初创公司尽调
尽调报告 Quantum Computing / Hardware pre-IPO private 2026-05-26

Quantinuum

IPO 前量子龙头,有 Honeywell 背书、Helios 动能和 $10B 私募估值

Quantinuum 是最强的 pre-IPO 量子平台之一;但 2025 年收入只有 $30.9M、净亏损 $192.6M,估值却到 $10B,Honeywell 控制条款也未解决,标的更适合观察而不是买入。

封面要素

最后一轮私募估值 01
10000 USD M [CO023]
2025 年 9 月宣布融资 02
600 USD M [CO023]
S-1 披露的 B 轮 03
838.8 USD M [CO025]
2025 年收入 04
30.9 USD M [CO027]
2025 年净亏损 05
192.6 USD M [CO027]
现金 06
677 USD M [CO029]
员工数 07
700 employees [CO033]
成立时间 08
2021 year [CO001]

公司概况

Quantinuum 于 2021 年 11 月 30 日由 Honeywell Quantum Solutions 与 Cambridge Quantum 合并而来,拼出一家全栈量子计算公司:离子阱硬件、中间件、化学与开发者软件、网络安全产品都在同一栈里。公司的主要行政办公室位于 Colorado 州 Broomfield,Cambridge 仍是其欧洲枢纽。Quantinuum 现在通过直销和云渠道销售 H-Series 与 Helios 系统访问,并提供 TKET、InQuanto、Nexus、Quantum Origin 等产品。到 2026 年 IPO 申报时,它已经是一家商业化真实存在但仍处早期的公司:2025 年收入 $30.9M,现金余额很大,蓝筹伙伴强,但亏损也大,并继续依赖 Honeywell 治理和后续技术里程碑。

官网
www.quantinuum.com
成立时间
2021-11-30
创始人
Ilyas Khan
创立地点
Cambridge, UK / Broomfield, Colorado, USA
总部
Cambridge, UK / Broomfield, Colorado, USA
产品
Quantinuum 通过 H-Series 系统和 Helios 平台销售离子阱量子计算访问,并提供 TKET、InQuanto、Nexus、Quantum Origin、Guppy、lambeq 等软件和工作流产品。
客户
主要面向金融、制药、汽车、化工、网络安全和高级研究场景中的大型企业、国家实验室,以及公共部门或战略型用户。
商业模式
变现方式包括系统订阅和云访问、直接 HaaS 式算力合同、软件与网络安全许可证,以及围绕量子工作流的咨询或联合开发项目。
阶段
pre-IPO private
融资情况
Quantinuum 于 2024 年 1 月完成 $300M 融资,投前估值 $5B;2025 年 9 月又宣布一轮约 $600M 融资,投前估值 $10B。其 2026 年 5 月 S-1 还披露,2025 年 11 月发行了 $838.8M 的 B 轮优先股,意味着合并后外部资本至少约 $1.14B。
[CO001, CO005, CO007, CO020, CO023, CO025, CO027, CO029]

执行摘要

主要优势

  • 从离子阱硬件、开发者工具、化学软件到网络安全,Quantinuum 走全栈路线,商业化触点比多数量子同行更多。
  • Honeywell 背书叠加 JPMorgan Chase、Amgen、Mitsui、Quanta Computer、NVentures、QED 等投资者,带来少见的战略可信度和融资厚度。
  • Helios 时代的技术进展、逻辑量子比特里程碑,以及云端 / 本地部署证据,让平台比许多纯概念量子故事更可信。
  • S-1 证明公司有真实运营业务、可观现金储备和活跃 IPO 路径,而不是只有叙事的私募轮。

主要风险

  • Honeywell 仍是控股股东,IPO 前少数股东保护、减持时点和长期治理都没有定论。
  • 当前估值仍按里程碑定价,因为公开基本面只显示收入有限、亏损很大。
  • 客户集中度、未披露客户数和有限的试点转生产证据,让当前商业牵引力偏脆弱。
  • Quantinuum 仍需把 Helios 势头转成更大规模的容错系统,同时应对出口管制、供应商依赖和更强竞争对手。

未决问题

  • 最终 IPO 价格区间、股数、股权结构机制,以及 Honeywell 减持或锁定条款,到本次报告日期仍未公开。
  • 公开资料仍未披露硬件、云和软件各线的收入结构、实际定价、毛利率或续约指标。
  • 客户集中度、确切付费客户数,以及旗舰试点转为耐久经常性收入的情况仍说明不足。
  • 公开证据能清楚确认 2025 年 9 月最初 $600M 融资,但不能确认后来扩至约 $800M 总额的每个细节。
  • 供应商集中度、出口许可历史,以及 Honeywell 相关商业支持的准确经济性仍需尽调。

目录

Chapter 01

01公司概况

1.1 身份、合并起源与产品栈

Quantinuum 把自己定位为全栈量子计算公司,而不是单卖一颗处理器:离子阱硬件、中间件、开发者工具和应用软件都被装进同一产品体系。公司于 2021 年 11 月 30 日由 Honeywell Quantum Solutions 与 Cambridge Quantum 完成合并后成立。这个起点至今仍在塑造业务:Honeywell 带来离子阱硬件谱系,Cambridge Quantum 带来软件、中间件和应用栈。公司的主要行政办公室位于 Colorado 州 Broomfield,公开材料也把英国 Cambridge 作为延续中的欧洲总部。到 2026 年,Quantinuum 还公开列出其在美国、英国、德国、日本和新加坡的设施或运营。 产品栈并不只是 H-Series 算力访问。Quantinuum Systems 把离子阱 QCCD 硬件、全连接和中途测量打包在一起;客户可以直接向 Quantinuum 获取硬件,也可以通过 Microsoft Azure,或通过 RIKEN 这类本地部署使用。围绕硬件,Quantinuum 销售 Quantum Origin,用于基于软件的量子随机性和后量子安全;销售 InQuanto,用于量子化学和材料模拟工作流;还销售 TKET、Guppy、lambeq 等开发者工具。这个一体化定位就是本报告后续的核心商业模型:在通用容错完全到来之前,Quantinuum 试图先把纵向一体化的硬件加软件栈变现。[CO001, CO004, CO005, CO006, CO007, CO008]

FO002: 公司快照逻辑

Honeywell 控制权、离子阱系统、软件产品和旗舰合作伙伴如何拼出 Quantinuum 的商业化模型。

[CO003, CO007, CO008, CO009, CO018, CO023]

1.2 领导层、治理与 Honeywell 控制

领导层已经从创始人主导的整合,转向规模化执行。Raj Hazra 在 Micron 和 Intel 长期负责运营后,于 2023 年 2 月接任 CEO;这一步与 Quantinuum 从合并后整合转向商业化同步。Cambridge Quantum 创始人 Ilyas Khan 并未离开:他继续担任董事,转任副董事长,并在 2026 年作为高管团队中的特殊项目负责人出现。2026 年 3 月,Quantinuum 从 SoundHound AI 招来 Nitesh Sharan 担任 CFO;这次任命看起来既是补内部财务纪律,也是为资本市场做准备。 Honeywell 仍实质性影响治理。公开材料列出 Honeywell CEO Vimal Kapur 担任董事长,Anne T. Madden、Niels Nielsen、Greg Lewis、George Sherman 和 Manish Bhatia 进入董事会。Honeywell 不只是投资人;2021 年成立公告还把 Honeywell 描述为供应商和客户,Reuters 在 2026 年 IPO 流程中仍称 Honeywell 为多数股东。这个组合给投资人带来一体两面:一边是工业可信度、供应链支持和天然首批客户;另一边是控制权集中,而且关键人物风险仍真实存在——Hazra 是商业化和 IPO 执行的中心,Khan 仍与产品身份和公司软件传承紧密绑定。IPO 生效前,公开材料仍未充分说明委员会构成、少数股东保护机制,或发行后控制权的精确分配。[CO002, CO003, CO014, CO015, CO016, CO017]

领导层与创始人表
人物当前角色背景 / 前任关联创始人-市场匹配或覆盖范围关键人物依赖
Rajeeb Hazra总裁兼 CEO来自 Micron 和 Intel;2023 年 2 月出任 CEO商业化、企业销售、公开市场叙事
Ilyas Khan副董事长;特别项目负责人2014 年创立 Cambridge Quantum;Quantinuum 创始 CEO软件脉络、外部可信度、战略叙事
Nitesh Sharan首席财务官2026 年从 SoundHound AI 加入,此前曾任职 Nike 和 HP资本市场、资金管理、规模化纪律
Vimal Kapur董事长Honeywell CEO,并担任 Quantinuum 董事会主席母公司监督和工业支持
Anne T. Madden副主席About 页面公开列示的董事会副主席董事会连续性和治理支持
Jenni Strabley副总裁兼总经理,Compute Platforms GroupInQuanto、Quantum Systems 和 Nexus 的 P&L 负责人计算栈产品交付和运营归属

覆盖截至 2026-05-26 公开具名、且与所有权、商业化和创始人连续性关系最密切的高管和董事会角色。

[CO014, CO015, CO016, CO017, CO018, CO019]

1.3 资本结构、规模指标与商业成熟度

用户强调的两个价格点上,Quantinuum 已经具备可信的独角兽身份。2024 年 1 月融资规模 $300M,投前估值 $5B,由 JPMorgan Chase 领投,Mitsui、Amgen 和 Honeywell 参与;按 Quantinuum 自身说法,这使成立以来累计融资达到约 $625M。2025 年 9 月 4 日,Honeywell 又宣布一笔标志性融资:约 $600M、投前估值 $10B,Quanta Computer、NVentures 和 QED 加入,原有支持方包括 JPMorganChase、Mitsui、Amgen 和 Honeywell。随后 2026 年 5 月 S-1 披露,Quantinuum 在 2025 年 11 月发行了 $838.8M 的 B 轮优先股,意味着 2025 年末融资实际扩张幅度明显超过 $600M 的公告口径。 IPO 披露也首次给出较清晰的经济图景。Quantinuum 2025 年收入只有 $30.9M,净亏损 $192.6M;2024 年收入 $23.0M,净亏损 $144.1M。截至 2026 年 3 月 31 日的季度,收入降至 $5.2M,亏损扩大到 $136.6M,但 2025 年融资后现金仍有 $677.0M。员工数从成立时接近 400 人,扩至 2024 年接近 500 人、2025 年末超过 630 人、2026 年 3 月约 700 人。因此,人员、设施、系统和融资规模都是真实的;商业成熟度还不是同一量级。准确付费客户数仍未披露,S-1 显示收入明显集中在少数账户,这一点在接受 IPO 式估值倍数前必须重点尽调。[CO020, CO021, CO022, CO023, CO024, CO025]

KPI 快照表
指标数值 / 状态日期置信度缺口 / 备注
成立合并完成2021-11-30Honeywell Quantum Solutions 与 Cambridge Quantum 合并
主要总部Broomfield, Colorado2026-05-08S-1 披露地址为 303 S Technology Court
Honeywell 持股成立时约 54%;截至 IPO 申报仍为多数股东2021-11-30 / 2026-05-08上市后精确稀释尚未披露
S-1 披露的 2025 年融资$838.8M B 轮优先股2025-11-01超过 2025 年 9 月最初公告的 $600M
参考估值投前估值 $10B2025-09-04最近一次已公告私募轮估值
2025 年收入$30.9M2025-12-31见于 S-1 / Reuters
2025 年净亏损$192.6M2025-12-31亏损显著高于当前收入基数
现金$677.0M2026-03-312025 年末融资后现金余额较大
员工数约 700 名员工2026-03-18公开口径从 2021 年约 400 人升至 2026 年约 700 人
商用系统4 套上线系统;第 5 套计划在新加坡部署2026-05-08三套位于 Colorado,一套位于 RIKEN
精确付费客户数未公开披露2026-05-26存在具名部署,但精确客户数仍未公开
债务 / 信贷额度已披露历史可转债;当前独立融资额度未说明2026-05-08需要超出留存公开来源的财务尽调

混合列示历史成立事实与最新披露的运营和融资指标;未披露数值明确标注,而非估算。

[CO001, CO002, CO003, CO005, CO023, CO025]
利益相关方或投资者地图
利益相关方角色控制或经济重要性当前证据尽调请求
Honeywell多数股东 / 董事长 / 供应商-客户控制锚点;创建业务且仍为多数股东成立时约 54%;2026 年 IPO 报道中仍为多数股东厘清 IPO 后投票权 / 经济权益拆分和关联方协议
JPMorganChase投资者、客户、研究伙伴锚定 2024 轮,并验证金融用例2024 年具名投资者,2025 年继续为股东;认证随机性合作伙伴确定商业合同价值与研究赞助的区别
Mitsui & Co.投资者和分销伙伴通向日本和亚太市场的战略入口2024 年 $50M 投资和分销协议量化渠道转化和收入贡献
Amgen投资者和 Helios 早期用户生命科学验证加战略投资者信号2024 轮和 Helios 发布批次中具名评估合作是 R&D、试点还是生产
NVentures / NVIDIA投资者和生态伙伴补充 AI 生态信号和研究中心访问2025 年新投资者;NVAQC 创始合作澄清战略品牌之外的商业条款
Quanta Computer投资者支持制造 / 亚洲硬件生态信号2025 轮具名新投资者了解 Quanta 是否也支持硬件供应链
QED Investors投资者围绕 JPMorgan 用例补充金融科技资本网络2025 轮具名新投资者核查 QED 是纯财务投资者还是商业上活跃
Microsoft云与逻辑量子位伙伴通向开发者和混合 HPC 分销的关键路径Azure 可用性和 12 逻辑量子位合作衡量 Azure 驱动的订单和获客
RIKEN系统客户 / 安装站点大额收入集中和旗舰系统部署四套商用系统之一位于日本 RIKEN 校区建模客户集中度和续约风险

梳理留存 2024-2026 年来源中明确具名、经济重要性最高的控制、资本、云和部署利益相关方。

[CO003, CO018, CO021, CO024, CO034, CO035]
FO003: 快照 KPI

定义 Quantinuum 当前成熟度和投资争论的关键规模与风险指标。

净亏损项按绝对值展示;详情说明它是亏损。客户数不纳入,因为保留的公开来源没有披露精确数字。

[CO025, CO027, CO029, CO033, CO034, CO048]

1.4 里程碑、合作伙伴与反向指向

Quantinuum 的里程碑记录在技术进展和蓝筹生态构建上最强。2021 年合并后,公司将 Quantum Origin 和 InQuanto 产品化,刷新 lambeq,并把 H-Series 硬件从早期商用 H1 系统推进到 H2,再推进到 Helios。2024 年,公司推出 56 量子比特 H2-1 系统,宣布 2030 年 Apollo 路线图,并与 Microsoft 合作创建 12 个逻辑量子比特,同时把 InQuanto 接入 Azure Quantum Elements。2025 年,公司加深了与 NVIDIA、JPMorganChase、Mitsui、Amgen、BMW、SoftBank 和新加坡国家量子机构的生态关系,并在 2025 年 11 月商业发布 Helios,配备 98 个物理量子比特和 48 个逻辑量子比特。 反向指向在于,商业化仍落后于技术进展。2026 年 IPO 文件公开说明,Quantinuum 仍是一家早期商业公司:收入不大、亏损很大,最激进的估值结果还依赖 Sol、Apollo 等未来系统。独立分析也指向具体结构性风险:离子阱未必会成为胜出架构;收入集中在 RIKEN、政府关联账户等少数客户;同位素富集材料和氦等关键投入带来供应端脆弱性。换句话说,Quantinuum 带着量子公司中少见的强技术可信度进入公开市场,但仍有前沿计算公司的经典风险:架构选择、客户集中度和变现时点。[CO038, CO039, CO040, CO041, CO042, CO043]

里程碑表
日期事件类型金额 / 估值 / 状态参与方含义
2021-11-30Quantinuum 成立创立合并完成Honeywell Quantum Solutions 与 Cambridge Quantum 双方创建 Honeywell 多数控制的全栈量子平台
2021-12-01Quantum Origin 全球发布计划产品网络安全产品发布窗口Quantinuum确立网络安全作为早期变现切入口
2022-03-29lambeq 可访问性更新产品工具包重大更新Quantinuum AI / NLP 团队扩大开源软件版图
2022-05-24InQuanto 独立发布产品商业软件发布Quantinuum、BMW、Honeywell、JSR、Mitsui 等合作方展示从研究走向化学软件收入的路径
2023-02-14Raj Hazra 出任 CEO治理领导层交接Quantinuum;Ilyas Khan叙事转向规模化执行
2023-12-13Mitsui / EAGLYS / Quantinuum 网络安全合作合作Quantum Origin 集成进 DataArmorEAGLYS;Mitsui;Quantinuum展示在亚洲的网络安全商业化
2024-01-16$300M 融资完成融资$300M,投前估值 $5BJPMorgan Chase、Mitsui、Amgen、Honeywell 等投资方独角兽估值正式确立
2024-06-05H2-1 发布产品56 个离子阱量子位Quantinuum;JPMorganChase技术领先主张增强
2024-09-10路线图 + 12 个逻辑量子位产品2030 年前 Apollo 路线图;12 个逻辑量子位Quantinuum;Microsoft抬高商业容错预期
2025-03-18NVIDIA 研究中心合作合作创始合作方身份Quantinuum;NVIDIA将路线图连接到 AI 和 CUDA-Q 生态
2025-09-04$600M 融资公告融资投前估值 $10BQuanta、NVentures、QED、JPMorganChase、Mitsui、Amgen、Honeywell 等投资方估值较 2024 轮翻倍
2025-11-01S-1 披露 B 轮融资融资$838.8M B 轮优先股Quantinuum 投资者显示 2025 年末融资超过最初头条金额
2025-11-05Helios 商业发布产品98 个物理量子位 / 48 个逻辑量子位Quantinuum、Amgen、BMW、JPMorganChase、SoftBank 等合作方把下一代系统推向市场
2026-03-18CFO 任命治理Nitesh Sharan 加入Quantinuum补强面向上市公司的财务领导力
2026-04-22秘密提交 S-1 消息披露监管注册声明草案已提交Honeywell;Quantinuum启动正式 IPO 审查
2026-05-08公开 S-1 披露亏损和收入反向$30.9M 收入 / $192.6M 净亏损Quantinuum;SEC;Reuters把估值争议和商业化风险带到公众视野

