On one day in August 2026, the quantum computing field hit a three-front inflection point: error correction + domestic commercialization + new topological materials. This post analyzes the three milestones and the common engineering logic behind them.
Key points
1. Nord Quantique SPAM < 0.1%: The Canadian company's single-mode grid-state (GKP) qubit achieved a state-preparation-and-measurement error rate below 0.1% — a 100x improvement over conventional GKP systems and on par with superconducting transmons. 2. QuantumCTek (Guodun) first non-GAAP quarterly profit: 2026 H1 revenue of RMB 124 million (+183.96% YoY); Q2 non-GAAP net profit turned positive for the first time at RMB 14.6 million; quantum computing revenue (RMB 66.22 million, +18.34%) became the largest income source. 3. Non-Abelian anyons in bilayer graphene: Columbia University + Japan's NIMS, publishing in Nature, detected e/4 charge at filling factors ν = -5/2, -1/2, and 3/2 — the first confirmation of non-Abelian anyons in a 2D material. 4. Unified logic: All three results represent the shift from scientific discovery to engineering implementation — "physicists exit, engineers enter."
1. Nord Quantique: GKP qubit SPAM below 0.1%
| Metric | Conventional GKP | Nord Quantique single-mode grid state | Superconducting transmon | |---|---|---|---| | SPAM error rate | ~10% | < 0.1% | < 0.1% | | Relative improvement | 1x | 100x | Par with transmon | | Protocol | Real-time correction + complex classical control | repeat-until-success stabilization | Standard transmon control | | Magic state preparation | Resource-intensive | Integrated with error correction | External protocol required |
The core innovation is the repeat-until-success protocol: rather than correcting errors in real time, prepare a state, verify success, and either keep it or discard and retry. This avoids dependence on real-time classical control and — critically — embeds magic state preparation (needed for non-Clifford operations, previously considered the most resource-intensive challenge) into the error correction protocol itself.
> GKP (Gottesman-Kitaev-Preskill) qubits encode quantum information in the position-momentum phase space of a harmonic oscillator. Theoretically robust against noise, but historically hard to control in practice. Nord Quantique's result makes GKP workable in engineering terms.
Roadmap to 2030: single-mode GKP SPAM < 0.1% (done) → fault-tolerant logical qubit engineering (2027–2028) → utility-scale fault-tolerant quantum computing (2029–2030). If it holds, the path to utility-scale fault tolerance shifts from "5000+ physical qubits" to "~100 physical qubits + repeat-until-success."
2. QuantumCTek: first positive quarterly non-GAAP profit
| Financial metric | 2026 H1 | YoY | |---|---|---| | Revenue | RMB 124 million | +183.96% | | Net loss attributable to parent | RMB 23.18 million (narrowed) | Reduced loss | | Non-GAAP net loss | RMB 41.19 million (narrowed) | Reduced loss | | Q2 non-GAAP net profit | +RMB 14.6 million, first ever positive | Historic first | | Quantum computing revenue | RMB 66.22 million | +18.34% (largest segment) | | Quantum communication revenue | RMB 49.51 million | Stable | | Quantum measurement revenue | RMB 4.53 million | Slight decline |
Simultaneously disclosed commercialization progress:
1. Delivery and acceptance of Tianyan-287, a superconducting quantum computer of the same design as Zuchongzhi 3.0, to China Telecom 2. Steady progress on an overseas 25-qubit project 3. Continued delivery of new dilution refrigerators and thousand-qubit control systems 4. Co-founding a RMB 1.5 billion quantum industry fund with China Telecom (QuantumCTek contributing RMB 300 million)
Given the company's Q4-heavy seasonal revenue concentration, Q2 non-GAAP profitability makes full-year non-GAAP profit highly probable (to be confirmed in Q4 reporting). Combined with Huairou Zhongke Liangyi's modular cryogen-free dilution refrigerator winning third prize at HICOOL 2026, a Chinese quantum equipment supply chain (systems + infrastructure) is forming.
