Equality Saturation for Quantum Circuit Optimization
summary
Quasar builds a compiler pass that constructs two complementary e-graphs (graph and sequence views), infers a compact 40 atomic rewrites set from thousands, and guarantees a step-limited optimal circuit under a bound on rewrite iterations. It implements equality saturation with new soundness and scalability techniques, runs on IBM Eagle benchmarks, and reports geometric-mean improvements: 20.2% fewer 2-qubit gates, 33.5% fewer total gates, 24.2% less depth and 21.4% higher fidelity, outperforming state‑of‑the‑art on most circuits.
AI-written summary · about
chapters (13)
- 0:00Quasar and equality saturation
- 1:13Quantum circuit optimization
- 2:39Current rewrite methods
- 3:20Sequential optimization limits
- 5:07Phase ordering problem
- 6:53Quasar and equality saturation
- 9:43Rewrite redundancy challenges
- 11:27Atomic rewrite rules benefits
- 13:15Quasar benchmark evaluation
- 15:03Quasar’s contributions summary
- 16:04Q&A: Initial rewrites
- 17:28Q&A: Probabilistic rewrites
- 18:11Q&A: Clermont comparison
related talks
24mSynthesizing Quantum-Circuit Optimizers
22mHow to Synthesize Quantum-Circuit Optimizers
32mOptimal Quantum Programs
39mBuilding a Compiler for Quantum Computers
19mExtracting maximum value from today's quantum computers while prep
20mCompilation of Qubit Circuits to Optimized Qutrit Circuits
15mAssertion-Based Optimization of Quantum Programs
18mAutomated distribution of quantum circuits via hypergraph partitioning