Academy and learning paths

Academy and research learning paths

How to navigate

The academy should be a prerequisite/equivalence graph, not a forced sequence or a list of every planned notebook. Learners can enter as beginners, computational physicists, engineering students, advanced quantum researchers or HPC scientists.

Track	Foundations	Progression and capstone
Scientific foundations	Python, linear algebra, numerics, uncertainty	verified classical reference, uncertainty and reproducibility
Classical HPC	CPU, profiling, parallel algorithms, MPI	Slurm, GPU portability, strong/weak scaling, HPL measurements
Quantum computing core	states, gates, channels, circuits, measurement	simulation, noise, transpilation, ISA, Hamiltonians, QEC
Algorithms	Grover, QFT, phase estimation, VQE, HHL	error/resource analysis, hardware-aware comparisons
QML / neural networks	classical ML baselines, kernel methods	QCNN/QSVM/QINN, trainability, data encoding and rigorous benchmarks
Finance	probability, derivatives, PDE, portfolio optimization	classical pricing/optimization -> HPC/ML -> quantum comparisons
Nuclear / HEP / medicine	Hamiltonians, nuclear methods, Monte Carlo	lattice/spin/field models, nuclear simulation, validated domain cases
CFD / PDE	finite difference/volume/element, solvers	HPC benchmark -> quantum linear systems/resource estimation
Photonics / quantum optics	Maxwell/Jones/diffraction, detectors	PhotonicsLab bench, coherence, single photons, QKD, noise
QPU modalities	superconducting, ions, neutral atoms, photonic, spins	native SDK, controls, error models, QEC and resource estimates

Hardware and software tracks should retain distinct foundations: classical engineering first, then quantum modeling, then resource/performance evidence. Planned modality tracks include analog/digital ions and neutral atoms, superconducting circuits, photonics, semiconductors, dual rail and qudits. QEC includes surface and LDPC/bicycle-type codes and relevant fault-tolerance overhead.

Availability labels

Committed now: the intro notebook, Bell CPU/GPU/QPU reference notebook where dependencies permit, three computational photonics N01–N03 notebooks, and core runner source. Historical/staged, not guaranteed in main: the larger Zero-to-Hero and QEC curriculum archives referred to by planning docs. Proposed: full domain academies and capstones. Count notebook files on the actual commit before asserting delivery.

Learning outcome contract

Each reviewed lesson needs audience, prerequisites, theory/derivation, CPU credential-free path, expected outputs, analytic checks, numerical convergence, tests, optional authorized accelerated/hardware path, safety/cost estimate, references and exercises. An LLM can offer tutoring but cannot certify mastery by itself.

See master academy, learning path, and curriculum ingestion policy.