这是本章截至 2026-05-26 关于公司成立、融资、产品、合作、治理、监管和反向里程碑的公开记录时间线。

[CO001, CO011, CO014, CO020, CO023, CO025]
FO001: 公司里程碑时间线

从 2021 年合并到 2026 年 IPO 申请及不利财务披露的公开里程碑记录。

仅披露月份的里程碑,如果保留来源没有给出更精确日期,则使用当月第一天。

[CO001, CO010, CO017, CO020, CO023, CO040]

1.5 图表

Chapter 02

02市场分析

2.1 市场边界与纳入口径

分析 Quantinuum 时,应把它放进窄口径的全栈量子计算市场,而不是整个 HPC、网络安全或高级软件市场。公司自己的产品面覆盖离子阱硬件、量子计算即服务访问、开发者工具、化学 / 材料工作流,以及量子随机性安全产品。外部市场报告使用类似的分层分类:硬件平台、软件开发工具、云服务和行业应用。这个框架很重要,因为它把核心市场绑定在可识别的量子专项预算上,而不是把所有相邻算力支出都当成可触达市场。 因此,纳入口径应覆盖 H-Series 系统的直接订阅或预留访问;Nexus 等云访问和编排层;围绕 TKET/Qermit/lambeq 类工作流的编译器、工作流和开发者工具支出;化学 / 材料与优化试点的应用层支出;以及与 QRNG 或后量子迁移准备相关、窄口径定义的网络安全支出。排除项应包括通用超算容量、通用云基础设施、传统 HPC 软件、经典优化工具,以及未购买量子专项产品的广义网络安全或半导体预算。 最重要的相邻边界是 HPC。Airbus 明确把量子计算描述为与传统 HPC 协同服务航空航天仿真,而不是取代它。对 Quantinuum 而言,这意味着航空航天、化学和公共实验室需求往往先作为现有研究或创新预算中的增量科目出现,而不是直接替代一家企业的完整经典计算栈。[CM001, CM002, CM003, CM005, CM006, CM007]

市场定义表
细分 / 类别包含支出排除支出买方 / 付款方对 Quantinuum 的意义
量子硬件访问订阅、预留系统时间、直接访问 H-Series 离子阱系统通用超算集群、半导体晶圆厂、非量子实验室设备量子 R&D 负责人、公共实验室、创新预算核心市场:直接硬件变现
量子云访问与编排Nexus、多后端访问、模拟器、工作流管理、混合执行没有量子层的通用云计算或存储创新团队、HPC 研究人员、中央 R&D 预算核心市场:降低采用摩擦,拓宽上市路径
量子开发者工具围绕 TKET 类栈的编译器、SDK、工作流和算法工具支出通用 DevOps、MLOps 或经典编译器支出量子软件团队、技术负责人、研究小组核心市场:先落地再扩张的入口
化学 / 材料应用用于分子、催化剂或材料模拟的量子专用软件和服务没有量子工作流的传统化学信息学和经典模拟许可计算化学负责人、工业 R&D、先进材料项目高度相关:当前最强应用证明
优化和金融试点物流、电信、风险管理、欺诈或建模中的量子专用优化试点通用运筹研究工具和经典优化支出量化团队、运营创新负责人、银行研究团队中高相关:试点可见,但仍偏选择性
量子网络安全 / PQC 赋能QRNG、熵源,以及 Quantum Origin 等后量子就绪产品整个网络安全栈或一般网络安全支出CISO、平台安全团队、公共密码学相关方中等相关:范围窄,但相邻市场可变现
HPC / AI / 半导体相邻市场更大计算项目中的量子专项支出通用 HPC、AI 加速器、数据中心资本开支和芯片制造支出工程、计算或基础设施预算仅属相邻市场:不应整体计入 TAM

纳入的支出刻意限定在量子专项。广义 HPC、AI 和安全预算,只有在买方采购量子访问、工具、应用或 QRNG/PQC 产品时才计入。

[CM001, CM005, CM006, CM007, CM014, CM040]

2.2 用多重视角测算,而不是套一个炒作数字

当前市场规模证据支持谨慎的多视角方法。QED-C 的 2026 年行业报告把 2025 年量子市场规模定为 $1.9B,同时显示生态仍然很小:556 家纯量子公司、16,482 名纯量子从业者、8,261 个新增量子相关岗位。这是本章最保守的当前收入视角,因为它锚定行业普查,而不是长期供应商模型。相比之下,The Business Research Company 估计 2025 年市场为 $3.62B,并预测 2030 年达到 $16.27B。ResearchAndMarkets 的公开摘要拉得更宽,把硬件、软件、云服务和行业应用都纳入 2026–2046 年维度。 这些数字不能互换。QED-C 的数字更像实测当前市场;供应商报告更像品类扩张模型,纳入未来采用和更宽的范围定义。公共资金又是第四个视角:NIST 2026 年宣布约 $2.013B 的意向书,说明公共部门投入生态的资本已经可以与实测当前市场规模相当。对 Quantinuum 来说,这一点重要,因为政府和公共实验室是少数愿意在广泛商业 ROI 被证明之前就为能力买单的买方。 实务结论是,Quantinuum 近期 SAM 小于最乐观 TAM 标题暗示的规模。严谨模型应从今天实测的技术栈出发,只加入那些 Quantinuum 有直接工作负载匹配、且公开证据显示买方意图的垂直领域。[CM008, CM009, CM010, CM011, CM012, CM013]

TAM / SAM / SOM 或规模测算视角表
发布方 / 视角年份地区数值 / 指标增长方法可信度局限
QED-C 当前市场视角2025全球$1.9B 市场规模年均增长 30%覆盖量子全栈的行业普查和生态跟踪目前公开的当前市场视角中最有参考价值,但仍取决于联盟如何界定可计入市场收入
QED-C 生态构成视角2025全球7,420 家机构;556 家纯量子公司;16,482 名员工;8,261 个职位空缺机构数较 2024 年 +14%;纯量子公司 +8%;职位空缺 +11%围绕机构、公司、劳动力和职位空缺的生态普查有助于判断规模和瓶颈,但不是直接收入指标
The Business Research Company 市场视角2025 基准 / 2030 预测全球2025 年 $3.62B;2030 年 $16.27BCAGR 33.7%自上而下的市场模型,覆盖硬件、软件、服务、部署模式和终端用户类别口径可能宽于当前已兑现的供应商收入,也可能把相邻服务纳入基准年
ResearchAndMarkets 公开摘要视角2025 快照 / 2026-2046 周期全球2024 年全球投资 >$1B;2025 年平均单轮融资额 >$50Mn/a覆盖硬件、软件工具、云服务和应用的品类扩张视角公开摘要没有给出单一的当前市场头部规模数值,因此更适合作为广度视角,而不是基准 TAM
NIST / CHIPS 公共资金视角2026 公告美国$2.013B 拟议意向书n/a面向代工厂和量子计算公司的直接公共部门激励政策承诺不等同于经常性客户收入,但会影响该行业的真实支出基底

本章有意并用多种量化视角。QED-C 是最保守的当前市场收入视角;供应商市场报告属于更宽的品类扩张视角;公共资金则是独立的付款方视角,不是客户收入。SAM 和 SOM 应先从保守视角搭起,再只在 Quantinuum 已有直接工作负载证据的地方扩展。

[CM008, CM009, CM010, CM011, CM012, CM013]

2.3 买方、用户与付款方分层

Quantinuum 的市场与其说按公司规模分层,不如说按工作流和预算所有者分层。在制药、化工和材料领域,用户通常是计算化学家或研究科学家,付款方则在 R&D、高级材料或创新预算中。在金融领域,用户是量化、研究或欺诈 / NLP 团队;付款方是中央创新、安全或量化研究预算。在航空航天和国防领域,公开证据显示工程和高级计算团队围绕流体力学、导航或优化问题工作,付款权更可能落在长周期工程或公共项目预算,而不是主流 IT。 政府和公共实验室构成独立分层,因为它们购买的是国家能力、开放访问测试床和人才培养,而不只是即时应用 ROI。网络安全是另一块不同市场:Quantum Origin 卖的不是“全部网络安全”,而是一种窄口径的密码强度增强能力,买方是 CISO、平台安全团队或公共密码学相关方。最后,开发者工具和云访问创造了软件主导的入口,工具链所有者或算法团队可以先成为初始买方,再释放更大的硬件或应用预算。 这种分层很重要,因为 Quantinuum 的变现路径很可能是先落地、再扩张。早期买方通常从访问、实验和工作流验证开始,之后才升级到更大的化学、优化或安全部署。[CM018, CM019, CM020, CM021, CM022, CM023]

细分市场 / 买方地图
细分市场买方用户付款方工作流预算负责人采用触发点
制药 / 化学 / 材料计算化学负责人或先进材料研发负责人计算化学家、仿真科学家研发或材料发现预算分子模拟、催化剂设计、反应路径建模科学组织的研发 SVP / CTO量子-经典工作流相较现有近似方法,展现更高保真度或更快筛选速度
金融量子战略负责人、量化研究负责人,或欺诈 / 创新团队量化人员、风险研究员、欺诈 / NLP 团队创新、量化研究或安全预算投资组合 / 风险建模、优化、欺诈检测、语言工作流全球量化研究负责人 / 创新负责人试点在高价值工作流上证明差异化表现
航空航天 / 国防先进计算、工程或导航研究负责人航空航天工程师、仿真团队、国防研究人员工程研发、项目办公室或公共合同流体动力学、飞行力学、燃料电池化学、优化、导航工程副总裁 / 项目高管量子工作流在不取代 HPC 的前提下,改进受约束的仿真或优化问题
政府 / HPC 研究人员机构项目经理、国家实验室负责人,或研究中心主任研究人员、测试床运营方、国家实验室科学家联邦项目、实验室预算或公共研究资助开放测试床、任务科学、战略能力建设机构或实验室项目办公室对本土能力、人才培养或开放访问基础设施有战略需求
网络安全 / PQCCISO、平台安全负责人,或公共密码学办公室安全工程师、密码学团队安全架构或合规预算熵生成、密钥管理加固、后量子就绪CISO / 首席产品安全官需要更强随机性,或更早推进 PQC 迁移准备
量子开发者 / 软件主导买方算法负责人、平台架构师、创业公司 CTO量子开发者、研究人员、ISV 团队研发工具或开发者平台预算电路构建、优化、编译、实验CTO / 研究负责人在绑定某家硬件供应商之前,需要跨多个量子后端做原型

同一家企业可能出现在多行。实践中,Quantinuum 往往先从研究或开发者预算切入,再放大到更大的垂直领域项目。

[CM018, CM019, CM020, CM021, CM022, CM023]
可服务细分市场成熟度表
细分市场当前公开证据预算路径近期契合度主要扩张堵点对 Quantinuum 的意义
制药 / 化学 / 材料强:InQuanto 和 QIDO 是明确的产品与合作伙伴证据工业研发和发现预算需要可重复 ROI 和更大的问题规模目前最契合全栈软件 + 硬件 + 混合工作流
金融 / 优化中等:已有 HSBC 和 JPMorgan 证据,但生产转化不透明创新、量化研究或安全预算中高试点到生产的转化,以及受监管模型验证有助于拿下高端灯塔客户和思想领导力
航空航天 / 国防中等:Airbus 已阐明用例,但采购周期长工程研发或公共项目预算量子仍是 HPC 的补充,公共合同推进慢对长周期战略价值重要,但近期不太可能贡献规模
政府 / 国家实验室 / HPC 研究强:NQCO、DOE、Berkeley Lab 和 CHIPS 激励都很明确公共项目、资助或实验室预算采购复杂,同业争夺公共资金政府是少数会在广泛 ROI 出现前为产能买单的付款方
网络安全 / PQC中等:Quantum Origin 范围窄,但已有具体产品市场契合安全架构或合规预算迁移窗口已经紧迫,但广泛量子破解日期仍不确定如果安全买方接受其架构,可在量子计算全面实用化之前变现
开发者 / 软件主导切入强:TKET / 开发者工具和云访问降低进入门槛研发工具预算开源替代品和间接变现是后续硬件或应用扩张的关键导流细分市场

成熟度不等于绝对 TAM。从现有证据看,政府 / 公共研究和化学最强;航空航天和金融具备战略意义,但仍偏选择性。

[CM020, CM021, CM022, CM023, CM024, CM025]
FM001: 买方 / 细分市场准备度地图

按序数展示 Quantinuum 在六个目标细分市场中,哪里当前证据最清楚、预算负责人最清晰、约束最重。

这些标签是序数判断,不是数值评分。它们按细分市场概括当前公开证据强度、预算可见度和约束强度。

[CM020, CM021, CM022, CM023, CM024, CM025]

2.4 采用驱动因素与约束

Quantinuum 的增长逻辑可信,但边界很紧。正面看,公司的离子阱架构提供全连接、中途操作等特性,这些特性与早期采用者最常提到的优化和仿真类别有关。云和混合编排进一步降低采购摩擦,因为买方无需自建基础设施即可实验。政府战略也是强顺风:在一般企业需求成熟之前,国家安全和产业政策参与者仍会继续资助生态。 约束同样重要。QED-C 仍预计首批有用应用将在约三到五年内出现,并称容错量子计算仍需多年。这一时间表使市场高度依赖工作负载选择。人才也是硬瓶颈:行业仍需要能打通物理、工程和软件的人,公开指引也明确指出对外国人才的依赖。出口管制和视同出口规则给跨境团队带来直接摩擦,可能放慢销售和研究合作。最后,受监管买方面临数据主权和国家安全规则带来的额外负担,尤其是在金融、制药、航空航天 / 国防和公共部门项目中。 实际上,采用不是一条简单线性的硬件曲线。它是闸门式过程:工作负载选择、云访问、出口管制审查和商业案例验证都必须通过,试点才会变成经常性支出。[CM023, CM027, CM028, CM029, CM030, CM031]

增长驱动因素与约束表
驱动因素 / 约束方向时间影响尽调追问
囚禁离子灵活性(全连接、中途电路操作)增长驱动因素当前支撑契合 Quantinuum 产品叙事的优化和仿真工作负载用真实客户工作负载做基准,不只看设备层指标
云端和混合访问模式增长驱动因素当前让买方无需承诺本地硬件即可实验客户扩张中有多少比例始于云端,随后转化为直签合同?
政府战略资金增长驱动因素2026+在广泛企业 ROI 出现之前,公共付款方可以为产能和生态成熟度买单公共资金中有多少可被 Quantinuum 获取,而不是流向组件供应商或同业?
跨垂直领域的可见试点证据增长驱动因素当前金融、化学和公共部门证据让市场比泛泛 TAM 演示更可信哪些试点已经转化为多年期经常性收入?
容错时间表约束近期至中期在更大规模容错系统出现前,价值被限制在少数工作负载哪些产品线在容错前就依赖实用性,哪些要等到容错之后?
人才短缺约束当前拖慢客户部署、供应商招聘和生态扩张Quantinuum 如何跨物理、工程和软件招聘?流失率如何?
出口管制和视同出口约束当前增加研究合作、招聘和国际商业扩张的复杂度按地区、国籍结构和受出口管制的工作流拆分敞口
数据主权和行业监管约束当前提高金融、制药、航空航天 / 国防和政府工作负载的合规负担哪些地区和行业需要本地部署、特别审批或工作流重设计?
市场平台和生态竞争约束当前买方可借助 AWS、Azure、IBM 或其他网络触达其他硬件和软件供应商衡量直接 Nexus 访问相对于第三方平台访问的胜率

最强的驱动 / 约束组合,是政策资金推动的动能与工作负载选择性技术成熟度之间的拉扯。这个市场可以继续增长,同时仍让泛计算叙事落空。

[CM027, CM028, CM029, CM030, CM031, CM032]
FM002: 采用漏斗或价值链地图

Quantinuum 的采用通常从工作负载假设开始,经过访问采购、混合验证和合规审查,之后才进入规模化项目预算。

该流程是概念性而非按时间缩放。它突出试点在变成经常性支出前最常卡住的位置。

[CM005, CM023, CM028, CM030, CM032, CM033]

