Optimizing Ancilla-Based Quantum Circuits with SPARE
summary
SPARE is a rewrite-based optimizer that detects and restructures compute–uncompute (CTU) patterns in QPL-lowered circuits and uses ancilla state propagation to collapse temporary qubits. The paper proves correctness of the rewrites and evaluates SPARE on Unqomp/Spire-generated circuits, reporting up to 27.3% fewer qubits, 56.7% fewer 2-qubit gates, 68.2% fewer 1-qubit gates and 73.9% less depth versus Unqomp and comparable large reductions versus Spire and gate-level tools.
AI-written summary · about
chapters (15)
- 0:00SPARE introduction
- 1:38Quantum language limits
- 2:39Uncomp compute pattern
- 3:20Tower control flow
- 4:21Inefficiencies in circuits
- 5:24Optimizing gate swaps
- 6:47Merging gate patterns
- 7:49Ancilla rewrite equality
- 9:14Proving gate deletion
- 9:57SPARE compiler workflow
- 11:00SPARE vs other tools
- 12:21Gate-level vs program-level
- 13:46Abstraction vs optimization
- 14:49Key takeaways
- 15:32Q&A
related talks
37mResource-Efficient Quantum Computing by Breaking Abstractions
24mSynthesizing Quantum-Circuit Optimizers
15mAssertion-Based Optimization of Quantum Programs
19mEquality Saturation for Quantum Circuit Optimization
20mCompilation of Qubit Circuits to Optimized Qutrit Circuits
24mA Verified Optimizer for Quantum Circuits & Verified Translation Between Low-Level Quantum Languages
17mCompiling Conditional Quantum Gates without Using Helper Qubits
22mHow to Synthesize Quantum-Circuit Optimizers