Quantum computing, with the classical baseline in view
Where quantum methods stand today, measured rather than promised: what each method is, where it applies in industry, and how it compares with the classical method it would replace. Every page comes with eigenkit code to rerun the test on your own data, and every result names the run or the paper it comes from. EigenOrbit Benchmarks shows which of today's QPUs could run a given circuit and at what price, and EigenOrbit Router runs it on one.
Applications
Industry use cases, each with the classical method it has to beat.
Application
Molecular optimization with VQE: bond lengths of H₂ and LiH against exact chemistry
Can VQE find a bond length? We scanned H₂ and LiH on a simulator: exact from a Hartree-Fock start, trapped on LiH from random angles, slower than FCI.ApplicationQuantum machine learning for wind power forecasting: a hybrid quantum network against persistence and classical models
Can quantum machine learning forecast wind power? We tested a 5-qubit hybrid network on two years of Danish grid data, 1 and 6 hours ahead. How to test yours.ApplicationQuantum machine learning for fraud detection: a hybrid quantum classifier at the real fraud rate
Does quantum machine learning detect card fraud better? We tested a 5-qubit hybrid model at the real 0.17% fraud rate against classical ones. How to test yours.ApplicationQuantum reservoir computing for volatility forecasting: what the evidence says
Can a quantum reservoir forecast volatility? We tested a 6-qubit reservoir against HAR on two series and checked the published S&P 500 study. How to test yours.ApplicationQuantum computing for option pricing: three routes and when each is real
Can quantum computers price options? Amplitude estimation needs quadratically fewer samples but a fault-tolerant machine; tensor networks run on GPUs now.ApplicationQuantum portfolio optimization: the three routes, what each has shown, and what to run instead
Does quantum portfolio optimization work yet? We ran QAOA on a 14-asset selection against exact solvers and reviewed all three routes. How to test yours.ApplicationVaR and CVaR with quantum amplitude estimation: a practical guide for risk teams
Can amplitude estimation compute VaR and CVaR? We ran 7-qubit IQAE against Monte Carlo: quadratically fewer queries, not yet a shorter risk run.ApplicationBreast cancer classification with a quantum SVM: a benchmark against classical models
Can a quantum SVM classify breast cancer better? We tested 4- and 8-qubit quantum kernels against tuned classical models over 10 splits. How to test yours.ApplicationQuantum machine learning for defect detection: a quantum kernel SVM against classical models on steel plate faults
Can quantum machine learning classify steel defects? We tested 4- and 8-qubit quantum kernels against classical models on 1,941 plates. How to test yours.ApplicationQuantum machine learning for predictive maintenance: a hybrid quantum classifier against classical models on AI4I 2020
Can quantum machine learning predict machine failures? We tested a 5-qubit hybrid classifier on AI4I 2020 at the real failure rate. How to test yours.Comparisons and guides
Which method to pick, and how to run a fair comparison yourself.