2.5 相互矛盾的估算与尽调缺口

这个市场的主要尽调风险是虚假精确。公开估算差异很大,因为有些来源衡量当前已实现收入,另一些则建模未来技术栈,把基础设施、软件、云访问和长周期行业应用都纳入其中。因此,QED-C 的 2025 年 $1.9B 市场规模,可以与 The Business Research Company 的 2025 年 $3.62B 估计并存,也可以与其他供应商更宽的多十年市场叙事并存。应把这些保留为彼此矛盾的视角,而不是平均成一个人为合成的“共识 TAM”。 Quantinuum 自身的不透明进一步放大问题。公司没有公开拆分硬件、Nexus、InQuanto 和 Quantum Origin 的收入、订单、定价或产品组合贡献。公开合作伙伴公告证明金融、化学和材料工作流有兴趣,但不披露合同规模、续约行为,或研究试点转为规模化生产支出的转化情况。同样,公共部门承诺金额很大,但 Quantinuum 相对同行能拿到的具体份额没有披露。 因此,估值时正确姿态应是受证据约束。按当前实测需求、云访问采购和已有直接证据的垂直工作流来承销市场;其余都应视作上行情景,仍需证明算法准备度、人才供给、出口合规和预算转化。[CM016, CM017, CM037, CM038, CM043, CM044]

矛盾估计与尽调缺口表
问题保守视角扩张视角差异原因尽调追问
当前市场规模2025 年 $1.9B 市场规模(QED-C)2025 年 $3.62B 市场规模(TBRC)当前市场口径与更宽供应商市场分类的差异在估值或市场份额计算中使用前,先统一各来源口径
增长周期3–5 年内出现有用应用;FTQC 仍需数年(QED-C)2026–2046 年多十年商业化叙事(ResearchAndMarkets)运营就绪时间比长期资本市场叙事更窄将近期投资假设与乐观情景可选性分开
公共资金与客户收入2026 年美国公共激励 $2.013B单个供应商能捕获多少私人经常性收入并不清楚商业需求尚未广泛出现前,政府就可以大规模支出要求管理层把收入敞口映射到公共项目和商业客户
网络安全时点PQC 迁移现在就很紧迫量子计算机距离破解当前密码体系,可能仍需数年甚至数十年迁移时间表与硬件时间表不是一回事将 Quantum Origin 建模为就绪 / 安全产品,而不是一般量子计算实用性的代理
细分市场经济性合作伙伴证据显示金融和化学有兴趣公共合同金额、定价和续约仍未披露用例证据并不说明变现质量要求提供产品线定价、ACV、试点转化率和续约数据

这个市场出现矛盾不是写作错误,而是其本质特征。矛盾来自市场边界、时间周期和付款方假设不同。

[CM016, CM017, CM031, CM035, CM037, CM043]

2.6 图表

Chapter 03

03竞争格局

3.1 竞争版图、技术模态与真实买方替代项

Quantinuum 的竞争版图比显而易见的离子阱对标更宽。直接模态对手包括离子阱领域的 IonQ,超导系统领域的 IBM、Google、Rigetti、IQM 和 Oxford Quantum Circuits,中性原子领域的 QuEra,以及光子领域的 PsiQuantum。D-Wave 与其说是正面对抗的架构同行,不如说是买方替代项——这些买方真正要解决的是近期优化,而不是门模型容错。现状选项也依然强大:硬件市场仍不成熟时,许多企业完全可以理性地继续使用经典 HPC、模拟器或内部探索团队。这很关键,因为 Quantinuum 不是只靠强硬件就能赢。它必须同时打败三类替代:资金更足的通用硬件平台、更容易购买的替代方案,以及默认继续等待的选择。按今天的公开记录,Quantinuum 在买方重视高保真离子阱性能和企业部署选项的场景最强;但市场仍足够开放,技术模态多样性本身就是竞争风险,而不是胜者已定的证据。[CP001, CP002, CP003, CP004, CP017, CP018]

竞争对手画像表
竞争对手 / 替代方案模态商业触点规模信号目标买方战略角度 / 局限
Quantinuum囚禁离子直接订阅、Azure 分销、云端与本地部署 Helios / H-Series 访问Helios 以 98 个物理量子比特、云端 + 本地部署访问,以及 BMW、JPMorganChase、SoftBank、Amgen 等企业客户背书对外营销化学、材料、网络安全、金融和先进研发领域的企业与研究买方目前最好地结合了囚禁离子性能和企业级包装,但公开标价仍不透明
IBM Quantum超导通过 IBM Quantum Platform 提供开放、按量付费、Flex、Premium 和本地部署计划30+ 个 100 量子比特以上系统,2,300+ 个可用量子比特,97% 正常运行时间,3.9T+ 个电路希望使用成熟平台工具的研究人员、企业和 HPC 相关机构公开产品中采购准备度最高,但 Qiskit Runtime 会提高工作流粘性
Google Quantum AI超导选择性 Willow 早期访问,加上 Cirq 软件和研究合作2026 年 Willow 访问仍未公开,仍需按提案筛选追求前沿实验的高端研究伙伴和机构长期生态威胁强,近期采购压力较弱
IonQ囚禁离子直连云、兼容 AWS / Azure 的工作流,以及 Forte Enterprise 本地部署系统2026 路线图目标为 100-256+ 个物理量子比特和 12 个逻辑量子比特;2026 年 Q1 收入达到 $64.7M需要多种访问模式的囚禁离子系统的企业和政府与 Quantinuum 商业重叠最接近,尤其是在直接云访问和本地部署都重要的场景
Rigetti超导QCS 云、混合 API 和本地部署 Novera 采购2026 年 4 月部署 Cepheus-1 108Q;两套 Novera 订单合计约 $5.7M希望深度访问技术栈的 HPC 相关团队和量子工程团队相比头部保真度领先,更侧重开放性和混合延迟竞争
IQM超导面向大学和 HPC 中心的本地部署系统,加上量子云访问官网对外展示 5、20、54 和 150 量子比特系统主权、学术和国家计算买方在主权 / HPC 采购上更强,在公有云心智上更弱
QuEra中性原子Amazon Braket 加高级直连访问Aquila 提供 256 个量子比特和每周 100+ 小时云使用;路线图称已实现 256 个物理量子比特和 >10 个逻辑量子比特专注仿真、优化和 ML 探索的研究人员与企业中性原子提供了可信的替代路径,但商业实用性主张仍处早期
Oxford Quantum Circuits超导云、主权托管和 API 嵌入官方信息强调当下可用的商业级系统,而不是公开量子比特数或收入披露金融、制药、国家安全和主权数据中心买方以部署地域和企业姿态竞争,但公开规模证据更薄
PsiQuantum光子没有广泛公开云访问;建设实用规模系统和政府支持设施官方技术栈列出数千片晶圆、>1M 个已测试器件、300mm 晶圆,以及 Chicago / Australia 实用规模项目愿意为未来架构押注买单的政府和长周期战略买方如果制造路线假设能转化为可用系统,长期威胁可能很大,但近期商业访问缺位
D-Wave / 优化替代方案超导退火 + 混合求解器Leap 云端和本地部署 Advantage2 系统官方系统图显示 4,400+ 量子比特、20 路连接、Leap 99.9% 可用性,以及百万变量混合求解主张优化负载很重、现在就需要生产使用,而不是等待通用门模型路线图的企业对部分工作负载是真替代方案,虽然并非通用门模型的一一对应竞争者
自建 / 经典 HPC 现状路径经典模拟器、HPC 和有限试点访问内部研究团队、公开模拟器,以及偶发的厂商额度或基于提案的访问通常比押注单一未成熟硬件厂商更便宜,也更容易治理正在探索量子准备度、但短期不部署生产系统的企业仍是耐久的默认选项,因为许多厂商主张仍领先于广泛商业可用性

各行聚焦 2026 年来源包中与 Quantinuum 相关的具名对手和实际替代方案;规模信号混合了官方硬件、云端和商业化披露,而不是采用单一归一化指标。

[CP001, CP002, CP003, CP005, CP006, CP008]
FP001: 竞争定位地图

序数评分显示,IBM 和 Quantinuum 在当前商业开放度与长期企业可信度上位置最高;Google 和 PsiQuantum 更偏长期战略价值,而不是今天容易采购。

评分是基于已引用产品、访问和商业化证据的序数综合,不是基准测量。

[CP005, CP010, CP012, CP018, CP021, CP029]

3.2 产品范围、云访问、软件栈与公开打包

IBM 和 IonQ 是商业上最可比的基准,因为两者都暴露具体采购路径,而不只是研究雄心。IBM 发布按分钟计费的 open、pay-as-you-go、flex、premium 和 on-prem 层级,并把它们与 Qiskit Runtime 和 Functions 绑定。IonQ 提供直接云访问、对所有主流 SDK 的支持,以及面向数据中心部署的 Forte Enterprise;其 2026 年业绩公告也显示出实质商业牵引,而不只是技术进展。Google 不同:Cirq 是开源的,但 Willow 访问仍需提交提案,而且明确不公开,因此 Google 眼下的威胁更多来自生态和人才引力,而不是采购规模。Rigetti 通过混合 HPC 集成和真实本地产品 Novera 竞争。IQM 更偏主权和 HPC 买方。QuEra 通过 AWS 加高级直接访问打包中性原子替代方案,OQC 则强调主权托管和云部署。在这个组合中,Quantinuum 在部署广度和软件成熟度上表现不错,但没有 IBM 和 AWS Marketplace 上架信息提供的同等公开价格透明度。[CP005, CP006, CP007, CP008, CP009, CP010]

功能 / 能力矩阵
采购标准Quantinuum证据最充分的替代方案竞争含义
硬件架构采用全连接和电路中测量的囚禁离子 QCCDIonQ 囚禁离子;IBM / Google / Rigetti / IQM / OQC 超导;QuEra 中性原子;PsiQuantum 光子Quantinuum 在当前囚禁离子定位上占优,但全市场的架构领导权仍未定
直接企业部署系统可通过直接订阅、本地部署和 Azure 访问IBM 本地部署和年度计划;IonQ Forte Enterprise;Rigetti Novera;OQC 托管部署;D-Wave 本地部署Quantinuum 在这里很强,但并非唯一选择
开放软件 / 编译器层pytket 可导入并编译到多种格式和后端Qiskit、Cirq、IonQ 多 SDK 支持、Rigetti Quil / pyQuil软件生态帮助采用,但也削弱应用层锁定效应
云市场触达Azure 渠道加直接服务AWS Braket 覆盖 IonQ、IQM、QuEra、Rigetti;IBM 直接平台渠道触达重要,因为买方在标准化前越来越倾向多处下注
公开定价透明度以报价为主,而非明确标价IBM 披露分钟档位;AWS 披露按 shot / 任务 / 预留定价;Rigetti 公开 Novera 商业化即便技术定位很强,不透明定价也会拖慢采购速度
当前商业化证明企业客户名单以及云端 / 本地部署可用性IBM 机群和计划;IonQ 收入和 RPO;D-Wave 生产主张;Rigetti 云端历史商业压力最大的是同时具备访问能力和付费证据的厂商
逻辑量子比特 / FTQC 叙事Quantinuum 围绕 Helios 上的实时纠错和逻辑量子比特做市场叙事IBM 模块化路线图、IonQ 2026 逻辑量子比特目标、QuEra 逻辑量子比特路线图、PsiQuantum 制造论点路线图已经是入场券;买方仍需要证据证明它能转成可用系统
当下工作负载契合度化学、材料、网络安全、金融和高级研发QuEra 和 D-Wave 更偏仿真 / 优化;IBM 和 Rigetti 更偏广泛门模型实验Quantinuum 覆盖面广,但替代架构仍可能赢下更窄的高价值任务

该矩阵总结的是公开证据中支撑最充分的部分,不包括隐藏客户条款或未公开基准数据;证据不足的单元格以定性方式表述,而不是猜测。

[CP002, CP004, CP007, CP009, CP010, CP011]
定价 / 包装对比
提供方 / 路径公开包装信号公开价格信号证据支持的注意点买方含义
Quantinuum 直接 / Azure订阅、赠款、Azure 访问和本地部署咨询引用页面未公开标价访问模式清楚,但企业买方仍需要销售流程确认确切经济性企业包装很强,公开采购透明度较弱
IBM Quantum Platform 平台开放、按量付费、Flex、Premium 和本地部署免费开放层;PAYG $96/min;Flex 每年 400 min 起 $72/min;Premium 每年 5,200 min 起 $48/min;本地部署报价IBM 定价按分钟而非结果计费,并假设使用 IBM 平台这一组里最透明的通用门模型方案
Google Willow选择性早期访问研究计划无公开价格2026 年该计划基于提案,且不公开采购入口受限,商业压力因此延后
IonQ 直接云按需或预留,加直接云控制台和企业工作流工具官方页面披露访问模式,但未披露标价经济透明度通过 AWS 市场价格提升,而不是来自直接页面费率即便直接标价不透明,包装宽度仍不错
AWS Braket 上的 IonQ按任务、按 shot 或按小时预留IonQ Forte 示例:$0.30/task 加 $0.08/shot;误差缓解至少需要 2,500 shots市场经济性可能让低规模实验变得昂贵是买方比较囚禁离子访问成本的有用基准
Rigetti Novera可立即发货的本地部署 QPU,加合作伙伴生态商业产品已存在,但引用页面没有标价公开商业化信号强于公开单价即使云价格不易比较,也帮助 Rigetti 参与硬件采购竞争
AWS Braket 上的 Rigetti按任务、按 shot 或预留使用定价示例:Rigetti Cepheus 上 $0.30/task 加 $0.000425/shot对云端实验来说,示例披露比总体系统经济性更清楚超导实验的每 shot 成本明显低于 AWS 上的囚禁离子示例
AWS Braket 上的 QuEraAWS 按需访问加高端直接访问AWS 列出 QPU 的按任务 / 按 shot 定价结构,但买方支出仍高度取决于工作负载设计中性原子访问已经可用,但成本可比性取决于问题类别,而不是原始量子比特数量无需定制企业合同,QuEra 就能成为探索阶段的实际选项
D-Wave Leap / Advantage2通过混合求解器提供即时云端和本地部署访问官方页面营销访问和试用,而不是在引用页面给出标准化公开按工作负载定价有用的生产姿态比通用门模型可比性更清楚买方看重即时优化结果而非通用性时,这是强替代方案

定价证据混合了直接标价和市场示例。缺少准确价格时,本行记录包装路径,而不是杜撰实际单位经济性。

[CP003, CP005, CP008, CP010, CP011, CP015]
FP002: 功能广度 / 能力地图

大体格局是,多数厂商能勾上几个选项,但只有少数能同时给出直接云、本地部署路径、开放开发者工具和透明公开定价信号。

[CP003, CP004, CP005, CP008, CP010, CP015]

3.3 切换成本、多栖采购、分发权力与伙伴访问

反驳持久赢家通吃硬件护城河的最强证据,是访问层已经相当标准化。AWS Braket 让买方通过一个控制平面和一个计费界面对比 IonQ、IQM、QuEra 和 Rigetti;Quantinuum 自己也把 Azure 访问与直接订阅、本地部署并列营销。IonQ 明确支持 Qiskit、Cirq、tket 和 Q#,Quantinuum 的 pytket 也被设计成可以导入其他指令格式并跨目标编译。这意味着软件侧切换成本真实存在,但不是绝对壁垒。买方可以对冲、实验,并让团队保持生产力,而无需完全押注单一硬件供应商。剩下的锁定效应主要是硬件特异性的:连接性、噪声画像、模拟与门模型行为、预留经济性,以及某个工作流实际跑得最好的位置。实务上,这有利于当前性能更强或渠道位置更好的供应商。对 Quantinuum 而言,护城河因此比“拥有整个栈”更窄,更接近“同类最佳离子阱性能,加上足够多部署选项,使买方多栖采购时仍把它留在短名单里”。[CP024, CP025, CP026, CP027, CP028, CP035]

3.4 护城河耐久性、商品化风险与第三方怀疑视角

反向证据并不是某个对手已经决定性击败 Quantinuum,而是整个品类仍太早,护城河叙事容易跑在商业现实前面。IEEE Spectrum 2026 年 1 月的观点说得很明确:行业不会在 2026 年实现广泛有用的商业量子计算;即便首批面向客户的纠错系统,也被框定为科学优势,而不是商业优势。这个怀疑很重要,因为它限制了任何供应商仅凭路线图主张短期防御性的空间。在这一背景下,Quantinuum 当前优势真实存在:离子阱保真度、全连接、企业证明点和多条部署路径。但长期威胁也可信。IBM 的模块化路线图是大型企业供应商中最清晰的公开容错扩展计划;QuEra 正在快速推进逻辑量子比特主张;PsiQuantum 则试图用半导体式制造论题绕开当前云竞赛。因此,竞争结论是有利但不能自满:Quantinuum 看起来位居今天企业短名单的一线,但市场结构仍可能在工作流层商品化,也可能被另一种胜出架构替代。[CP021, CP022, CP023, CP029, CP030, CP031]

