The Systems That Made Headlines in 2025
Three quantum systems dominated the conversation in 2025. Each operates on a different physical architecture, targets different applications, and signals a different dimension of China’s capabilities. Understanding them individually is more useful than treating them as a single undifferentiated advance.
Zuchongzhi 3.0: 105 qubits and what that actually means
Developed by a team at the University of Science and Technology of China (USTC), Zuchongzhi 3.0 is a superconducting quantum processor operating at 105 qubits. The research team reports that it completed a specific random circuit sampling task in minutes — a calculation they claim would take the world’s fastest classical supercomputers an impractical length of time to replicate.
It is worth being precise about what that claim does and does not establish. Random circuit sampling is a benchmark task designed to be hard for classical computers, not a commercially useful computation. The result demonstrates quantum advantage on a narrow, engineered problem — not general-purpose quantum computing capability. Peer-reviewed independent replication of the full result has not yet been widely confirmed. That caveat aside, 105 superconducting qubits with the reported error rates represents a credible technical milestone, and places Zuchongzhi 3.0 alongside — rather than clearly ahead of — comparable systems from IBM and Google.
Hanyuan-1: the first commercial neutral-atom deployment
The more strategically significant announcement may be Hanyuan-1, which the developers describe as the first commercially deployed neutral-atom quantum computing system in China. Neutral-atom computing uses individual atoms — typically rubidium or caesium — trapped by laser arrays as qubits, rather than the superconducting circuits used in Zuchongzhi.
The architecture matters because neutral-atom systems offer longer coherence times and greater qubit connectivity than many superconducting designs, potentially making them more suitable for real-world optimisation and simulation tasks. More importantly, Hanyuan-1 has reportedly been deployed in a data centre environment rather than a research laboratory. According to coverage from The Conversation‘s academic commentary on 2025 quantum advances, this transition from lab to operational setting is one of the most meaningful developments of the year — regardless of the specific qubit count involved.
The photonic quantum AI chip entering real data centres
The third milestone is a photonic quantum chip designed to accelerate AI workloads within classical data centre infrastructure. Unlike gate-model quantum computers, photonic chips process information using photons — particles of light — which can operate at room temperature and integrate more readily with existing optical networking hardware.
The practical implication is a hybrid quantum-classical architecture in which quantum coprocessors handle specific matrix operations or optimisation subroutines, while conventional processors handle everything else. This is arguably the nearest-term route to commercial quantum advantage, and the fact that Chinese developers are deploying photonic quantum chips in production data centre environments signals a shift from research demonstration to industrial application.
How These Systems Compare to the Global Competition
China’s 2025 quantum progress is real, but the comparative picture is more nuanced than either triumphalist or dismissive readings suggest. The ITIF’s 2024 analysis of innovation in China’s quantum sector, and the McKinsey Quantum Technology Monitor, both provide useful structured frameworks for the comparison.
Where China leads: quantum communications dominance
China’s clearest global lead is in quantum communications — specifically quantum key distribution (QKD) over fibre and satellite links. The Micius satellite network remains the most advanced operational quantum communications infrastructure in the world. This is a distinct discipline from quantum computing, and conflating the two distorts the overall picture. China leads in quantum communications; it does not yet lead in quantum computing.
China has also accumulated significant strengths in quantum sensing — the use of quantum systems to measure physical phenomena with extreme precision — and holds a large volume of quantum-related patents, though the ITIF analysis notes that patent count alone is a weak proxy for genuine innovation depth.
Where China still lags: fault-tolerant quantum computing vs. the US
The United States retains a meaningful lead in fault-tolerant quantum computing — the regime in which error correction enables reliable computation at scale. Google’s 2024 Willow chip demonstrated error correction improvements that represent a genuine step towards fault tolerance. Microsoft’s topological qubit programme and IBM’s roadmap to 100,000 physical qubits are both oriented towards this goal.
China’s current systems, including Zuchongzhi 3.0, operate in the NISQ era — noisy intermediate-scale quantum — where error rates limit the depth and reliability of computations. The gap between NISQ performance and fault-tolerant quantum computing is not merely technical; it is the difference between a system that can win engineered benchmarks and one that can solve industrially relevant problems reliably. China has not yet publicly demonstrated fault-tolerant operation at meaningful scale.
The qubit quality vs. qubit count debate
Qubit count is the metric most commonly cited in press coverage, but experts and the McKinsey Quantum Technology Monitor consistently emphasise qubit fidelity — the accuracy with which quantum gates are executed — as the more operationally meaningful measure. A system with 1,000 low-fidelity qubits is, in most practical scenarios, less useful than one with 100 high-fidelity qubits with robust error correction.
