrary uploaded code are rejected clearly.

What is estimated?

Success probability uses the selected independent error model. Generic profiles are illustrative models, not current device specifications; custom values are treated as user-supplied assumptions.

Example uses

Compare line versus grid routing, inspect a small QFT, find a congested coupling, or quantify the overhead introduced by a sparse topology.

Method and reference

The input language is grounded in the Open Quantum Assembly Language paper. QAtlas deliberately implements only its documented static subset and runs routing locally; it does not submit work to quantum hardware.

Frequently asked questions

Does Circuit X-Ray access quantum hardware?
No. Compilation, routing, and metrics run locally in your browser without vendor credentials or hardware submissions.
Does a lower estimated error prove a better device result?
No. It is a transparent model comparison under the supplied errors and does not include calibration drift, crosstalk, or all compiler optimizations.