护城河耐久性 / 竞争风险清单
护城河主张反向证据 / 对手压力严重性什么会改变结论
当前最好的囚禁离子企业包装IonQ 现在已在直接云、多 SDK 支持和本地部署 Forte Enterprise 上重叠Quantinuum 需要更清楚的性能或商业证明,且能经得起与 IonQ 的直接比较
软件生态降低 Quantinuum 用户的锁定效应同样的开放性意味着 pytket、Qiskit、Cirq 以及 AWS/Azure 路径也会降低竞争对手的锁定效应只有当 Quantinuum 把软件转化为独特的更好结果,而不只是兼容性,护城河才会改善
云和合作伙伴触达具有防御性AWS Braket 已经聚合多种竞争模态;IBM 运营自己的成熟平台;Azure 不是独家分发独家渠道不太可能出现,所以渠道优势必须来自更好的服务和企业采用
高保真度创造耐久硬件护城河IEEE 和竞争对手路线图显示,该品类仍未形成共识,IBM、QuEra 和 PsiQuantum 正在推进不同扩展路径如果 Quantinuum 持续把保真度转化为经过基准验证的逻辑量子比特和应用胜利,耐久性会上升
近期买方紧迫感利好 QuantinuumD-Wave 和经典 HPC 现状路径今天就能解决部分优化任务,无需等待通用门模型成熟Quantinuum 需要拿出应用证明,显示通用门模型能在价值兑现速度上击败替代方案
路线图支持长期领导地位Google、IBM、QuEra、IonQ 和 PsiQuantum 都在营销激进的未来状态叙事只有当路线图转化为付费访问、稳定性能和可复现工作负载,它才具有防御性
企业级部署让 Quantinuum 差异化OQC、Rigetti、IBM 和 D-Wave 都在营销云端加本地部署或托管部署模式只有部署质量伴随可衡量的性能优势,Quantinuum 才能守住优势
市场热度本身帮助在位厂商IEEE 对 2026 年的怀疑式框架表明,尚无厂商拥有广泛商业优势,这会压缩所有人的护城河主张整个品类需要可复现的经济性胜利,任何护城河叙事才值得积极承销

严重性是基于引用证据包的判断,不来自统计模型;各行聚焦 Quantinuum 当前优势最容易被商品化或架构替代挤压的位置。

[CP010, CP018, CP024, CP026, CP029, CP032]
FP003: 护城河 / 就绪度 KPI

综合 KPI 视图对 Quantinuum 当前硬件护城河和部署就绪度有利,但在价格透明度和防多归属能力上明显偏弱。

KPI 标签是基于引用证据包综合形成的判断,不是供应商报告的指标。

[CP029, CP032, CP035, CP036, CP037, CP038]
Chapter 04

04财务情况

4.1 收入模型与定价现实

Quantinuum 的申报文件终于说明,公司不是纯软件业务。S-1 称收入来自三大类:专业量子计算硬件、带维护和支持的访问合同,以及围绕 Quantinuum 系统联合开发算法的咨询工作。公开网页又把图景铺得更宽:Quantinuum Systems 直接订阅、Azure 订阅、云和 HaaS 交付、InQuanto 许可证、Nexus 平台访问,以及 Quantum Origin 安全软件。承销问题不在于是否存在变现面——显然存在——而在于公开证据几乎没有说明实际定价、折扣和收入流组合。 唯一清晰的公开标价在 Azure Quantum 上:Quantinuum 的 Standard 和 Premium 方案每月分别为 $125,000 和 $175,000,另加基础设施成本,并有单独的按量付费 HQC 公式。这说明企业访问可以开出六位数月度标价,但仍不能揭示实际 ASP、直销定价或合同层面折扣。与此同时,S-1 显示硬件交易会主导收入时点:2025 年收入受到一笔 $16.5M 销售型租赁拉动,2026 年 Q1 收入在上一年一次性确认没有重演后大幅下滑。因此,公开收入模型真实存在,但在组合和变现质量上高度不透明。[CI008, CI009, CI010, CI011, CI012, CI026]

收入流表
收入流机制单位 / 合同基础当前公开状态质量评估尽调要求
专用量子计算硬件量子系统的设计、开发、建造、销售、交付和调试;可按销售型租赁确认系统销售 / 租赁 / 调试里程碑S-1 已确认;2025 年包括一笔 $16.5m 销售型租赁交易低–中:收入真实,但确认时点偏一次性,且取决于具体合同提供硬件订单积压、已安装基础经济性、保修 / 支持成本和复购特征
Quantinuum 系统云访问含维护和支持服务的量子系统随时可用访问固定费用期限,加超过最低量的可变使用费S-1 和直接系统页面已确认;Azure 提供月度计划和 HQC 按量付费中:存在经常性结构,但实际定价和续约数据未公开提供合同最低量、超额用量定价、续约率和市场收入分成
直接订阅 / HaaSQuantinuum 直接订阅和面向系统访问的 HaaS企业订阅系统页面明确提供直接订阅和云 / HaaS中:渠道存在,但没有公开直接标价提供直接销售价目表、平均合同期限和支持义务
Nexus 平台软件带模拟器、notebooks、配额和协作功能的云工作流平台平台订阅 / 企业工作区Nexus 拥有 150+ 组织和 750+ 用户,但未公开收入中:有采用代理指标,变现不透明提供 Nexus 年经常性收入(ARR)、付费席位数,以及与硬件合同的附加率
InQuanto 软件随系统访问和合作伙伴驻留一起销售的量子化学平台许可证 / 企业合作产品页面确认平台和合作伙伴驻留模式;无公开定价中:可信的软件产品面,经济性未披露提供许可证结构、服务成分,以及按制药 / 材料客户拆分的收入
Quantum Origin 网络安全软件二进制加量子衍生种子,用于熵和后量子工作流软件许可证 / SDK / 连接器部署产品和文档页面确认软件交付;无公开定价中:产品有差异化,但收入可见度接近为零提供客户数、定价模型、经销商经济性和毛利率
咨询和共同开发服务算法共同开发、入门支持、垂直领域科学协作和咨询服务里程碑 / 服务工作说明书S-1 和领导层页面明确提到咨询服务低–中:明确存在,但服务收入金额未披露提供服务利用率、混合日费率、与产品销售的附加关系,以及向经常性软件的转化

各行列出公开文件和官方页面可见的变现入口。公开状态指存在性被披露,不代表收入组合或实际合同经济性已披露。

[CI008, CI026, CI028, CI029, CI033, CI034]
定价 / 变现表
产品面公开价格 / 单位标价与实际来源支持的注意点含义尽调要求
Azure Standard Plan每月 USD125,000 + Azure 基础设施成本,含 10k HQCs 和 100k eHQCs仅公开标价Microsoft 警告,提供方定价可能变化,且可能滞后于最新工作区定价证实六位数月费企业访问在商业上说得通要求披露扣除额度、超额费用和私下折扣后的实际价格
Azure Premium Plan每月 USD175,000 + Azure 基础设施成本,含 17k HQCs 和 170k eHQCs仅公开标价仍未披露实际价格或用量承诺数据暗示 H2 访问采用企业分层模型要求披露客户数、升级组合和平均 HQC 使用率
Azure 按量付费HQC 公式:HQC = 5 + C(N1q + 10N2q + 5Nm)/5000仅公开基于使用量的标价公式公式展示计费机制,但没有给出每个工作负载的有效 $ 或谈判后的企业包有助于发现标价,对收入预测作用较弱要求披露每 HQC 实际确认收入,以及从免费试用到付费使用的转化
Quantinuum 直接订阅公开提供,但未披露价格实际价格未知系统页面提供直接订阅,但没有标价直接渠道可能与 Azure 经济性存在实质差异要求提供直接企业销售的标准 MSA 和价目表
专用硬件租赁 / 销售2025 年有 $16.5m 收入来自一项 45 个月协议下的销售型租赁实际合同经济性部分可见,但不是可重复标价单笔交易实质性扭曲了收入确认时点硬件交易可能带来大额一次性收入峰值要求披露硬件 ASP、部署成本、维护负担和续约选项
AWS Braket 访问公开页面显示按任务、按 shot 或按小时预留模式,但未给出 Quantinuum 特定设备价格Quantinuum 的实际价格和标价都不清楚抓取到的页面未露出 Quantinuum 行AWS 是真实渠道,但 Quantinuum 在该渠道上的经济性不透明要求披露 Braket 合同经济性,或确认访问是否仅限私有 / 预留
InQuanto / Nexus / Quantum Origin未找到公开标价实际价格未知官方页面确认包装和交付,但未披露价格软件上行在战略上重要,但公开数据不足以承销要求提供各软件线的价目表、席位 / 租户数和续约指标

该表区分公开标价机制和实际经济性。空白或未披露字段意味着抓取的公开材料中不可得,不代表经济上不重要。

[CI009, CI010, CI012, CI029, CI030, CI031]
FI001: 收入模型桥接图

公开披露显示 Quantinuum 有多种变现界面,但一次性硬件交易与更经常性的访问或软件合同之间,经济性差异很大。

这张桥接图是定性分析,因为公开来源没有披露各输入节点的收入组合。

[CI008, CI009, CI010, CI012, CI026, CI029]

4.2 GTM 动作与销售效率代理指标

Quantinuum 的 GTM 更像长周期企业销售、伙伴渠道和科学服务的混合体。S-1 把业务框定为硬件主导、软件增强,官网显示直接订阅、Azure 订阅、咨询服务、InQuanto 伙伴驻场,以及多年期研究合作。公开牵引力代理指标存在,但不是常见 SaaS 效率指标。Quantinuum 披露 2025 年订单额(bookings)为 $79.3M,年末剩余履约义务约 $80.7M;Honeywell 另行披露,归属于 Quantinuum 的积压订单(backlog)约 $81–82M。到 2026 年 3 月,Nexus 也拥有超过 150 家组织和 750 名活跃用户。 这些指标方向上有用,但必须附带重大警示:公开客户集中度极高,生产规模转化仍不清楚。仅 RIKEN 就贡献了 2025 年收入的 60%,政府关联需求在 2026 年 Q1 占比更高。S-1 明确警示,许多大型关系是试点、研究合作或拨款资助项目,而不是持久生产合同。BMW 和 RIKEN 是可信标杆客户,也是商业相关性的有用证据,但还没有证明部署可以大规模重复。本章正确的销售效率解读是:Quantinuum 有强企业访问和技术可信度,但在可重复性、队列行为、CAC 或回本周期上的公开证据很差。[CI003, CI005, CI006, CI027, CI037, CI038]

GTM / 销售效率代理指标表
代理指标公开数值 / 状态为什么重要读数尽调要求
2025 年签约额79.3m USD除已确认收入外披露得最好的需求信号显示客户签约活动超过 2025 年收入,但签约额到收入的转化不透明提供签约额到已确认收入的桥接,以及取消 / 修改情况
剩余履约义务FY2025 为 80.7m USD,31% 在 12 个月内到期前瞻收入可见度代理指标提供了一定合同可见度,但时点仍取决于交付和里程碑完成按收入流、客户类型和预期毛利率拆分 RPO
Honeywell 报告的订单积压FY2025 为 ~82m USD;2026 Q1 Corporate and All Other 中为 ~81m USD订单可见度交叉验证订单积压大体印证 S-1 的 RPO 数字解释订单积压、签约额和 RPO 定义之间的差异
Nexus 采用情况截至 2026 年 3 月,150+ 用户组织和 750+ 活跃用户软件漏斗代理指标开发者 / 企业参与度信号不错,但每用户变现未知提供付费与未付费组织、活跃付费用户和使用强度
客户集中度RIKEN 贡献 2025 年收入的 60%;政府相关实体主导 2026 Q1收入质量和销售效率警示商业牵引更集中,而不是广泛铺开提供前 10 大客户组合、续约率和集中度趋势
具名企业合作BMW 多年合作;RIKEN 本地部署;Honeywell、JPMorgan、Amgen、Mitsui 活跃合作企业访问证明和长销售周期客户名单很强,但大多是合作信号,而非披露的生产合同提供哪些合作是付费试点、生产使用或赠款资助工作
员工规模~700 名员工,其中包括 ~410 名硬件专家和 ~105 名软件专家成本收入效率代理指标大技术团队意味着高固定成本和长销售支持周期提供人均收入、服务利用率和预期经营杠杆

这些是代理指标,不是标准化 SaaS 效率指标。公开证据在需求可见度和客户名单上最强,在回本周期、销售配额产出和续约行为上最弱。

[CI003, CI005, CI006, CI027, CI037, CI038]

4.3 成本结构、毛利驱动因素与资本强度

Quantinuum 的成本结构足以显示强度,但还不够清晰,无法判断毛利质量。S-1 称收入成本包括运营和可靠性人员、客户支持、系统折旧、云和设施基础设施,以及第三方承包商。2025 年直接收入成本只有 $4.73M,但摊销又增加 $11.36M,管理层没有发布能拆分硬件、云、软件和服务的毛利口径。真正的成本中心是研发:2025 年研发支出 $165.4M,而收入只有 $30.9M;销售和营销加 G&A 又合计增加 $48.7M。 资本强度也很明确。仅 2026 年 Q1,Quantinuum 就使用 $62.9M 经营现金和 $22.7M 投资现金,资本开支与量子系统开发和租赁物业改良相关。连营运资本桥也显示公司依赖实体建设和关联方支持,包括向 Honeywell 预付款和进行中的租赁物业改良。因此,公开来源支持一个简单结论:Quantinuum 目前正在为硬件、实验室、基础设施和人才承担重成本,且早于广泛商业规模。公司最终会变成软件占比更高,还是继续偏服务和硬件重型,仍是未解尽调问题,因为收入流级别的毛利没有公开。[CI013, CI014, CI015, CI016, CI017, CI018]

单位经济模型表
指标公开数值 / 状态为什么重要置信度尽调要求
2025 年净收入30.9m USD所有利润率和估值工作的基准规模提供月度收入桥接和收入流组合
2025 年直接收入成本4.73m USD展示摊销前披露的直接服务交付成本说明这是否排除了重要支持成本或系统折旧分摊
2025 年摊销费用11.36m USD与收购技术和无形资产相关的重大非现金成本按业务线提供含摊销和不含摊销的毛利桥
2025 年 R&D 费用165.4m USD衡量维持路线图成本的最佳公开代理指标将路线图 R&D 与客户专项工程或服务工作拆开
2026 年 Q1 经营性现金使用62.9m USD短期烧钱速度的最佳代理指标提供剔除一次性项目和权证噪音后的标准化季度现金消耗
按业务线划分的毛利率未公开披露核心承销判断指标仍未公开按硬件、云、软件、网络安全和咨询提供毛利率、支持负担和利用率
按客户集中度调整后的收入质量已披露期间高度集中,由 RIKEN 和政府关联需求主导单一客户依赖会扭曲表面单位经济提供客户队列数据,以及按集中度调整后的流失 / 续约情况

该表有意将已报告的收入直接成本与摊销拆开,因为监管文件没有按产品线给出管理口径的毛利率桥。

[CI001, CI002, CI014, CI015, CI016, CI017]
FI002: 单位经济模型桥接图

公开材料对成本桶披露较强,但对投资人最关心的利润率输出披露偏弱。

由于文件没有提供真正的毛利率瀑布图,图中使用已披露成本桶,而非完整毛利桥接。

[CI013, CI014, CI015, CI016, CI017, CI018]

4.4 资本充足性、Honeywell 支持与融资依赖

披露的现金余额很大,但 Quantinuum 还不能自我供血。2026 年 3 月 31 日,公司现金及现金等价物为 $677.0M,此前 2025 年 9 月融资 $600M、投前估值 $10B。这给公司带来有意义的近期灵活性,但同一份文件显示,业务仍在大量烧钱,并继续投入系统、设施和可分拆上市的公司基础设施。IPO 旨在再加一层融资,但截至本报告日期,S-1 尚未设定价格或发行股数,投资人不能把它视作已经承诺的现金跑道。CHIPS R&D 意向书作为政策支持信号有帮助,但也不是已到账现金。 Honeywell 仍是资本充足性的中心。Honeywell 文件继续把 Quantinuum 放在 Corporate and All Other 板块,披露 Quantinuum backlog,并把 Quantinuum 收入描述为 Honeywell 持股多数的软硬件量子业务的一部分。Quantinuum 自己的文件称,Honeywell 已贡献基础设施、供应链关系、管理专长和早期客户需求,并打算在 IPO 后继续作为战略客户和合作伙伴。这种支持降低执行风险,但也证明 Quantinuum 尚未作为完全独立的商业实体运营。因此,资本充足性结论是有条件的:当前现金可以买时间,但未来独立性仍取决于另一场资本事件、更顺的商业化,或两者兼具。[CI004, CI006, CI007, CI017, CI018, CI019]

