GCRI · 量子技术
量子计算
Quantum computers use qubits that can be in superpositions and become entangled, which in principle lets some problems be solved far faster than on classical computers. Current machines are noisy and small, and useful fault-tolerant computing requires error correction that is only beginning to be demonstrated. This page sets out how quantum computers work and fail, how progress is measured honestly, and what evidence serves the institutions deciding how to engage.
Quantum computing: how the risk develops.
Hardware, algorithms and actual maturity. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
Qubits are fragile
Qubits lose their quantum state through interaction with their environment. Gate errors and decoherence limit how long a computation can run before results become noise.
Error correction needs overhead
Fault-tolerant computing encodes each logical qubit in many physical qubits and corrects errors continuously. It only helps once physical error rates fall below a threshold.
Advantage is problem-specific
Quantum algorithms speed up particular problems, such as simulating quantum 系统 or factoring. For many tasks classical computers remain as good or better.
Claims need classical comparison
Demonstrations of quantum advantage are often followed by improved classical methods. Honest comparison requires the best classical approach, not a convenient one.
Quantum testbed.
团队跟踪该主题时会使用的界面：其指标、状态与决策。名称为示例，数值仅作示意。
Quantum testbed
Quantum computing · Testbed lab · Device QD-28
量子比特
| Q0 | 121 | 49 | 0.03 | Drifting |
| Q1 | 118 | 24 | 0.03 | Drifting |
| Q2 | 94 | 41 | 0.03 | Drifting |
| Q3 | 73 | 93 | 0.05 | Drifting |
| Q4 | 50 | 48 | 0.02 | Drifting |
Device
- Topology
- heavy-hex, illustrative
- 2Q fidelity
- 0.994
- 访问
- by committee
- 标签
- 观测
Quantum computing matters to the domains through simulation and through cryptography.
- 能源耦合Simulation of battery and catalyst chemistry is a proposed early use.What to watchDemonstrated against classical baselines
- 卫生耦合Future cryptanalysis threatens long-lived health data.What to watchMigration of health systems
- 食品耦合Fertiliser catalysis is often cited as a target problem.What to watchClaims separated from results
Status reviews are designed for the 指数级技术系列, separating demonstrations from roadmaps. No date for a 具有密码学意义的量子计算机 出现的日期表述为已知。
- Peer-reviewed results with classical comparisons.
- Corrections to over-stated claims.
How quantum computing is modelled, measured and tested.
模型、其假设、数据和不确定性，以科学家或工程师会核查的形式呈现。结构遵循已发表的方法。所取数值仅为示意，不是研究发现。
系统模型
The couplings between the systems in quantum computing: signs of influence, a reinforcing and a balancing loop, and cross-links where stress jumps.
衡量什么，以及衡量得如何。
每个指标都注明其单位、频率、背后的来源类型以及数值将附带的证据标签。此处不显示数值：数值由计划依据其自身权限提供。
| 指标 | 单位 | 节奏 | 来源类型 | 标签 |
|---|---|---|---|---|
| Physical qubit count | qubits | per announcement | Published device results | 已报告 |
| Two-qubit gate fidelity | percent | per device | Benchmark results | 已报告 |
| Logical qubit error rate | errors per logical operation | per experiment | Peer-reviewed experiments | 观测 |
| Coherence time | microseconds | per device | Device characterisation | 已报告 |
| Quantum volume or similar holistic benchmark | benchmark score | per device | Published benchmarks | 已报告 |
| Demonstrations surviving classical challenge | 计数 | 每次审查 | Literature review | 估算 |
证据来自何处。
具名公共数据集与机构记录，各自注明许可类别。依协议获得的来源仅按其条款使用，绝不再次发布。
| 来源 | 它提供什么 | 许可类别 |
|---|---|---|
| arXiv quant-ph and peer-reviewed journals | Primary experimental and theoretical results | Open access preprints; journal licence per article |
| Metriq benchmark platform (Unitary Foundation) | Community-submitted quantum benchmark results | Open, licence to check |
| Vendor technical documentation and roadmaps | Device specifications and stated targets | Public documents, terms per vendor |
| National academies and agency assessments of quantum computing | Independent assessments of progress and prospects | Public, terms per publisher |
| QED-C reports | Industry consortium analyses of use cases and benchmarks | Public reports, terms to check |
由谁决定，以及证据服务于什么。
Quantum Technologies prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.
