Scalable Quantum Compilation With Jasp
summary
Jasp traces Crisp programs into a Jasper IR and lowers through Catalyst to MLIR/QIR, enabling full compilation of large workflows (iterative QPE, hybrid VQE, HHL, LCU) with real-time classical feedback. Resource-estimation runs via an interpreter 1 billion times faster than Qiskit and reports roughly an order‑of‑magnitude fewer CX gates; Hamiltonian simulation cuts arbitrary-angle gates by factor 20. Scaling comes from call graph compression and dynamic loop capturing, and the system emits simulators, QIR, boolean simulators and accurate resource counts.
AI-written summary · about
chapters (22)
- 0:00JASP introduction
- 1:05Crisp quantum framework motivation
- 2:58Quantum circuit programming dead end
- 7:11Crisp quantum programming overview
- 9:23Quantum compilation scalability
- 12:59High-performance Python approaches
- 17:55Jax and JASP interpreter tracing
- 21:06JASP architecture and IR
- 25:05Quantum compilation optimizations
- 27:31JASP vs Crisp programming
- 29:18JASP real-time computations
- 31:04JASP control flow
- 32:13Dynamic loop capturing performance
- 34:00Call graph compression in JASP
- 36:29Measurement-based uncomputation
- 37:36Quantum phase estimation
- 38:21NISQ hybrid algorithm compilation
- 39:25HHL algorithm compilation
- 40:30JASP reinterpretation infrastructure
- 42:16Boolean simulation and resource estimation
- 46:16JASP summary and future
- 47:00Q&A
related talks
26mQrisp: A Framework for Compilable High-Level Programming of Gate-Based Quantu...
16mCompiling and Running High-level Quantum Programs
20mJKQ: JKU Tools for Quantum Computing
01h 10mThe Rise & Collapse of a Quantum State
17mQuantum Applications with Classiq
22mSimuQ: A Framework for Programming Quantum Hamiltonian Simulation with Analog Co...
31mQuantum compilation using the ZX calculus
19mExtracting maximum value from today's quantum computers while prep