China’s published figures for Zuchongzhi 3.0 include competitive gate fidelity numbers, but independent benchmarking of these claims remains limited. Until results undergo rigorous third-party verification, treating the headline qubit count as a direct measure of capability overstates certainty.
China’s Industrial Strategy: From Research Lab to Factory Floor
Perhaps the most consequential development in 2025 is not any single system, but the structural shift in how China is approaching quantum technology — from academic achievement to industrial deployment.
What the 15th Five-Year Plan actually mandates for quantum
China’s 15th Five-Year Plan, covering 2026–2030, explicitly identifies quantum technology as a strategic priority industry. This follows the pattern established in AI and semiconductors: identify a sector of long-term national importance, direct state capital towards it, and build domestic supply chains to reduce external dependencies.
The plan’s quantum provisions go beyond research funding. According to CSIS analysis of China’s quantum ambitions published in early 2026, the policy framework mandates progress in quantum commercialisation — not just publication output or patent filings. Government procurement, national laboratory deployment, and state enterprise adoption are all mechanisms being used to create domestic demand for quantum systems, bypassing the slower market-pull dynamic that governs commercial technology development in most Western countries.
The shift from R&D funding to commercial deployment
The deployment of Hanyuan-1 in a commercial data centre environment is a direct product of this industrial policy logic. Rather than waiting for market demand to pull quantum systems out of laboratories, Chinese policy is pushing systems into operational environments — accepting imperfect performance in exchange for accumulated deployment experience and supply chain development.
This approach has precedents in Chinese solar, electric vehicles, and high-speed rail. It generates deployment data and engineering learning that purely research-oriented programmes do not. It also creates domestic customers and ecosystems that can sustain the technology through its commercially awkward adolescence.
State investment figures and the widening gap with Western rivals
Precise figures for China’s quantum investment are difficult to verify, given the opacity of state budget allocations and the blending of military and civilian spending. The McKinsey Quantum Technology Monitor estimates that China has committed in excess of $15 billion USD to quantum technology since 2016 — a figure that, if accurate, exceeds the combined public investment of the United States and European Union over the same period. The US National Quantum Initiative has authorised multi-billion-dollar funding, but annual appropriations have consistently fallen short of authorised levels.
The investment asymmetry matters not because spending guarantees results, but because it shapes the talent pipeline, the infrastructure base, and the domestic supplier ecosystem that sustains long-term capability. The ITIF report notes that China produces more quantum-focused PhD graduates annually than the US, though the report also observes that citation impact — a proxy for research quality — still favours American institutions in several quantum sub-disciplines.
What the Breakthroughs Actually Mean for Global Tech
The systems described above matter in their own right. But the more urgent question for technologists, policymakers, and organisations is what these developments imply for infrastructure, security, and competitive position.
Cryptography and the ‘store now, decrypt later’ threat
The most immediately relevant implication for most organisations is cryptographic. Current public-key encryption standards — RSA, elliptic curve cryptography — are mathematically vulnerable to a sufficiently powerful quantum computer running Shor’s algorithm. That machine does not exist yet. Q-Day, the point at which a quantum computer could break widely deployed encryption at scale, is not an imminent threat in 2025.
However, the store now, decrypt later (SNDL) attack is already operative. State-level actors — and China is the most frequently cited concern in Western intelligence assessments — are assessed to be harvesting encrypted data today with the intention of decrypting it once sufficiently powerful quantum systems become available. Sensitive communications, intellectual property, and long-lived government secrets transmitted now are potentially exposed to future decryption. The timeline for that risk to mature is measured in years to decades, but data captured today will still be decryptable when Q-Day eventually arrives.
Post-quantum cryptography (PQC) is the established response. NIST finalised its first set of post-quantum cryptographic standards in 2024, providing algorithm specifications that are believed to be resistant to quantum attack. Migration to these standards is technically achievable now, though it requires systematic cryptographic inventory, system upgrades, and vendor coordination across complex supply chains.
Implications for AI acceleration and data centre architecture
The photonic quantum chip development signals a near-term architectural shift in data centres that is separate from — and in some ways more immediately practical than — gate-model quantum computing. Quantum AI coprocessors that accelerate specific matrix operations or sampling tasks within otherwise classical pipelines could deliver meaningful performance improvements on AI training and inference workloads without requiring fault-tolerant quantum computing.