资本充足性表
项目公开数值 / 状态解读依赖 / 风险尽调需索取
账面现金677.0m USD,截至 2026 年 3 月 31 日2025 年融资后资产负债表弹性充足单靠这一项不足以证明现金跑道提供月度现金瀑布和最低现金约束条款假设
2025 年经营性现金消耗160.3m USD显示年度现金消耗已超过 2025 年收入的 5 倍IPO 营销开始前,现金消耗已经上升提供 2026 年计划现金消耗和预期拐点
2026 年 Q1 经营性现金消耗62.9m USD显示当期消耗率仍然很重季度可能包含一次性项目;准确稳态未知提供标准化季度现金消耗和营运资本季节性
2026 年 Q1 资本开支22.7m USD确认系统和设施仍在投入资本强度不止来自薪酬和云成本按系统代际和设施提供多年资本开支计划
Honeywell 支持控股母公司、积压订单持有人、战略客户和服务交易对手降低执行风险,并帮助打开商业渠道造成关联方依赖和转让定价不透明提供 Honeywell 协议经济条款和 IPO 后分拆模型
2024 年融资里程碑300m USD,投前估值 5b重置合并后的资本基础历史融资不能解决当前现金跑道问题提供 2024 年资本用途完成报告
2025 年融资里程碑~600m USD,投前估值 10b提供当前现金垫,并支撑 Helios / 容错路线图如果现金消耗和资本开支维持高位,仍可能不够提供 2025 轮融资剩余非受限现金
2026 年 IPO 过桥已提交文件,但股数和价格尚未确定潜在下一笔流动性来源截至报告运行日,IPO 资金尚非已承诺资本提供基准、下行和无 IPO 融资情景
CHIPS R&D 意向书拟议的联邦支持和供应商合作可能以非稀释方式抵消未来 R&D / 供应链支出LOI 不是已到账现金,时间不确定提供拨款规模、里程碑和配套资金要求

资本充足性仅根据已披露现金和现金消耗评估。公开来源没有给出管理层现金跑道预测、IPO 后现金预算或按业务线的资本开支桥。

[CI004, CI017, CI018, CI019, CI020, CI021]
FI003: 资本强度 / 现金流地图

Quantinuum 目前现金充足,但多条现金流出通道和仍未确定的 IPO 募资额,让融资故事仍带条件。

这是一张方向性的现金依赖地图,不是正式现金流预测;公开记录没有管理层现金跑道模型。

[CI004, CI017, CI018, CI019, CI022, CI024]

4.5 公开财务缺口与承销结论

Quantinuum 的公开披露足以证明业务真实存在,但还不足以让人舒服地按基本面承销。最强正面信号包括:变现面多元,蓝筹企业和政府关系存在,Azure 至少暴露一张公开价格卡,backlog / RPO 指标在 Quantinuum 与 Honeywell 文件之间大致能对上,公司在 2025 年重大融资后仍有大额现金余额。最强负面信号同样清楚:收入相对烧钱规模很小,季度业绩不平滑,客户集中度尖锐,外部观察者已经质疑拟议公开估值究竟锚定当前经济性,还是锚定一台尚未被证明的未来机器。 因此,财务结论应保持谨慎。今天的收入质量更像混合而非强韧:一部分是硬件确认事件,一部分是云使用,一部分是研究 / 支持服务,一部分是软件雄心,还有一部分仍未披露的专业服务。毛利路径看起来可行但未被证明,因为实际定价和收入流级成本结构都不公开。资本充足性可以支撑下一段执行,但更深层看仍依赖融资,因为公司烧钱速度要求它要么 IPO 成功,要么经济性收紧,要么继续获得类似母公司的支持。本章的主要卡点是毛利率、定价兑现、收入流组合和 Honeywell 分离经济性。[CI014, CI017, CI023, CI037, CI038, CI040]

公开财务缺口表
缺失指标 / 披露重要性对承销判断的影响具体尽调路径
按业务线划分的收入结构拆开一次性硬件确认与经常性云或软件收入缺少结构时,收入表面值可能高估经常性质量索取 2024 年至 2026 年 Q1 按硬件、云、软件、网络安全、咨询和补助划分的收入
实际成交价格、折扣和最低承诺标价不等于实际经济收益无法有把握地建模毛利或定价权索取直销、Azure,以及任何 AWS 或本地部署渠道的合同样本
按产品线划分的毛利率检验软件能否跑赢硬件服务成本强度的核心指标缺少它,无法从质量角度支撑估值索取按业务线划分并含支持成本分摊的经审计毛利桥
客户续约和生产化转化试点偏重的合作会带来订单额,却不一定带来持久复购商业牵引力可能强于持久收入现实索取付费试点转化率、续约队列和按客户类别的流失率
IPO 后 Honeywell 经济条款分拆后,关联方支持可能发生实质变化独立运营利润率和现金需求可能明显变差或改善索取战略服务和供应协议,以及任何过渡服务安排
基准和下行情景下的现金跑道现金余额本身不是现金跑道预测没有预算和里程碑计划,下一轮时点仍只能猜索取 2026-2028 年经营计划和融资情景

每一行都是已抓取公开证据包中的明确缺口,不是锦上添花的信息请求。这些缺失数字会直接影响对收入质量、利润率和资本充足性的判断。

[CI014, CI032, CI037, CI038, CI040, CI041]

4.6 图表

Chapter 05

05产品与技术

5.1 全栈产品组合与面向用户的模块

Quantinuum 推向市场的不是单一离子阱盒子,而是一套分层产品组合。公开产品面把 H-Series 硬件、新旗舰 Helios、TKET 编译器栈、InQuanto 化学套件、Quantum Origin 密码产品,以及用于量子自然语言处理工作的 lambeq 组合在一起。官网和开发者工具页面明确表示,公司希望买方按终端用户工作流思考,而不是只看量子比特数量:化学团队在 Quantinuum 系统上使用 InQuanto,开发者使用 TKET 和 Guppy,安全团队使用 Quantum Origin,研究用户则可用 lambeq 这个相邻开源工具包实验。 这种宽度具有战略意义。Quantinuum 至少可以通过四个不同入口把潜在客户带进技术栈:直接硬件访问、领域软件、开源开发者工具,或网络安全产品集成。InQuanto 是最清晰的应用包,公开定位围绕分子与材料模拟、Quantinuum 系统访问和专家驻场支持。Quantum Origin 是一条可单独变现的安全产品线,围绕量子生成种子和 Bell-test 支持的随机性增强,为现有产品服务。lambeq 更小、更偏研究,但它仍让 Quantinuum 拥有一个可见的开源 QNLP 产品面,从而超出纯硬件。净效果是,公司把硬件、中间件、应用和网络安全产品放进了同一品牌栈。[CE001, CE017, CE020, CE021, CE022, CE023]

产品模块 / 资产矩阵
模块 / SKU主要用户当前状态差异化尽调缺口
System Model H1量子算法团队商业访问 / 第一代平台20 量子比特线性俘获离子 QCCD,具备全连接、MCMR 和量子比特复用公开定价和装机基数细节未披露
System Model H2研究人员和企业试点当前一代 QPU56 量子比特赛道式 QCCD,语法检查器和仿真器路径复刻托管编译流程公开正常运行时间和 SLA 指标未披露
Helios企业混合量子用户新旗舰 / 云 + 本地部署98 量子比特 Ba+ 系统,配套 Guppy、GPU 联动控制和逻辑量子比特定位公司和合作伙伴证据之外,独立基准测试仍有限
TKET / pytket量子开发者活跃维护的开源项目平台无关编译器,带扩展模块和栈级优化商业支持边界与社区使用的划分未完全公开
InQuanto化学和材料团队商业应用软件面向化学工作流、误差降低和 Nexus 集成的领域套件案例 ROI 和定价未公开
Quantum Origin安全架构师和 OEM商业网络安全产品纯软件 QRNG 增强,种子质量由 Bell 测试支撑未找到公开合规材料包
lambeq研究型开发者活跃开源工具包与 TKET 集成的 QNLP 流水线,仍在持续发新版本相对 TKET,商业需求看起来更小众

状态标签反映截至 2026-05-26 的公开证据;它们区分活跃公开产品界面,与更广泛的商业化证明或企业运营披露。

[CE002, CE004, CE007, CE011, CE017, CE020]
工作流 / 用例表
用户任务当前工作流Quantinuum 模块可衡量或声称的收益限制
量子开发者用 Python 或 Q# 编写电路,并准备在硬件上执行TKET + 语法检查器 + 仿真器在稀缺 QPU 使用前,编译到原生门,并在同一栈上验证托管端点和配额仍限制生产访问
化学研究员建模分子或材料,运行混合协议,缓解噪声InQuanto + H-Series 或 Helios公司称,相比领先开源替代方案,准确性和资源效率最高提升 10×公开证据多由公司主导,而非中立基准
安全产品团队增强现有产品内的密钥生成Quantum Origin无需额外硬件或云连接;量子种子由 Bell 测试支撑未留存公开认证材料
NLP 或 QNLP 研究员把文本转换为图和电路,并训练模型lambeq + TKET端到端 QNLP 工作流,带教程和可扩展性相比 TKET,公开采用规模较小
企业混合量子团队开发混合量子与加速经典计算的程序Guppy + CUDA-Q + Helios实时解码和混合控制路径,支持云或本地部署相比 SaaS 式工具,需要更深集成和可用性规划

收益要么有直接文档支持,要么明确标注为公司说法;限制强调公开运营证明仍比产品叙事薄的地方。

[CE017, CE018, CE020, CE022, CE023, CE029]
FE001: 产品架构图
[CE001, CE007, CE011, CE017, CE020, CE027]

5.2 离子阱架构与部署模型

硬件线条是连贯的,不是碎片化拼接。H1 是第一代系统:一台 20 量子比特的 171Yb+ 离子阱机器,采用单线性架构,配有五个并行门区、全连接、中途测量、条件逻辑和量子比特复用。H2 保持相同的离子阱逻辑模型,但扩展到 56 个物理量子比特,并转向跑道式 QCCD 架构,包含两个相连线性区段和四个并行双量子比特区。Helios 再次沿用同一设计语言,推进到 98 个物理 137Ba+ 量子比特、八个门区、一个圆形存储环加线性门控区段,并加入实时算术和控制流特性,以支持重度纠错工作负载。 同样重要的是,Quantinuum 不是把硬件作为裸金属商品端点暴露,而是通过托管访问层提供。Nexus 是公司的硬件和编译服务控制平面。Azure Quantum 以 Quantinuum 为提供方,提供语法检查器、模拟器和 H2 QPU,OLCF 文档则给出从语法检查器到模拟器再到真实硬件的推荐工作流。Helios 增加了更企业化的部署姿态,因为 Quantinuum 公开提供云和本地访问,并把硬件与 Guppy 和 GPU 联动控制绑定。架构故事因此不只是离子阱加激光,而是离子阱加托管控制平面、编译路径、验证端点和伙伴运营入口渠道。[CE002, CE003, CE004, CE005, CE006, CE007]

技术 / 运营架构表
角色关键实现依赖风险
量子比特硬件在俘获离子上执行门H1 和 H2 使用 171Yb+;Helios 使用 137Ba+ QCCD 阱,具备全连接激光和控制栈校准,加上离子传输扩展取决于持续硬件升级和校准纪律
原生操作支持动态程序中途电路测量、条件逻辑、量子比特复用、任意角 ZZ 和可选 SU(4) 纠缠门稳定的编译器和运行时语义高级功能推高集成和测试负担
编译运行时优化并路由工作负载TKET 编译到原生门,并暴露优化级别和扩展模块编译器正确性和扩展维护托管默认值可能让用户看不清实际电路转换
访问控制平面调度作业和仿真Nexus、Azure 目标、OLCF 工作流、语法检查器和仿真器门户正常运行时间、配额、身份和伙伴集成访问窗口和配额系统可能拖慢生产工作
应用和网络安全层封装领域工作流InQuanto、Quantum Origin、lambeq、Guppy,以及与 CUDA-Q 联动的工作流客户赋能和伙伴共创覆盖广,但各模块公开使用深度差异很大

这是一张层级图,不是物料清单。重点是公开产品如何拼成一个运营栈,以及主要依赖落在哪里。

[CE007, CE008, CE011, CE027, CE028, CE029]
FE002: 客户工作流 / 运营流程
[CE012, CE019, CE027, CE028, CE029, CE032]
FE003: 关键依赖图
[CE008, CE027, CE028, CE029, CE030, CE038]

5.3 软件栈、基准与路线图成熟度

Quantinuum 的软件层比简单 SDK 封装更深。TKET 是真实的编译器 / 运行时层,拥有公开 C++ 核心、Python 绑定、扩展包,以及活跃到 2026 年 5 月的发布节奏。公开文档把它描述为构建-编译-运行技术栈,H1 和 H2 数据表也明确称 TKET 会在托管栈中优化提交电路。这里的开发者信号有意义:PyPI 显示 pytket 有跨平台打包,Pepy 显示累计下载量达到数百万,说明 TKET 是组合中最成熟的公开软件面。 基准阶梯也强于许多量子创业公司。H1 公开 20 个量子比特和 1×10^-3 的典型双量子比特不保真度;H2 公开 56 个量子比特、相同典型双量子比特不保真度区间,以及更大的跑道式 QCCD;Helios 公开 98 个量子比特,并把典型双量子比特不保真度提升到 8×10^-4,同时加入实时控制特性。路线图成熟度也不只是未来承诺。Quantinuum 和 Microsoft 公开了 H2 上的逻辑量子比特结果,Quantinuum 与 NVIDIA 现在也把 Helios 描述为 GPU 联动实时解码和混合量子-AI 工作流的起点。不过,路线图应被视为正在成熟,而不是已经完成:TKET 已广泛公开,InQuanto 已产品化但许多收益指标来自公司主张,lambeq 活跃但小众,Helios 则仍是早期企业产品,而非已被广泛证明的公用服务。[CE011, CE012, CE013, CE014, CE015, CE016]

路线图 / 发布 / 开发阶段表
日期 / 阶段功能或里程碑状态含义来源
2025 年当前代H1 数据表 8.0.1 版,20 量子比特线性 QCCD活跃基准第一代平台仍是俘获离子工作流的可用参考系统H1 数据表
2025 年当前代H2 数据表 4.00 版,56 量子比特赛道式 QCCD活跃第二代平台在保留全连接和动态电路功能的同时扩大量子比特H2 数据表和 Azure 提供商
2024 年里程碑,2026 年仍相关Microsoft 在 H2 上的逻辑量子比特演示已完成里程碑支撑纠错运行路线图可信度,但不是广泛商业 SLA 证明PR Newswire 新闻稿
2025 年 11 月发布 / 2026 年 1 月规格Helios 98 量子比特系统,配套 Guppy 和 GPU 集成控制已发布 / 早期客户阶段路线图从 H-Series QPU 转向混合量子-AI 系统架构Helios 页面、数据表和 DCD
2026 年 5 月软件节奏TKET v2.17.0 和 v2.18.0 发布活跃开源维护编译器层随硬件路线图快速发版GitHub 发布记录和 PyPI
2026 年 5 月软件节奏lambeq 0.5.0 发布,新增 DisCoCircReader、OncillaParser 和 PytorchQuantumModel活跃但小众显示核心编译器之外仍在投入研究工具GitHub 发布记录和 Pepy

日期是本章留存的公开里程碑;它们显示路线图仍在推进,但不能替代生产可靠性证据。

[CE014, CE025, CE033, CE034, CE035, CE036]
FE004: 产品成熟度 / 能力地图
[CE003, CE005, CE008, CE014, CE016, CE018]

5.4 差异化、信任姿态与产品边界

Quantinuum 的主要差异化是整合。公司把全连接离子阱硬件、动态电路特性、第一方编译器、化学套件、密码产品和开源研究工具放在一起。相比依赖外部编译器、单一云市场上架项,或只有一个窄应用的供应商,这个全栈叙事更强。Quantum Origin 的 Bell-test 定位、公开语法检查器和模拟器工作流、GitHub 发布节奏、PyPI 来源记录,都增加了信任信号,说明这个栈是在作为软件运营,而不只是物理演示。 边界同样可见。公开证据在技术验证上远强于企业运营。Azure 和 OLCF 文档描述了队列、配额和开发周期,但保留来源没有提供历史可用时间、正式 SLA 或详细信任中心材料。Helios 本地部署已公开提供,但保留来源中设施和支持要求仍不够具体,无法承销。Helios 旗舰“全球最准”的表述也更多由公司和合作伙伴主导,而非在保留材料中得到独立基准验证。净结论:Quantinuum 在全栈离子阱执行上看起来差异化且技术领先,但买方承销的仍是一套托管、稀缺、企业重型的平台,而不是透明的商品化量子云。[CE020, CE021, CE028, CE030, CE038, CE039]

