Qadence: a differentiable interface for digital-analog programs
summary
Qadence is a high‑level, differentiable interface that composes Hamiltonian block analog evolutions with single‑qubit digital gates and compiles them to backends (simulator or QPU). It embeds circuits as PyTorch models so variational parameters use the automatic‑diff engine and a generalized parameter‑shift rule for digital‑analog operators. The system targets Rydberg neutral‑atom hardware with reconfigurable registers, aims for near‑term executability on devices scaling toward 10,000 atoms, and is already used internally at Pasqal.
AI-written summary · about
chapters (11)
- 0:00Qadence intro and paradigm
- 2:10Digital vs analog Quantum Computing
- 3:55Qadence’s hardware-agnostic framework
- 8:14Qadence core architecture
- 10:28Building Quantum circuits with blocks
- 13:15Hamiltonian programming in Qadence
- 15:25Differentiable parameterized circuits
- 17:38Summary and future plans
- 19:06Q&A: measurement-based Quantum Computing
- 19:48Q&A: abstraction level for devs
- 21:33Q&A: hardware specificity
related talks
17mQuantum Applications with Classiq
40mQ# - going beyond quantum circuits
27mA Haskell Adiabatic DSL: Solving Classical Optimization Problems on Quantum Hardware
16mGeqo - A Sovereign End-to-End Software to Design, Build and Execute Quantum Programs
28mAnalog and Digital Quantum Computing with Neutral Atoms
18mType-safe (Variational) Quantum Programming in Idris
14mQuantum approach to classical optimization: why bother and what to do?
19mExtracting maximum value from today's quantum computers while prep