Two-qubit depth
17.6% lower125 fewer two-qubit layers
Bars start at zero. Lower is better.Weylattice
AI-assisted qubit layout, routing and native-gate compilation for your selected hardware. Access Weylattice in Spinorio or through the Spinorio Python SDK.
Measured compilation · IBM Fez
See the measured difference between Weylattice and standard Qiskit on a five-layer QAOA circuit, compiled for IBM Fez.
125 fewer two-qubit layers
Bars start at zero. Lower is better.232 fewer two-qubit gates
Bars start at zero. Lower is better.In this run, the two-qubit reductions came with higher total gate count, higher overall circuit depth and longer compilation time. All five metrics are reported below.
All five metrics
Both outputs were checked against IBM Fez’s native gates and allowed connections. Metrics were independently recomputed from the downloaded circuits.
Download benchmark & circuits| Metric | Standard Qiskit | Weylattice | Change |
|---|---|---|---|
| Native two-qubit gates | 7,245 | 7,013 | −3.2% |
| Two-qubit depth | 711 | 586 | −17.6% |
| Total circuit depth | 1,376 | 1,662 | +20.8% |
| Total gates | 20,537 | 29,225 | +42.3% |
| Compilation time | 1.28 s | 155.35 s | +154.07 s |
The compiled circuits
Compare where the two outputs place their two-qubit operations across the same 156-qubit processor.
Click a connection for exact counts. With a map focused, use Left and Right to inspect connections and Escape to clear the selection.
Two-qubit gate activity across IBM Fez. Brighter connections indicate more operations. Click a connection for exact counts.
100 logical qubits · 150 graph edges · five QAOA layers · 1,350 input gates
AI-assisted qubit placement and routing adapt a circuit’s interactions to the selected processor’s connections.
Convert output to supported native gates, then validate every gate and two-qubit connection against the target configuration.
Inspect native two-qubit gates, depth, total gate count and compilation time against the same-source standard baseline.
The benchmark compares Weylattice with standard Qiskit transpilation using the same circuit, hardware target and optimisation level.
This is a single compilation benchmark. No physical QPU execution, fidelity measurement, MaxCut solution or formal circuit-equivalence verification was performed. Results shown are for this 100-qubit QAOA benchmark on IBM Fez.
Read the full benchmark reportAvailable in Spinorio and through the Spinorio Python SDK.
Ask the agent to compile a circuit for your selected target, or use the Spinorio Python SDK directly in a notebook.
Try Weylattice in SpinorioExplore the available configuration, comparison and output options before running a compilation.
Read the guideTo simplify the use of AI powered quantum compilation, you can use the Spinorio SDK.
View documentation