信任 / 质量 / 合规表
控制项或信号状态范围证据缺口
硬件数据表和已发布规格公开H1、H2 和 Helios 功能,以及非保真度区间官方产品数据表留存证据中没有公开 SLA 或正常运行时间历史
语法检查器加仿真器工作流公开硬件执行前预检验证Azure Quantum 和 OLCF 文档最终证明仍需要稀缺硬件时间
开源编译器来源公开TKET 发布节奏和 PyPI 可信发布轨迹GitHub 发布记录和 PyPI 包元数据商业支持保证未完全披露
量子随机性验证公开Quantum Origin 种子质量和部署模型Quantum Origin 页面称通过 Bell 测试,且无需额外硬件或云未留存公开审计包或认证包
云和本地部署说法公开营销和合作伙伴文档Helios 云 / 本地部署,加上通过 Azure 或 OLCF 访问 H2Helios 页面、DCD、Azure、OLCF设施要求和正式信任中心细节披露不足

最强的公开信任面偏技术,而非重合规:数据表、验证端点、发布来源和加密机制描述先于企业保障材料出现。

[CE020, CE021, CE027, CE028, CE029, CE039]

5.5 图表

Chapter 06

06客户情况

6.1 按垂直行业、买方与地域划分客户

Quantinuum 披露的客户基础,更适合理解为一组垂直行业参考账户,而不是已经规模化的席位型装机基础。买方、用户和付款方拆分会随细分市场变化。在金融领域,JPMorganChase 和 HSBC 以创新或安全买方身份出现,研究和网络安全团队是直接用户。在制药和化学领域,Amgen、Chugai、Panasonic 和 JSR 以 R&D 用户身份测试量子增强发现或仿真工作流。在出行和材料领域,BMW 和 Airbus 把 Quantinuum 用在围绕燃料电池和催化剂化学的高级研究项目中。在公共部门,新加坡的 NQO 和 NQCH 以及日本的 RIKEN 充当付款方或基础设施赞助方,研究人员则是直接用户。S-1 又把图景拓宽,点名公司在制药、材料科学、金融服务、政府和工业市场有活跃客户合作。仍然缺的是分母:Quantinuum 没有披露总客户数、云客户与本地部署客户拆分,也没有说明多少 Q-Net 用户会转化为付费账户。[CU001, CU002, CU018, CU019, CU021, CU027]

客户分层表
客群买方 / 用户 / 付款方具名证明当前用例部署成熟度变现缺口
金融与网络安全买方:银行创新或安全负责人;用户:量化研究和网络安全团队;付款方:企业技术预算JPMorganChase;HSBC经认证随机性、金融分析、量子安全代币化黄金试点到研究部署无公开 ACV、席位数或转化数据
制药 / 生命科学买方:R&D 技术或化学负责人;用户:发现科学家;付款方:药企 R&D 预算Amgen;Chugai生物制剂发现、反应路径探索研究合作方 / QIDO beta无公开生产部署或 ROI 指标
移动出行 / 材料买方:先进研究团队;用户:材料科学家;付款方:企业研发部门BMW Group、Airbus、Panasonic、JSR 等行业客户燃料电池催化剂、ORR 模拟、材料研究多年联合开发,加上 Beta 验证商业收入条款未披露
公共部门 / 国家级研究买方 + 付款方:国家量子办公室、研究机构和政府;用户:研究人员和 HPC 运营方Singapore NQO/NQCH;RIKEN;美国政府本土 Helios 访问、量子-HPC 混合、获得资助的量子项目已采购或由公共资金支持的基础设施使用强度和合同期限大多未披露
工业工程 / 软件渠道买方:工程软件或平台合作伙伴;用户:企业工程师Synopsys 与 Microsoft Azure Quantum Elements 软件生态工作流集成与私有预览访问渠道 / 生态扩张终端客户经济性取决于合作伙伴
生态 / 社区买方:多元;用户:研究人员、开发者和初创公司Q-Net 成员;初创公司合作伙伴培训、工具、协作、工作负载发现漏斗顶部采用信号2,500+ 活跃用户并非已披露的付费客户

各行按买方、用户、付款方和垂直领域汇总公开证据。社区和公共部门行反映的是采用信号和基础设施资金,不一定是已签约的经常性收入。

[CU001, CU002, CU018, CU019, CU021, CU024]
FU001: 客户旅程图

Quantinuum 通常把客户从生态发现带入试点或共同开发,再走向选择性部署和硬件路线图扩张。

[CU003, CU018, CU020, CU024, CU025, CU035]

6.2 具名客户证据,以及生产化与试点的分野

在量子计算领域,具名客户证明的质量高于平均水平,因为多起案例包含客户侧表述和具体工作负载,而不只是 logo 页。JPMorganChase 是最干净的金融证明:该银行共同撰写并单独介绍了在 Quantinuum H2 系统上运行的认证随机性工作流;Quantinuum 后来又把 JPMorgan 认定为研究工作流中的 Guppy 采用者。BMW 是最强的工业持续性案例:公开材料显示,双方合作自 2021 年持续,2023 年 BMW、Airbus 与 Quantinuum 跑出氧还原工作流,2026 年又签下多年延期,并明确覆盖 Helios、Sol 和 Apollo。HSBC 通过使用 Quantum Origin QRNG 的代币化黄金试点,给出近期安全场景证明,但公开证据仍把它定义为试点。Amgen 明确是具名生命科学合作方,但公开记录还没给出生产经济性或规模化工作流指标。RIKEN 是最强的公共部门部署证明,因为它在使用 H1 后实际采购了 H2 升级。JSR、Panasonic 和 Chugai 的 QIDO beta 证言在日本补充了有用证明,但也说明许多客户关系仍停留在早期验证,而不是规模化生产。[CU006, CU007, CU008, CU009, CU010, CU011]

具名客户证明表
客户细分领域公开证据用例生产部署 / 试点限制
JPMorganChase金融服务客户侧博客和联合新闻稿认证随机性和高级金融分析研究实时系统上的研究工作流未披露合同金额或规模化部署条款
HSBC银行 / 网络安全Quantinuum 公告和独立报道使用 Quantum Origin QRNG 的量子安全代币化黄金试点公开证据未显示更广推广指标
BMW Group移动出行 / 材料2021 年至今合作,2026 年多年续约燃料电池催化剂和 ORR 材料研究多年联合开发收入经济性仍未披露
Airbus航空航天 / 材料客户侧关于氢和燃料电池相关性的引述航空航天材料和可持续出行研究研发合作未披露公开采购或生产合同
Amgen制药 / 生命科学Helios 发布材料和申报文件表述用于生物制剂发现的量子-机器学习混合方法研究合作方未披露公开生产工作流或 ROI 指标
RIKEN公共部门 / 研究运行 H1 后采购 H2 升级用于化学和生物分子工作流的量子-超算混合已采购 / 已部署系统商业经济性未与研究任务拆分
Singapore NQO / NQCH公共部门 / 国家级枢纽本土 Helios 访问和研发中心生物、金融、材料、优化领域的国家量子应用赞助访问 / 计划部署企业用户转化未披露
JSR / Panasonic / Chugai工业 / 制药 Beta 用户日本发布前的 QIDO Beta 评价化学、材料和药物发现工作流验证Beta / 验证用户明确指出仍需技术工作

各行列出公开具名、且有直接工作流或采购证明的交易对手。覆盖并不完整,因为申报文件披露了集中度和行业广度,却未点名所有贡献收入的客户。

[CU006, CU007, CU009, CU011, CU013, CU015]
FU003: 客户证明矩阵

如果来源提供具名用户、具体工作负载和重复合作,而不只是 logo,证明质量最强。

[CU011, CU015, CU020, CU022, CU023, CU026]

6.3 采用轨迹、访问模式与社区广度

公开采用路径有清晰顺序。Quantinuum 先拿下研究或创新合作关系,让对方通过云访问、beta 项目或国家枢纽基础设施上手,之后才尝试转化为更深部署或重复采购。2023 年 BMW 和 Airbus 的材料工作流,以及 2024 年 Microsoft 逻辑量子位里程碑,先建立技术可信度。2024 和 2025 年,HSBC 的代币化黄金试点、QIDO beta 测试、JPMorgan 的认证随机性合作,显示终端用户已在具体工作负载上投入。随后 Helios 在 2025 年末把叙事往前推了一步:Quantinuum 公开点名 Amgen、BMW、JPMorganChase 和 SoftBank 为早期用户或合作方,并提到 SoftBank 和 JPMorgan 在全面开放前参加了两个月早期访问项目。漏斗顶部比具名客户名单更宽:Quantinuum 称 Q-Net 有 2,500+ 名活跃 Nexus 用户,Q-Net Connect 2026 吸引 170+ 名参会者。这些指标衡量生态活跃度,不是付费客户数或使用量,因此应视为采用信号的广度,而不是收入深度。[CU003, CU004, CU005, CU025, CU026, CU034]

客户增长 / 采用轨迹表
日期 / 阶段具名交易对手公开证据生产部署 / 试点含义缺失分母
2023-08BMW Group + AirbusH-Series 上的工业 ORR 模拟工作流商业化前研发工作流Helios 发布前已有具体材料科学用例未披露合同金额或部署规模
2024-04Microsoft可靠逻辑量子比特,以及后续 Azure Quantum Elements 集成合作伙伴 / 渠道里程碑提升生态触达和化学工作流访问未表述为直接客户销售
2024-09HSBC用于代币化黄金的 Quantum Origin 试点试点显示近期网络安全变现路径未披露更广部署量
2025-03JPMorganChase + 国家实验室H2 上的认证随机性实时系统上的研究工作流金融服务领域强客户侧证明未披露合同金额
2025-08通过 QIDO 的 JSR / Panasonic / Chugai正式发布前的 Beta 测试Beta / 验证显示工业和制药用户愿意测试化学工具无付费推广指标
2025-11Amgen、BMW、JPMorganChase、SoftBank + 新加坡Helios 发布、早期用户和本土合作早期访问 / 发布证明扩大垂直领域和地域覆盖无总客户数分母
2026-04 to 2026-05RIKEN 和 BMWH1 到 H2 升级,以及 BMW 多年续约重复采购 / 续约所审阅来源中最有持续性的公开证据仍无 NRR、GRR 或头部客户身份

本表跟踪公开可见的客户采用里程碑,而不是已确认收入。多行由合作伙伴或试点驱动,不应误读为规模化生产部署。

[CU005, CU007, CU011, CU015, CU018, CU020]
FU002: 采用 / 部署漏斗

公开可见的采用从广泛社区活动急剧收窄到少数具名且持续的部署。

前两个阶段是公司报告的社区指标,不是付费客户数。后续阶段是从留存公开证明中得出的保守计数,因此漏斗展示的是证据质量收窄,而不是真正的商业转化数据集。

[CU004, CU005, CU007, CU020, CU025, CU026]

6.4 持续性代理指标与扩张路径

Quantinuum 不披露 NRR、GRR、流失率或合同期限,持续性只能从重复行为推断。BMW 合作是最清晰的重复参与代理指标,因为合作始于 2021 年,并续签到覆盖多代硬件的多年路线图。RIKEN 从 H1 升级到 H2 更强,因为这是可见的采购升级,不是新 logo。JPMorgan 也显得稳定:它在 S-1 中是客户和创新伙伴,出现在 Helios 早期访问中,后来又获得 Q-Net 采用奖。Honeywell 是另一种有用代理指标,文件称其既是测试场,也是早期客户,并打算在 IPO 后继续作为战略客户。扩张路径可见,但由合作伙伴中介。Microsoft 通过 Azure Quantum Elements 私有预览扩大触达,Mitsui 在日本分销 QIDO,Synopsys 把 Quantinuum 嵌入工业工程工作流,新加坡打造本地枢纽模式。这些路径能扩大采用,但也说明 Quantinuum 仍依赖联合销售、生态项目和硬件路线图执行,而不是简单的自助式软件增长路径。[CU007, CU008, CU013, CU020, CU024, CU025]

留存 / 重复使用 / 满意度表
指标 / 代理指标公开数值细分领域置信度尽调问题
总客户数所有客户按产品、地区,以及公共部门 / 企业拆分披露付费账户
NRR / GRR / 流失所有收入客户提供按队列或产品线拆分的续约、流失和扩张指标
BMW 合作时长代理指标自 2021 年起合作;2026 年多年续约移动出行 / 材料要求提供年度支出、里程碑门槛和续约经济性
RIKEN 重复采购使用 H1 后于 2026 年采购 H2公共部门 / 研究澄清合同期限、使用量和升级经济性
JPMorgan 重复合作早期访问、2025 年论文和 2026 年 Guppy 奖金融展示从研究合作到产品部署的付费扩张
Honeywell 战略连续性早期客户 / 测试场;产品推出后的战略客户工业 / 化学将关联方需求与独立第三方收入拆分
社区活跃度2,500+ 名活跃 Nexus 用户和 170+ 名 Q-Net Connect 参会者生态展示社区用户向签约收入的转化

null 表示所审阅公开来源未披露该 KPI。非 null 行是耐久性代理指标,不是正式 SaaS 留存指标。

[CU006, CU007, CU020, CU027, CU028, CU033]
合作伙伴 / 渠道依赖与采购摩擦表
上市路径 / 依赖项支撑内容依赖风险客户含义尽调问题
企业直销通过联系人驱动销售提供 Helios 云端和本地访问重服务导入和定制合同采购较慢,且未见公开自助购买动作询问平均销售周期和最低 ACV
Q-Net / 生态项目社区、培训和初创公司合作伙伴计划社区规模未必转化为付费扩张对漏斗顶部教育有用,但变现能见度有限要求提供 Q-Net 和初创公司合作伙伴转化为合同的情况
Microsoft / Azure Quantum Elements私有预览化学和逻辑量子比特生态Quantinuum 依赖合作伙伴路线图,以及预览版到 GA 的转化可扩大对企业和 HPC 买方的触达询问线索或收入中经 Microsoft 而来的占比
Mitsui / QIDO 分销日本独家分销,加上与工业用户共创 Beta分销商经济性和客户归属未披露可加快化学和制药领域的本地访问要求提供分销商利润率,以及直销 / 渠道管线拆分
国家级枢纽赞助新加坡和 RIKEN 形成基础设施驱动的采用预算周期和政治优先级会重塑需求大型标杆账户可能伴随较长采购周期要求提供已签署条款、续约选项和本地用户采用情况

本表刻画上市路径结构,而不是客户标识。几条扩张路径由合作伙伴中介,因此在扩大触达的同时也增加执行依赖。

[CU003, CU019, CU024, CU025, CU035, CU036]

6.5 集中度风险、采购摩擦与明确缺口

负面客户信号是集中度。Quantinuum 的 S-1 披露,一个客户贡献了 2025 年收入的 60%、2024 年收入的 63%;2026 年 3 月季度,一个客户贡献 47% 收入,而 2025 年 3 月季度的 90% 收入来自另一个客户。同一文件还显示,美国政府贡献了 2025 年收入的 16%、2024 年收入的 9%、2026 年第一季度收入的 24%。这不是广泛企业客户基础,而是少数大客户模式,并带有显著主权敞口。公司销售方式也暴露出采购摩擦。Helios 访问需要通过云端或本地销售联系人,Microsoft 商业化仍处在私有预览,QIDO 在全面发布前经过 beta 测试,许多具名客户证明是共同开发或试点项目,而非按价目表部署。因此,公司商业前景可信,但运营上仍早期:垂直行业和地域证明可靠,公开材料却仍缺客户数量、最大客户身份、定价、ACV 和留存统计。承销客户故事,必须同时接受集中度风险和企业或公共部门漫长转化周期。[CU027, CU028, CU029, CU030, CU031, CU032]

扩张与集中度风险表
扩张驱动因素集中度 / 摩擦影响证据尽调路径
硬件路线图增购BMW 明确从 Helios 扩展到 Sol 和 Apollo 世代灯塔客户内扩张潜力高BMW 多年协议要求按硬件世代提供账户级支出承诺
公共部门重复采购RIKEN 从 H1 迁移到 H2,新加坡正在建设本土 Helios 访问公共部门买方能成为大型部署的锚点RIKEN 和新加坡公告澄清未来收入中可能来自主权买方的比例
收入集中度单一客户贡献 2025 年收入的 60% 和 2024 年收入的 63%单一账户流失会重创业绩S-1 客户集中度脚注要求提供前五大客户结构和趋势方向
季度波动另一客户占 Q1 2025 收入 90%,另一个客户占 Q1 2026 收入 47%几个账户就能让业绩大幅波动S-1 季度集中度披露获取逐季度集中度桥表
政府依赖美国政府占 Q1 2026 收入的 24%、2025 年收入的 16%公共采购节奏会显著拉动收入波动S-1 披露拆分拨款、合同和商业企业收入
试点到生产转化HSBC 试点和 QIDO Beta 显示早期牵引,但尚非规模化推广转化周期偏长的风险HSBC 和 QIDO 公告要求按垂直领域提供管线阶段数量和转化率
定价 / 身份不透明头部客户身份、合同期限和价目表未披露难以支撑持久性和 ACV 假设申报文件和公司材料均缺少披露在 NDA 下获取主服务条款、合同期限和集中账户名称