Guide
Quantum computing for chemistry: what has run, what beats it today, and what it would take
Quantum computing for chemistry, by the evidence: what has run on hardware, how it compares with classical methods, and what FeMoco and drug design need.GuideQuantum simulation of materials and quantum matter: what quantum computers have done, and what classical computers matched
Quantum simulation of materials and many-body physics: every beyond-classical claim, which ones classical methods matched, and which still stand.Comparisoneigenkit, marqov, quantum-executor and SuperQuantX: four ways to run one circuit on many backends
Four open-source Python libraries that run one quantum program on many backends: what eigenkit, marqov, quantum-executor and SuperQuantX each do, and which to pick.ComparisonDoes a quantum layer help? HQNN vs a classical network on the same data
Does a quantum layer help a neural network? We tested a 5-qubit hybrid against a classical twin of the same size on two datasets. How to test yours.ComparisonMonte Carlo or quantum amplitude estimation: when the quadratic speed-up is real
Does amplitude estimation beat Monte Carlo? We measured IQAE's quadratic query advantage on a simulator, and what still stands between it and less time.ComparisonQAOA vs MILP: when an exact classical solver beats a quantum heuristic
QAOA vs MILP: a mixed-integer solver proves its optimum, QAOA samples candidates. We ran both on one 14-asset problem: what it shows and what could flip it.ComparisonQiskit vs Cirq vs PennyLane: which quantum SDK to choose
Qiskit, Cirq or PennyLane? What each quantum SDK is built for, which hardware it reaches, how the same circuit looks in all three, and how to choose.ComparisonQuantum kernel vs RBF kernel: how to run a fair QSVM benchmark
QSVM vs SVM: we tested 4- and 8-qubit quantum kernels against tuned RBF and linear SVMs on breast cancer data. How to run a fair benchmark, with code.ComparisonQuantum annealing vs QAOA: two ways to solve the same QUBO, and how to choose
Quantum annealing vs QAOA: both minimise a QUBO, one on analog annealers, one with gates. How they differ, how to choose, and a worked example on one problem.ComparisonQuantum annealing vs simulated annealing vs simulated bifurcation: what the benchmarks show
Quantum annealing, simulated annealing, simulated bifurcation, digital annealers and CIMs take the same QUBO. What benchmarks found, and how to test yours.ComparisonPennyLane vs Qiskit and other quantum machine learning frameworks, and where eigenkit sits
PennyLane vs Qiskit, CUDA-Q, Braket, TensorCircuit and Qulacs for quantum machine learning: licence, gradients, hardware, PyTorch and scikit-learn support.ComparisonSuperconducting vs trapped-ion vs neutral-atom qubits: how the types of quantum computer compare
Superconducting vs trapped-ion vs neutral-atom qubits, plus photonic and analog machines: gate fidelity, speed, connectivity and size of named QPUs today.GuideHow much does quantum computing cost? QPU prices, billing models and worked examples
How much does quantum computing cost? QPU prices from IBM, AWS Braket, Azure Quantum and IQM as of November 2026, the billing models and worked examples.GuideIs quantum machine learning useful? When it makes sense, and when it does not
Is quantum machine learning useful? For most problems on classical data today, no. Where it makes sense, and how a benchmark on your own data decides it.GuideQuantum computing use cases: where quantum computers will help first, ranked by the evidence
Quantum computing use cases ranked by evidence: chemistry, materials and physics simulation first, then codebreaking, with optimisation and ML far behind.GuideQuantum computing for logistics: routing and scheduling, and what the trials show
Routing, scheduling and loading can be written for a quantum computer. What the published trials ran, how fast the qubit count grows, and what to do today.Concepts
The ideas under quantum computing, with the circuits that show them.
Concept
Ansatz: the trial circuit behind every variational quantum algorithm
An ansatz is the parametrised circuit a variational quantum algorithm searches over. The main types, the trade-off between them, and how to choose one.ConceptBarren plateaus: why quantum neural networks stop training
A barren plateau is a training landscape whose gradients vanish exponentially as qubits are added. What causes it, how to avoid it, and the catch in doing so.ConceptQuantum advantage: what it means, every major claim, and what has held up
Has quantum advantage been achieved? Yes for tasks built to test it, not yet for useful ones. Every major claim since Sycamore, and what held up.ConceptQuantum error correction: logical qubits, the threshold, and how far hardware has come
Quantum error correction turns noisy qubits into logical qubits. What below threshold means, which machines have crossed it, and what is still missing.ConceptQuantum neural networks (QNN) explained: QNN vs hybrid QNN (HQNN), and how to test whether the quantum layer helps
A quantum neural network is a trainable quantum circuit used as a model. How a QNN differs from a hybrid QNN (HQNN), and how to test whether the quantum layer helps.ConceptQuantum superposition: what a qubit holds before you measure it
What quantum superposition is: a qubit holds an amplitude on 0 and on 1, measurement returns one of them, and interference is what algorithms use. With eigenkit code.Methods
How to apply a quantum method to a problem. The eigenkit docs have the API.
Method