3. Columbia + NIMS: non-Abelian anyons in bilayer graphene
The Nature paper (team including Mario Di Luca, Emily Hajigeorgou, Ning Ma, Alexandra Waldherr, Mitali Banerjee of Columbia, and Kenji Watanabe, Takashi Taniguchi of NIMS; arXiv: https://arxiv.org/abs/2608.27444v1) used bilayer graphene in the fractional quantum Hall regime with engineered antidot defects to confine charge carriers:
| Filling factor ν | Detected local charge | Significance | |---|---|---| | -5/2 | e/4 | Non-Abelian anyon (core of fault-tolerant QC) | | -1/2 | e/4 | Same | | 3/2 | e/4 | Same | | 2/3 | e/3 | Abelian fractional excitation |
The e/4 charge is the signature of non-Abelian anyons, whose braiding operations change the quantum state in a topologically protected way — the material basis for fault-tolerant quantum computing resistant to local noise. Previously such signatures had only been weakly observed in GaAs heterostructures at ν = 5/2. Additional findings: as antidot-bound states couple more strongly to edge channels, the system crosses over between two transport regimes, with gate-voltage periodicity doubling or halving — hints of quasiparticle bunching, relevant for device scaling. The authors note that if bilayer graphene can host and control non-Abelian anyons, inherently decoherence-resistant qubits could enable ultra-low-power quantum processors relevant to grid balancing and molecular simulation.
The antidot engineering is the key engineering move: artificial structures replace natural defects, making topological qubits designable, manufacturable, and scalable.
4. The hardware landscape on this date
| Hardware route | Representative players | August update | |---|---|---| | Superconducting transmon | IBM/Google/Zuchongzhi | Q-CTRL QFT 1.8% fidelity result | | Ion trap | Quantinuum/IonQ | Helios 99.921% fidelity (day before) | | Neutral atoms | QuEra/Infleqtion | Shunkai 500 qubits | | Photonic | PsiQuantum/IX | Photonic SHYPS | | Topological | Microsoft/D-Wave | Columbia + NIMS bilayer graphene | | GKP CV | Nord Quantique | SPAM < 0.1% | | Dilution refrigerators | QuantumCTek/Zhongke Liangyi/IBM | QuantumCTek Q2 profit + HICOOL award |
5. Two engineering philosophies: Nord Quantique vs IBM
- Nord Quantique (light hardware, heavy protocols): few physical qubits, repeat-until-success compensation — if it works, the next 5 years look like "100 physical qubits + strong software correction."
- IBM (heavy hardware, modular): 1000+ physical qubits, modular cryogenics — if it works, "10,000 physical qubits + weaker software correction."
Observations
1. Nord Quantique's SPAM < 0.1% (100x over conventional GKP, transmon parity) is GKP's key step from theory to engineering. 2. QuantumCTek's first positive Q2 non-GAAP profit and 183.96% revenue growth mark a commercialization inflection for China's quantum industry. 3. The bilayer graphene e/4 detection confirms non-Abelian anyons in 2D materials; antidot engineering makes topological qubits designable. 4. Together: quantum computing's industrial year begins not from 1000 physical qubits but from three concrete engineering breakthroughs — sub-0.1% SPAM, domestic non-GAAP profitability, and non-Abelian anyons in 2D materials.
References
1. The Circuit Journal Daily — Nord Quantique sub-0.1% SPAM: https://thecircuitjournaldaily.com/article/quantum-error-correction-breakthrough-nord-quantique-achieves-sub-0-1-spam-errors 2. Energy Innovation Review — Bilayer graphene breakthrough: https://energyinnovationreview.com/2026/08/28/breakthrough-in-bilayer-graphene-quantum-computing-reshapes-energy-tech 3. CFI.cn — QuantumCTek quantum computing scale-up: https://cfi.cn/p20260828002008.html 4. arXiv — Columbia + NIMS bilayer graphene paper: https://arxiv.org/abs/2608.27444v1 5. Toutiao — HICOOL 2026 Huairou Zhongke Liangyi modular dilution refrigerator: https://www.toutiao.com/article/7678869329349214735