扩张机会可见,但集中度和披露缺口主导投资判断风险,因为公开记录缺少判断客户广度或留存所需的分母。

[CU007, CU008, CU015, CU018, CU020, CU028]
Chapter 07

07风险

7.1 风险概览

Quantinuum 的风险栈由三项投资逻辑级暴露牵头:Honeywell 控制下的所有权与未解决的流动性路径、仍脆弱的量子市场中的商业化时点风险,以及把 Helios 时代里程碑转化为可重复容错效用的技术挑战。公司资金充足、技术可信,但公开材料仍要求投资者承销未来里程碑——Apollo 规模的逻辑量子位进展、更广泛生产部署、成功 IPO 流程——而不是收入持续性或独立治理的闭环。风险热力图把所有权 / 流动性、商业化延迟、规模化执行放在最高严重度单元,因为即便科学继续进展,每一项都能快速重定价这家公司。出口管制、外资审查、供应商集中、关键人物依赖,是把一条有前景的路线图变成更慢、更贵、更缺流动性公司的主要传导路径。[CR003, CR004, CR011, CR024, CR036, CR040]

按严重程度排序的风险摘要
排名风险可能性影响缓解成熟度剩余敞口投资含义
1Honeywell 控制权和未解决的 IPO / 流动性条款在 S-1 经济条款和控制权条款清晰前,不要把退出时点或少数股东保护纳入投资假设。
2商业化延迟 / 板块情绪重估低-中在生产部署和可复制收入证明扩展到旗舰客户之外前,用里程碑折价估值。
3Helios 到 Apollo 之间的扩展和纠错未达预期中-高严重把路线图视作执行风险,而不只是科学上行;将资本投放与里程碑达成绑定。
4出口管制和国家安全限制若 BIS 或 CFIUS 摩擦扩大,建模时应假设跨境招聘、合作和投资人灵活性都会放缓。
5对 NVIDIA / Microsoft / GF / Monarch 的生态和供应商依赖中-高若任一合作伙伴或供应商不可用或延迟,假设进度延误、成本上行。
6来自 Google、IBM、IonQ 和 QuEra 的竞争Quantinuum 必须持续拉大性能领先;里程碑速度走平会引发估值倍数压缩。
7客户集中度和试点到生产转化低-中公开客户标识证明是真实的,但经济性披露不足;在为广泛采用付费前,必须拿到客户经济性。
8关键人和治理集中度中-高中-高Hazra、Khan 或法务 / 政府关系负责人流失,会同时削弱外部可信度和执行力。

严重程度排序是截至 2026-05-26 的公开记录综合;它更重视可能性和影响,而不是标题式乐观,并假设无法访问保密 S-1 或客户收入文件。

[CR003, CR011, CR024, CR029, CR030, CR036]
FR001: 风险热力图
[CR003, CR011, CR024, CR036, CR040, CR045]

7.2 所有权、流动性与监管风险

Honeywell 仍位于 Quantinuum 治理和流动性故事的中心。公开申报公告显示,IPO 流程已推进到公开 S-1,但公司仍未披露股数、价格区间或最终时间,Honeywell 也称发行取决于市场条件。这意味着,在公开市场正检验量子计算是否配得上后期高倍数溢价的时点,公司退出路径、自由流通深度和少数股东保护仍是开放问题。与此同时,出口管制风险已从抽象政策噪音变成现实运营约束。BIS 如今在全球范围管制量子计算机、配套软件和技术;视同出口申报和未来可能的许可要求,会给量子公司依赖的国际人才池增加摩擦。法律分析还指向 CFIUS 敏感度和合规复杂度上升,这意味着所有权、招聘、跨境协作可能一起收紧,而不是彼此独立。[CR002, CR003, CR004, CR011, CR013, CR015]

监管 / 法律风险登记表
规则 / 事项管辖区当前状态发生概率严重度缓释措施剩余敞口尽调路径
IPO 生效、流通盘与控制权条款美国 / Nasdaq公开 S-1 已提交;价格区间和发行股数仍未披露Honeywell 背书和主要承销商提供流程支持审阅 S-1 草案、锁定期、投票控制权,以及任何分阶段 Honeywell 减持计划。
BIS 对量子物项的出口管制美国 / 全球出口自 2024 年 9 月起生效,覆盖硬件、软件、技术和组件专设法务 / 政府事务负责人,并为盟友国家保留 IEC 路径梳理产品 ECCN、目的地结构,以及已需或待批的任何许可证。
视同出口 / 外籍人员报告美国员工与实验室报告义务已生效;未来许可要求仍在评估中高内部跟踪和律师支持可降低流程失误中高索取受控岗位的公民身份结构,以及截至目前任何已触发报告的披露。
CFIUS 关键技术审查扩围美国所有权 / 融资法律分析提示,受覆盖技术的申报义务更广中高公开未披露不利申报审阅融资计划、外国投资者权利,以及与 IPO 或私募轮相关的任何 CFIUS 备忘录。
数据隐私、IP 与治理合规负担美国 / 欧盟 / 全球公开材料将这些义务归口到一个高管职能责任集中有助于加快决策索取隐私事件、专利纠纷、赔偿责任和监管往来材料。
诉讼及潜在索赔透明度缺口全球已审阅的公开来源无法提供可靠纠纷清单公开材料中未见中高索取外部律师摘要、保险清单和董事会诉讼更新。

各行按严重度排序,仅基于公开证据构建;诉讼行属于披露风险项, 而非已确认纠纷。

[CR003, CR011, CR012, CR013, CR015, CR016]

7.3 技术与运营规模化风险

Quantinuum 确有技术动能,但 Helios 与具商业决定性的 Apollo 之间的距离,仍是核心运营风险。公开发布和 Helios 论文显示出强劲的离子阱性能、逻辑量子位进展,以及迈向更大系统的可信路径。但 Quantinuum 和独立报道仍把当前一代定义为重要证明点,而不是完成的实用规模产品。路线图依赖逻辑量子位数量、错误抑制、实时解码和硬件可制造性的激进改进。这些里程碑并不孤立:Quantinuum 自身材料把未来进展绑定到 NVIDIA 支持的经典协处理,以及包括 GlobalFoundries 和 Monarch Quantum 在内的狭窄硬件供应链。只要其中任一环节滑坡——纠错、合作伙伴软件、光子学、组件良率或系统集成——公司仍可保住科学可信度,却错过估值和 IPO 叙事暗含的商业时间表。[CR019, CR020, CR022, CR023, CR024, CR029]

运营 / 质量 / 安全风险清单
失效模式证据发生概率严重度缓释成熟度剩余敞口未解缺口
从 Helios 到 Sol / Apollo 的路线图延误Helios 已上线,但 Apollo 级容错仍是前瞻目标中高极高需要内部里程碑燃尽表、良率数据,以及系统级 go/no-go 标准。
实时纠错栈未按计划扩展Quantinuum 现在把未来解码性能绑定到 NVIDIA GPU-经典计算集成上中高需要延迟基准、备用解码器,以及不依赖单一软件栈的方案。
硬件与光子组件单一来源瓶颈CHIPS LOI 将关键组件供应商圈定在少数几家低中中高需要备选供应商、库存缓冲,以及各组件认证状态。
本地部署和主权部署增加支持负担RIKEN 以及未来本国部署,把运营复杂度推高到纯云访问之外需要现场服务人员配置模型、正常运行时间义务,以及保修 / 服务经济性。
现有系统尚不能支撑投资人期待的广泛变现算法独立报道称,Helios 仍不足以跑通市场期待的商业算法低中需要超出试点或研究叙事的用例转化证明。

本清单聚焦运营扩张和质量风险,而非泛泛的实验室安全;公开证据最强的部分是路线图执行、 解码器集成和组件可得性。

[CR019, CR020, CR022, CR024, CR029, CR030]
FR002: 风险传导图
[CR003, CR011, CR015, CR020, CR024, CR036]

7.4 依赖、竞争与商业化风险

Quantinuum 的公开客户证明可信但偏窄。Helios 营销和合作伙伴公告反复回到同一批旗舰账户——BMW、JPMorganChase、Amgen、SoftBank、Microsoft、NVIDIA 和 RIKEN——却没有公开收入集中度、续约或试点转生产数据。集中度风险重要,因为更广泛市场仍在判断哪些量子用例能承载持久预算。SoftBank 自己的合作公告异常坦率:收入模式、定价逻辑、服务时点和具商业意义的用例仍在搭建中。与此同时,竞争在多种技术路线之间加剧,而不是收敛到单一赢家。Google 展示低于阈值的超导纠错,IBM 扩大机群可用性和数据中心架构,IonQ 向生产环境推销离子阱系统,QuEra 则主张中性原子不靠低温也能更干净地扩展。因此,公开市场审视会从两侧打击 Quantinuum:采用弱于预期会挑战整个板块;竞争对手里程碑更强,则会压窄投资者愿为 Quantinuum 领先地位支付的溢价。[CR009, CR025, CR026, CR027, CR028, CR032]

合作伙伴 / 依赖风险清单
依赖交易对方 / 节点角色集中度信号失效情景严重度缓释措施剩余敞口
治理与流动性支持方Honeywell控股股东、董事会主席、IPO 支持方一个母公司掌握控制权和退出时点IPO 条款或减持安排未给少数股东带来实质流动性母公司深度支持和承销团
混合计算与 QEC 栈NVIDIAGPU 加速、CUDA-Q、NVQLink、研究中心合作路线图材料反复依赖 NVIDIA 工具链解码器或集成延误会拖慢路线图,也削弱 GenQAI 叙事早期集成工作和多年合作中高
逻辑量子比特可信度与云分发Microsoft逻辑量子比特验证、Azure 访问、化学工作流关键路线图可信度主张由 Microsoft 共同背书共同里程碑停滞,或 Azure 访问优先级下降中高既有合作很深,且已有多个里程碑证明
关键硬件供给GlobalFoundries 和 Monarch Quantum半导体组件与集成光子CHIPS LOI 将其列为关键扩产供应商良率、认证或产能问题拖慢未来系统本土供应链支持与 CHIPS 对齐中高
旗舰客户证明集BMW、Amgen、JPMorgan、SoftBank、RIKEN 等具名客户标杆需求与验证公开证明集偏窄,反复使用同一批标识一两个旗舰项目未能转化为持续支出多年关系和高知名度合作伙伴
战略性政府支持Commerce / NEDO / 国家项目资金、部署可信度、生态入口政府相关项目在公开增长叙事中占比很高政策变化或主权要求压缩部署选择中高多个项目之间的地域多元化中高

当 Quantinuum 自己把合作伙伴或供应商列为路线图可信度、硬件扩产或旗舰需求证明的关键时, 公开依赖信号最强。

[CR002, CR004, CR029, CR030, CR032, CR033]
FR003: 依赖关系图
[CR004, CR015, CR029, CR030, CR032, CR034]

7.5 人员风险、缓释因素与投资逻辑失效触发点

自 2023 年以来,领导层厚度有所改善,但关键人物暴露仍然重要,因为 Quantinuum 的公开叙事依赖少数几个人,每个人都掌握一个关键接口。Raj Hazra 是规模化和公开市场门面,Ilyas Khan 仍是与产品愿景绑定的创始运营者,法务 / 政府关系职能则把治理、出口管制、隐私和 IP 汇聚到一名高管之下。公司的主要缓释因素同样清晰:Honeywell 支持、大型技术团队、锚定合作伙伴、政府参与,以及拥有运营经验的扩展高管班子。这些缓释因素足以支持持续关注,但不支持自满。投资者应监测 S-1 修订、Sol / Apollo 里程碑节奏、出口管制变化、生产客户转化和高管团队连续性。仍需补齐的尽调包很直接:股权结构机制、客户经济性、出口许可历史和供应商集中度。在这些文件完成审阅前,剩余暴露仍足够高,应把 Quantinuum 视为里程碑驱动投资,而不是已去风险的基础设施平台。[CR042, CR043, CR044, CR045, CR046, CR047]

人员 / 执行风险清单
角色 / 职能依赖或缺口发生概率严重度缓释措施剩余敞口尽调路径
Raj Hazra / CEO主要扩张操盘手和公开市场叙事者更宽的管理梯队和运营经验中高审阅继任计划、留任方案和授权决策权。
Ilyas Khan / 创始人与副董事长产品愿景、历史关系和创始人信誉仍较集中已完成过一次角色交接,且创始人未退出中高索取创始人留任条款,以及与 CEO 之间的产品治理分工。
首席法务 / 政府事务负责人治理、出口管制、隐私、IP 和合规集中在一个角色单一责任人能加快协调中高审阅二线合规人员配置和外部律师支持深度。
Honeywell 背景偏重的高管梯队运营纵深更宽,但多名高管有母公司履历团队厚度强于 2023 年评估独立性激励、董事会构成,以及 Honeywell 渠道之外的权限。

关键人风险不只是创始人风险;监管与合作伙伴管理能力也高度集中, 短期很难替代。

[CR016, CR042, CR043, CR044, CR045]
缓释措施与投资逻辑破裂触发点
风险可监测触发项阈值 / 事件行动含义
Honeywell 控制权与流动性悬顶压力S-1 修订和上市后治理流通盘小、超级投票权控制,或 Honeywell 减持路径不清按由赞助方控制的公司处理,在控制条款正常化前给予流动性折价。
路线图延误Sol / Apollo 里程碑节奏系统延误超过 12 个月,或逻辑量子比特目标反复下修按更慢商业化曲线和更低战略稀缺性溢价重新测算估值。
商业化延后生产部署和可重复付费用例到 2027 年仍未从同一批旗舰账户向外扩展假设业务仍由研究驱动,并大幅下调收入预期。
监管收紧BIS / CFIUS / 盟友管制更新新增许可要求限制招聘、海外合作或投资者参与提高监管折价率,并重新评估全球商业化落地假设。
供应商集中组件良率和备选来源认证任何单一来源延误推迟重大系统里程碑从时间表风险升级为执行红旗,在应急计划得到验证前推迟新资本。
关键人流失高管连续性Hazra、Khan 或法务 / 政府关系负责人意外离职且没有明确接任者升级到投资逻辑破裂复核,因为路线图、治理和市场可信度都会 立即受挫。

触发点设计成可从申报文件、合作伙伴公告、里程碑发布和尽调材料中监测, 而不是依赖主观叙事变化。

[CR003, CR029, CR036, CR041, CR045, CR048]

7.6 展项

Chapter 08

08估值

8.1 投资逻辑与反向逻辑

相比大多数已上市量子同行,Quantinuum 的定性故事更强。公司组建了横跨 JPMorgan Chase、Amgen、Mitsui、Quanta Computer、NVentures、QED Investors 和 Honeywell 的财团,这既是资本支持,也是覆盖金融、医药、工业和 AI 基础设施的分销信号。产品证明也比纯硬件叙事更宽:Helios 于 2025 年 11 月商业发布,具名客户包括 Amgen、BMW Group、JPMorganChase 和 SoftBank;Quantum Origin 给公司带来软件安全切入口;NVIDIA 合作以及 BMW、bp 伙伴关系显示路线图正在接入真实企业项目。反向逻辑是,公开证据仍未披露收入、利润率或客户集中度,所以公司要求投资者为里程碑质量和生态可信度付费,而不是为已证明的经济性付费。在量子板块情绪亢奋时这可能成立,但如果 S-1 显示收入有限,或公开量子倍数压缩,叙事会很脆弱。[CV002, CV005, CV011, CV012, CV016, CV017]

投资建议摘要表
维度取值依据
投资建议观察等 S-1 经济性披露,或入场价明显更便宜后再测算上行空间
置信度判断方向清楚,但缺失财务数据和条款清单压低确信度
风险评级板块倍数波动大、披露不完整,Honeywell 悬顶未解决
估值立场按公开可比公司看合理;按基本面看偏高$10B 标记落在火热的公开量子集群内,但缺少已披露收入或利润率支撑
目标回报 / 持有需要 4-6 年总回报 >2.0x;已知条款下无法支撑没有 S-1 经济性,当前估值标记还看不出足够可测算的上行空间

截至 2026-05-26 的决策摘要。回报目标是纪律门槛,不是有来源承诺。

[CV033, CV034, CV041, CV045]
投资逻辑 / 反向逻辑表
维度投资逻辑反向逻辑
财团质量战略支持方横跨金融、制药、电子、AI 加速器和工业战略背书不会自动转化为经常性收入或少数股东保护
商业证明Helios 将 Amgen、BMW、JPMorganChase 和 SoftBank 列为启动客户;bp 和 BMW 扩大合作具名用户仍可能只是试点或 R&D 项目,而非持久高毛利收入流
产品广度Quantum Origin 和纠错里程碑说明故事不止单一硬件广度也可能在核心计算业务证明独立经济性前分散焦点
相对定价$10B 标记低于 IonQ,接近 D-Wave,仍处在火热的公开量子板块内公开市场本身可能就是问题:可比公司更像叙事驱动,而非现金流驱动
退出路径S-1 已提交,流动性路径可见度提高价格区间、减持画像或股权结构细节尚未公开
投资测算依据今天观察;若 S-1 验证经济性,可转为买入现在买入依赖的是对里程碑的信任,而非已披露基本面

