Automated distribution of quantum circuits via hypergraph partitioning
summary
Claims an automated compiler pass that maps a quantum circuit to K QPUs while minimizing costly entangled pairs (ebits) by encoding multi-gate grouping into a hypergraph. It builds a hypergraph in O(#C-sets) (2× per control set), calls off‑the‑shelf hypergraph partitioning, and decodes partitions into distributed operations using C-set grouping and staged teleportation. Prototype implemented in Cooper and evaluated on five benchmarks, cutting ebit use up to ~50% for ground‑state estimation; partitioning quality depends on the NP‑hard backend solver.
AI-written summary · about
chapters (12)
- 0:00Intro
- 0:19Distributed quantum architectures
- 2:02Compiler perspective
- 2:41Quantum communication
- 3:53Implementing non-local CZs
- 7:19Problem definition
- 8:33Previous work
- 11:17From circuit to hypergraph
- 11:52From partition to distribution
- 13:10Connectivity changes over time
- 13:51Test suite
- 15:11CZ-grouping advantage
related talks
17mQVM: Quantum Gate Virtualization Machine
19mQuantum Virtual Machines
14mDependency-Aware Compilation for Surface Code Quantum Architectures
39mBuilding a Compiler for Quantum Computers
50mMaking the Leap to Quantum PCPs
19mEquality Saturation for Quantum Circuit Optimization
30mQuantum Constant Propagation
21mSoftware for Quantum Computing- Easing Solutions for Complex