If Chinese hyperscalers and state data centres begin operating with hybrid quantum-classical infrastructure at meaningful scale, they may achieve performance and energy efficiency advantages on certain AI workloads that are difficult to replicate with purely classical architectures. This is speculative at present — the photonic chip deployments reported in 2025 are early-stage — but it represents a plausible commercial differentiator within a three-to-five-year horizon.
Supply chain and semiconductor dependencies
China’s quantum programme faces its own supply chain constraints. Superconducting quantum processors require dilution refrigerators that cool qubits to near absolute zero — hardware in which Western suppliers, particularly Oxford Instruments and Bluefors, hold dominant positions. US export controls, tightened progressively since 2022, have complicated Chinese access to certain cryogenic and semiconductor fabrication equipment relevant to quantum manufacturing.
The photonic and neutral-atom architectures partly reflect a strategic response to these constraints. Both are less dependent on the extreme cryogenic infrastructure that superconducting designs require, and both offer cleaner integration with existing photonic and atomic physics manufacturing bases that China has developed domestically. Architectural choice, in this context, is partly an industrial policy decision.
What Organisations and Technologists Should Do Now
The honest answer is that most organisations are not yet directly affected by China’s 2025 quantum milestones. The systems described above are not, today, capable of breaking production encryption or outperforming classical computers on commercially relevant tasks at scale. But ‘not yet’ is the operative phrase, and the decisions made now will determine how exposed organisations are when the capability landscape shifts.
Is post-quantum cryptography migration urgent yet?
For most private sector organisations: migration is warranted, but not a crisis-level emergency for 2025. The priority depends on the sensitivity and longevity of the data involved. Organisations handling data that must remain confidential for ten or more years — healthcare records, legal documents, intellectual property, government communications — should treat PQC migration as an active programme, not a future planning item.
NIST’s finalised post-quantum standards — including CRYSTALS-Kyber (now ML-KEM) for key encapsulation and CRYSTALS-Dilithium (ML-DSA) for digital signatures — provide concrete migration targets. The UK’s National Cyber Security Centre (NCSC) has published guidance aligned with the NIST framework. Organisations should begin with a cryptographic inventory: identifying where RSA and elliptic curve keys are used across their infrastructure, and prioritising the highest-sensitivity, longest-lived data for early migration.
How to track quantum readiness as the landscape shifts
Quantum readiness is not a one-time assessment. The capability frontier is moving, and the threat model will evolve as fault-tolerant systems get closer. Useful practices include monitoring NIST’s ongoing PQC standardisation process, tracking NCSC and CISA advisories on quantum risk, and requiring quantum risk disclosure from critical vendors handling sensitive data.
For technology teams, the McKinsey Quantum Technology Monitor and the CSIS quantum tracker provide credible, regularly updated assessments of where the global capability frontier sits. Engaging with these primary sources — rather than relying on press coverage, which tends to oscillate between hype and dismissal — is the most reliable way to maintain an accurate picture of quantum risk assessment over time.
The broader competitive question — whether China’s quantum industrialisation strategy will translate into durable technological advantages — will take years to resolve. What is clear in 2025 is that the transition from research demonstration to commercial deployment is underway, and the organisations and governments that treat that transition as a future concern rather than a present one will find themselves managing risk reactively rather than proactively.
Frequently Asked Questions
Which country is currently leading in quantum computing?
No single country leads across all quantum disciplines. The United States holds the advantage in fault-tolerant quantum computing and overall research quality, with IBM and Google setting the pace on error correction. China leads in quantum communications and has made credible progress in quantum computing hardware. The quantum race in 2025 is competitive and multidimensional, not a simple ranking.
What is the current state of quantum computing in 2025?
Quantum computing in 2025 remains in the NISQ era for most systems — capable of impressive benchmark demonstrations but not yet delivering reliable, general-purpose commercial advantage. The frontier is advancing on multiple architectures (superconducting, neutral-atom, photonic). Fault-tolerant quantum computing, which would enable transformative real-world applications, remains a medium-term goal rather than a current reality.
Does China’s quantum progress threaten Western encryption?
Not immediately. Current quantum systems, including Zuchongzhi 3.0, are not capable of breaking production encryption. The real near-term risk is the store now, decrypt later threat: adversaries harvesting encrypted data now for future decryption. Organisations with long-lived sensitive data should begin migrating to NIST-standardised post-quantum cryptography, but a practical encryption vulnerability at scale is not an imminent 2025 threat.