Computation unbound from classical limits.
Wave Quantum Systems builds the full stack — superconducting qubits, real-time error mitigation, and applied research programs — so chemistry, finance, and security teams can take quantum advantage from a benchmark to a balance sheet.
A coherent quantum platform — from cryostat to compiler.
One vendor for the parts that have to be coupled tightly. Open SDKs for the parts that shouldn't be.
4,096-qubit QPU
Heavy-hex superconducting lattice with 500 ms median T2 coherence and 99.84% two-qubit gate fidelity, measured continuously and published every six hours.
Real-time error mitigation
Dynamic circuit compilation, pulse-level scheduling, and zero-noise extrapolation built into the runtime — no post-processing scripts to glue together.
Python & Rust SDK
A typed intermediate representation that targets WC hardware, hosted simulators, and classical fallbacks. Same code, dev to prod, single line change.
The WC-890 processor — a chandelier of physics.
Suspended in a 10 mK dilution refrigerator, the WC-890 is a 4,096-qubit heavy-hex lattice fabricated at our Yokohama foundry and tuned in San Francisco. It supports up to 12 distance-7 logical qubits with break-even error correction — enough to run the chemistry programs that classical hardware cannot finish.
- Topology
- Heavy-hex
- Qubits
- 4,096 physical
- Logical
- 12 distance-7
- Coherence
- 500 ms T2
- Fidelity
- 99.84% 2Q
- Connectivity
- All-to-all (compiler)
Six sectors. Twelve live programs. Real baselines.
Every Wave engagement starts with a measured classical baseline and ends with a benchmarked quantum result — never a slide deck. Below, the sectors where we already have programs in production.
Industry briefs →- 01PharmaceuticalsFeMoco · CYP450 · GPCR docking
- 02Quantitative financePortfolio · derivative pricing · VaR
- 03Energy & materialsCatalysts · electrolytes · carbon capture
- 04Logistics & supplyFleet routing · scheduling · packing
- 05National securityPost-quantum migration · secure comms
- 06Climate scienceCloud microphysics · ocean eddies
Active-space VQE on nitrogenase at chemical accuracy in 36 hours.
A joint Wave–Pfizer team mapped the FeMoco active space onto 154 qubits using a doubly-projected unitary coupled-cluster ansatz. With dynamic noise calibration on WC-890, ground-state energies converged to ±1 mHa in 36 wall-clock hours, against a classical DMRG baseline that exceeded 11 days at lower precision.
Measured results, signed off by the customer.
Each brief ships with a classical baseline, a quantum result, and a methodology PDF. No marketing math.
Field notes from the lab and the runtime.
Long-form writing from the Wave team — postmortems, applied research, the occasional opinion piece.
What we learned the year we put Wave OS into production at Goldman.
Year-one numbers from running quantum amplitude estimation in shadow against a real derivatives book — the wins, the surprises, and the thing we still cannot explain.
Why we rebuilt the Wave compiler around a typed intermediate representation.
After three years of an untyped IR we hit a wall. Here is what broke, what we replaced it with, and the one ergonomic loss we still have not solved.
Heavy-hex vs. hex-lattice: what we got wrong in 2023 and how WC-890 fixed it.
We bet on heavy-hex for two generations of processors. The data eventually said otherwise. A short note on why connectivity topology matters more than we admitted.
What people are writing about Wave.
Four reasons serious teams pick us over a research demo.
No advantage theatre. No slideware. Every claim ships with a number, a baseline, and a date.
Full-stack control
We own the qubit, the cryostat, the compiler, and the runtime. When something drifts at 10 mK, the fix ships from one team — not three vendors.
Production telemetry
Calibration data is published every six hours, not when it looks good. Customers see the same dashboards our SREs see.
Costed roadmaps
Every engagement ends with a quantum cost-per-result number versus your current classical baseline. No advantage, no invoice.
Long-horizon programs
Multi-year embeds with Pfizer, Goldman, and Maersk. We optimise for the second deployment, not the first press release.
A twelve-week path from question to costed quantum roadmap.
- Weeks 1–201
Problem framing
Joint workshops with your domain team. We pick one bounded problem with an existing classical baseline and an executive sponsor.
- Weeks 3–602
Classical baseline
Our applied research team reproduces, profiles, and optimises your classical solver. Most quantum work starts by improving your classical floor.
- Weeks 7–1003
Quantum implementation
Algorithm selection, ansatz design, and execution on WC-890 — with full pulse-level diagnostics shipped to your team each Friday.
- Weeks 11–1204
Costed roadmap
A written deliverable: measured speedup or honest negative result, projected scaling to fault-tolerant hardware, and the price per future run.
What partners say about working with Wave.
Wave is the only quantum vendor we have worked with who showed up on day one with the classical baseline already beaten. That set the tone for everything after.
Eighteen months of shadow-mode pricing on a real derivatives book. Every morning, a number we can diff. No PR, no theatre. Exactly what we asked for.
The Wave team treats fleet routing like a serious engineering problem, not a quantum demo. The result was a 9% reduction in empty-leg miles we can audit.
Start a program with the team building the useful quantum era.
Twelve-week scoping engagements that produce a measured baseline and a costed quantum roadmap — or your money back.