| 谁来决定 | 问题 | 本页面服务的证据 |
|---|---|---|
| Research funders and science ministries | How much to invest in quantum computing research and access | Progress summaries; independent assessments |
| Organisations exploring quantum use | Whether a problem is a plausible quantum use case | Use case assessments with classical baselines |
| Security agencies | How progress affects cryptographic migration timelines | Error correction and scale evidence |
该主题产生的产出
- Progress summaries that separate peer-reviewed results from roadmaps
- Benchmark interpretation guides
- Use case assessments with classical baselines
- Evidence notes for research funders
专家提出的问题。
The central question is what has been demonstrated, not what has been promised.
- How many qubits does a device have, and how good are they: what are their error rates and coherence times?
- Can errors be corrected faster than they accumulate, and at what cost in physical qubits per logical qubit?
- Which algorithms offer a demonstrated advantage over the best classical methods, and for which problems?
- Are claims of quantum advantage robust to improved classical algorithms?
- Which hardware approaches are likely to scale, and what limits them?
机制与驱动因素。
Quantum computers gain their power from interference: algorithms are designed so that wrong answers cancel and right answers reinforce. That power is fragile. Qubits lose their quantum state through interaction with the environment, and every operation introduces some error. Quantum error correction encodes one reliable logical qubit in many physical qubits, but only works if physical error rates are below a threshold, and it multiplies the hardware required.
Known algorithms with large demonstrated speed-ups are few. Factoring and discrete logarithms, which underpin public-key cryptography, are the best-known examples; simulation of quantum 系统 in chemistry and physics is another. For many optimisation and machine-learning problems, a practical advantage has not been established, and improved classical algorithms have repeatedly narrowed claimed gaps.
如何研究，以及方法止于何处。
| 方法或数据 | 显示内容 | 局限 |
|---|---|---|
| Benchmarking of gate fidelity and coherence | Quality of physical qubits | Benchmarks can be chosen to flatter a device |
| Error-correction experiments | Whether logical qubits outperform physical ones | Early demonstrations are small and short-lived |
| Algorithm complexity analysis | Theoretical speed-up and resource estimates | Resource estimates depend on hardware assumptions |
| Classical simulation and verification | Whether results could be reproduced classically | Verification becomes harder as devices grow |
| Peer-reviewed literature versus announcements | What has been independently assessed | Announcements often precede, or replace, peer review |
Where this page’s references lead.
阅读本页时所对照的机构和文件。状态和版本会变化；每个链接都指向发布机构。
问题与解答。
什么是量子比特？
What is quantum advantage?
What is fault-tolerant quantum computing?
Why does qubit count alone mislead?
Does the platform predict when large quantum computers will arrive?
从本页贡献证据。
报告 is the Consortium’s quarterly publication. What you know about this subject can become evidence, a case or a correction in the next edition, credited to you and published under CC BY 4.0 where it is your own work.
- 电子邮件网关 GW-GCRI从本页发送证据、更正、案例、方法、审查、翻译或概况投稿。它会带上本页标签送达，由负责的编辑接收。
- Zenodo开放存缴存缴并提交到社区用您的 ORCID iD 在 Zenodo 上发布您自己的记录，然后提交到那里的 Reports 社区。由策展人审核；DOI 仍归您所有。
- 报告GCRI查看所有贡献方式贡献类型、各自的评审方式、您获得的署名、许可协议，以及绝不接受的内容。
切勿发送个人数据、敏感基础设施细节或您无权分享的任何内容。贡献在使用前会经过审查；发送贡献不赋予署名权或编辑权，赞助也绝不能换取研究结论。
法律文件。
本网站上的地图、数据和图表仅作为背景和一般信息展示。它们不是建议、预测或预警；信息的局限见免责声明和数据说明。