投资逻辑和反向逻辑放在同一决策节点上配对,确保投资案例同时对价格和证据敏感。

[CV011, CV016, CV017, CV033, CV034, CV036]
FV001: 建议逻辑

从战略验证和板块情绪,到当前跟踪建议的决策链。

[CV036, CV037, CV038, CV039, CV045]

8.2 融资背景与入场纪律

融资路径推进很快:2024 年 1 月以 $5 billion 投前估值完成 $300 million 融资,随后在 2025 年 9 月以约 $10 billion 投前估值完成约 $600 million 融资。IPO 路径也已启动:Quantinuum 于 2026 年 2 月秘密提交 S-1,并在 2026 年 5 月公开递交。这些事实重要,但缺失事实对入场纪律更关键。已审阅公开材料没有披露最终 IPO 价格区间、股份数量、Honeywell 减持规模,或 2025 年融资背后的优先股堆叠。用户提供的说明称 2025 年融资后来扩大到 $800 million,方向上可信,但留存公开证据没有像最初 $600 million 公告那样清楚验证这一点。这让投资者只有一个头条估值,却很难看清普通股持有人实际会拥有什么,或该估值背后有多少稀释和抛压。至少,这支持等到 S-1 把经济条款讲清楚后再行动。[CV001, CV003, CV004, CV007, CV008, CV009]

乐观 / 基准 / 悲观情景表
情景概率信号关键假设估值区间$10B 入场隐含回报
乐观25%S-1 披露有意义的收入质量,Helios 将标杆用户转化,量子 IPO 窗口仍然打开16-22B1.6x-2.2x
基准50%S-1 显示真实进展但规模可见度有限;条款不惩罚投资者, 但也没有明确便宜证据8-12B0.8x-1.2x
悲观25%收入规模小、条款对投资者不友好,或公开量子倍数下修3-6B0.3x-0.6x
决策解读按概率加权后,当前结果仍达不到严格的后期私募回报门槛低于目标观察 / 等待

区间是情景估计,锚定当前公开可比公司市场情绪、IPO 准备事实, 以及 Quantinuum 财务未披露。

[CV007, CV009, CV034, CV041, CV042, CV043]
FV002: 估值敏感性

如果最重要的上行或下行投资测算假设发生变化,对名义 $10B 中点的方向性影响。

条形表示相对 $10B 中点的大致方向性变化,单位为十亿美元;这些变化不可相加,应按单因素敏感性解读。

[CV035, CV038, CV039, CV041, CV046]

8.3 可比公司组与情景区间

上市量子可比公司在这里更适合作为情绪标记,而不是干净的基本面锚。IonQ、D-Wave、Rigetti 和 Quantum Computing Inc. 在收入基数很小且持续净亏损的情况下,仍有数十亿美元市值,说明 Quantinuum 的 $10 billion 私募估值处在一个仍愿为可选性激进定价的公开市场中。在这个基础上,相比上市集群,Quantinuum 并不明显贵;它低于 IonQ、接近 D-Wave,同时拥有更强股东支持和更丰富客户证明。但同一组可比公司也正是承销危险所在:公开量子倍数太松,无法告诉投资者 Quantinuum 是否便宜,只能说明市场当前能容忍昂贵故事。使用可比表的正确方式,是框定估值区间。如果 S-1 披露验证商业规模且 IPO 窗口保持开放,上行存在;如果文件显示收入单薄或条款对投资者不友好,相比上一轮私募估值会有显著下行。[CV020, CV021, CV022, CV023, CV024, CV025]

可比估值表
可比公司状态估值或市值收入依据引用倍数或读数可比性说明
Quantinuum最近一轮私募新股融资$10B 投前 / 约 $10.6B 投后隐含公开收入未披露n/a该组中战略财团和客户证明最强,但基本面仍未公开
IonQ公开上市$23.75B 市值TTM 收入 $187.12M~126.9x 市值 / TTM 收入最接近的高知名度纯量子可比标的,但基本面估值仍极其昂贵
D-Wave上市公司市值 $10.88BTTM 收入 $12.44M~874.6x 市值 / TTM 收入披露收入小得多,市值却接近 Quantinuum
Rigetti上市公司市值 $8.78BTTM 收入 $10.02M~876.2x 市值 / TTM 收入硬件可比标的的警示样本,规模弱得多
Quantum Computing Inc.上市公司市值 $2.78BTTM 收入 $4.33M~642.0x 市值 / TTM 收入说明市场火热时,即便量子故事收入极小,也能撑住高报价

这是样本而非完整同业集;市值倍数为近似值,并与 Stock Analysis 在 2026-05-26 的 TTM 收入口径配对。

[CV020, CV021, CV022, CV023, CV024, CV025]
投资逻辑破裂与否决触发因素表
触发因素阈值或信号重要性操作含义
S-1 经济性偏弱收入或毛利率披露显示,对 $10B 标价支撑不足会印证该轮定价更像按里程碑给价,而非由基本面锚定不买入 / 转入悲观情景
对投资者不友好的 2025 年条款尽调或监管文件显示惩罚性优先权、老股占比高,或存在估值调整条款账面估值可能高估普通股价值要求更大入场折价,或放弃
Honeywell 悬压招股说明书显示大额减持、锁定安排宽松,或保留控制权过重会挤压流通盘、治理和上市后交易技术面回避 IPO,或缩小仓位
上市量子股估值下修上市同业市值大幅缩水,且没有基本面改善抵消Quantinuum 的相对定价防线会快速变弱下调情景表

这些是可监测的估值触发因素,不只是经营里程碑;每一项都会改变普通股买方实际拥有的东西,或未来可退出的市场。

[CV035, CV039, CV041, CV044, CV045]
FV003: 估值 / 回报区间

在不同披露和市场情绪结果下,围绕最近一轮私募估值的牛市、基准和熊市估值区间。

区间是情景估算,不是来自市场报价;Quantinuum 仍未披露足够财务细节,无法建立正式自下而上的模型。

[CV042, CV043, CV044]

8.4 建议、尽调请求与投资逻辑失效触发点

建议是观察,中等信心、高风险评级,估值立场为合理到偏高。相对意义上用“合理”没错,因为公开量子可比公司更贵;基本面意义上用“偏高”也没错,因为 Quantinuum 仍未披露能让投资者判断 $10 billion 是便宜还是陷阱的运营数据。换句话说,公司质量可能好于公开同行,但价格纪律问题仍未解决。因此,本章不支持只因为板块火热就买入已知私募估值。从观察转向买入的关口很直接:S-1 必须披露真实收入和收入质量,2025 年融资不能隐藏惩罚性优先条款,Honeywell 减持和锁定结构必须可投,Amgen、BMW、JPMorganChase、SoftBank 和 bp 等头部用户必须对应可重复收入,而不能只是试点观感。任一测试失败,投资逻辑都会很快转向悲观情景。[CV034, CV035, CV039, CV041, CV045, CV046]

最终尽调问题表
优先级待索取资料重要性
P1完整 S-1 财务报表:收入、毛利率、烧钱速度、现金、集中度缺少核心经济性,估值工作主要只能相对情绪定价,而不是扎在基本面上
P12025 年融资条款清单和股权结构表需要用来厘清清算优先权、估值调整、稀释,以及账面标价是否高估普通股价值
P1Honeywell 持股、老股交易和锁定安排需要评估 IPO 前后流通盘、控制权和未来供给悬压
P2Helios、Quantum Origin 和主要合作关系带来的客户收入桥需要区分试点与战略声量,还是可重复的收入转化
P2证明 2025 轮融资是止于 ~$600M,还是扩至总计 ~$800M 的文件后钱估值、现金跑道和稀释测算都需要它

这份尽调清单刻意保持短且能影响估值;没有一项只是形式问题。

[CV007, CV034, CV035, CV041, CV046]
FV004: 投资 KPI

面向投委会的快照,概括当前最影响可投性的变量。

这些 KPI 是决策辅助,不是经审计的公司指标;它们概括截至本次运行日期的公开证据状态。

[CV004, CV009, CV011, CV033, CV034, CV045]

8.5 展项

免责声明

本报告是基于公开证据的尽调快照,不构成投资建议。重要的财务、法律、技术和合同事实仍未公开;作出任何投资决策前,应直接向管理层和一手文件核验。

证据索引

结论
编号陈述可信度来源
CO001 Quantinuum was formed on November 30, 2021 when Honeywell Quantum Solutions and Cambridge Quantum completed their business combination. SO008, SO002
CO002 At formation Honeywell was described as Quantinuum's largest shareholder with an approximately 54% ownership stake. SO008
CO003 By the 2026 IPO process Quantinuum was still described externally as majority owned by Honeywell. SO024, SO025
CO004 Quantinuum is operating as a standalone subsidiary with its own capital raises and IPO registration while Honeywell remains the controlling shareholder. SO022, SO024, SO030
CO005 The 2026 S-1 lists Quantinuum's principal executive offices at 303 S Technology Court, Broomfield, Colorado 80021. SO030
CO006 Quantinuum's public materials describe Broomfield, Colorado as headquarters and list facilities across the United States, United Kingdom, Germany, Japan, and Singapore. SO002, SO012, SO022
CO007 Quantinuum describes itself as a full-stack quantum computing platform combining hardware, software, developer tools, and application-specific intellectual property. SO001, SO002, SO030
CO008 Quantinuum commercializes its systems through direct subscriptions, cloud access, Azure distribution, and on-premise offerings. SO003, SO013, SO028
CO009 Quantinuum's H-Series systems are trapped-ion QCCD machines that emphasize all-to-all connectivity and mid-circuit measurement. SO003, SO004, SO028
CO010 Quantum Origin is a software quantum random number generator that Quantinuum says can be deployed without extra hardware or a cloud connection. SO005, SO018
CO011 Quantum Origin became the first software QRNG to receive NIST validation in April 2025. SO018
CO012 InQuanto is Quantinuum's computational chemistry platform for molecular and materials simulation workflows. SO006, SO019
CO013 Quantinuum's developer-tool portfolio includes TKET, Guppy, and lambeq. SO007, SO015, SO019
CO014 Rajeeb Hazra became Quantinuum's CEO in February 2023 after senior roles at Micron and Intel. SO011, SO002
CO015 Ilyas Khan founded Cambridge Quantum in 2014, was the founding CEO of Quantinuum, and in 2026 is presented as vice chairman and head of special projects. SO002, SO011
CO016 Nitesh Sharan joined Quantinuum as CFO effective April 6, 2026 after finance leadership roles at SoundHound AI, Nike, and Hewlett-Packard. SO012, SO002
CO017 Quantinuum's 2026 About page lists Vimal Kapur as chairman, Anne T. Madden as vice chair, and Niels Nielsen, Greg Lewis, George Sherman, and Manish Bhatia as board members. SO002
CO018 Honeywell's influence extends beyond equity ownership because Honeywell is described as a supplier and customer, and Honeywell CEO Vimal Kapur chairs the board. SO008, SO002, SO024
CO019 Key-person dependence is concentrated around Hazra for commercialization and capital-markets execution and around Khan for product lineage and strategic credibility. SO011, SO012, SO022
CO020 Quantinuum closed a $300 million equity round at a $5 billion pre-money valuation in January 2024. SO009, SO026
CO021 The January 2024 round was anchored by JPMorgan Chase and included Mitsui, Amgen, and Honeywell. SO009, SO026
CO022 Quantinuum said the January 2024 financing brought total capital raised since inception to approximately $625 million. SO009, SO014
CO023 Honeywell announced an approximately $600 million Quantinuum financing at a $10 billion pre-money valuation on September 4, 2025. SO010, SO021, SO023
CO024 The September 2025 round added Quanta Computer, NVentures, and QED Investors while existing shareholders JPMorganChase, Mitsui, Amgen, Cambridge Quantum Holdings, Serendipity Capital, and Honeywell reinvested. SO010, SO021, SO023
CO025 Quantinuum's S-1 says the company completed a November 2025 funding round issuing $838.8 million of Series B convertible redeemable preferred stock. SO030
CO026 Combining the 2024 $300 million round with the $838.8 million Series B disclosed in the S-1 implies at least about $1.14 billion of post-merger external capital raised through late 2025. SO009, SO030
CO027 Quantinuum reported $30.9 million of revenue and a $192.6 million net loss for 2025, versus $23.0 million of revenue and a $144.1 million net loss for 2024. SO030, SO025, SO033
CO028 For the quarter ended March 31, 2026 Quantinuum reported $5.2 million of revenue and a $136.6 million net loss. SO030, SO033, SO035
CO029 Quantinuum had $677.0 million of cash and cash equivalents as of March 31, 2026. SO030
CO030 Quantinuum said it had a staff of almost 400 people at inception in late 2021. SO008
CO031 Public 2024 disclosures described Quantinuum as having almost 500 employees, including more than 370 scientists and engineers. SO009, SO014, SO026
CO032 Quantinuum described itself as having more than 550 employees in March-April 2025 and more than 630 employees by September-November 2025. SO016, SO018, SO010, SO013
CO033 Quantinuum's March 2026 CFO announcement said the company had approximately 700 employees worldwide. SO012, SO033
CO034 The S-1 states that Quantinuum operates four commercial quantum systems today, three in Colorado and one at RIKEN in Japan, with a fifth system expected in Singapore in late 2026. SO030, SO031
CO035 Retained sources publicly name customers or collaborators including Airbus, BMW Group, Honeywell, HSBC, JPMorgan Chase, Mitsui, Thales, Amgen, SoftBank, and RIKEN. SO009, SO013, SO024, SO026
CO036 Retained public sources disclose named deployments and collaborations but do not provide an exact current paying-customer count. SO009, SO024, SO030
CO037 The S-1 says Quantinuum historically funded operations primarily through convertible debt that subsequently converted to equity and preferred stock, but retained chapter sources do not identify a current standalone credit facility. SO030
CO038 Quantinuum launched InQuanto as a standalone platform in May 2022 and cited collaborations with BMW, Honeywell, JSR, Nippon Steel, and TotalEnergies. SO019, SO006
CO039 Quantinuum released a major lambeq update in March 2022, expanding accessibility for its open-source quantum natural language toolkit. SO015, SO007
CO040 Quantinuum launched H2-1 in June 2024 with 56 trapped-ion qubits and reported a 100x random-circuit-sampling improvement over Google's 2019 benchmark. SO014, SO027
CO041 In September 2024 Microsoft and Quantinuum said they had created 12 logical qubits on H2 and integrated InQuanto into Azure Quantum Elements. SO027, SO020
CO042 Quantinuum became a founding collaborator for NVIDIA's accelerated quantum research center in March 2025. SO016, SO010
CO043 Helios launched in November 2025 with 98 physical qubits, 48 logical qubits, cloud and on-prem access, and early users including Amgen, BMW Group, JPMorganChase, and SoftBank. SO013, SO030
CO044 JPMorganChase and Quantinuum announced a certified-randomness milestone in March 2025 using the 56-qubit H2 trapped-ion system. SO029, SO018
CO045 Quantinuum confidentially submitted its draft S-1 in February 2026, disclosed that confidential filing in April 2026, and publicly filed on May 8, 2026 under the ticker QNT. SO024, SO025, SO022, SO030
CO046 IPO-era analysis frames Quantinuum as an early-commercialization company whose public valuation case depends heavily on future Apollo and broader quantum adoption rather than current revenue scale. SO032, SO033, SO034
CO047 Independent analysis highlights unresolved risks around trapped-ion scalability, customer concentration, isotopically enriched materials, and helium supply. SO033, SO035
CO048 Independent valuation analysis argues that a $15-20 billion or $20 billion IPO target would imply an aggressive multiple relative to Quantinuum's modest revenue base and widening losses. SO033, SO034, SO035
CM001 Quantinuum publicly markets a full-stack quantum offering spanning trapped-ion hardware, cloud access, developer tooling, chemistry software, and quantum-randomness cybersecurity. SM001, SM002, SM003, SM004, SM009, SM010
CM002 Quantinuum’s H1 and H2 systems are trapped-ion quantum computers with all-to-all connectivity, mid-circuit measurement, conditional logic, and qubit reuse. SM009, SM010, SM011, SM012
CM003 Quantinuum’s current public hardware range spans H1 systems with N ≥ 20 qubits and H2 systems with N ≥ 56 qubits. SM011, SM012
CM004 Quantinuum Nexus is a cloud-based platform that offers multiple backends, dedicated simulators, JupyterHub, and collaboration features. SM004
CM005 AWS Braket, Azure Quantum, IBM Quantum, and Quantinuum Nexus show that cloud access is a primary procurement route and a status-quo substitute for direct vendor contracting. SM004, SM024, SM025, SM026
CM006 Airbus frames quantum computing as complementary to traditional HPC for aerospace simulation, not a wholesale replacement for classical high-performance computing spend. SM027
CM007 Quantum Origin is positioned as a deploy-anywhere QRNG product that requires no extra hardware or cloud connection, so the relevant cybersecurity spend is narrow cryptographic enablement rather than the whole security market. SM001, SM018
CM008 QED-C reports a 2025 global quantum market size of $1.9 billion. SM013, SM028
CM009 QED-C counts 7,420 quantum-engaged organizations and 556 pure-play quantum companies at the end of 2025. SM013, SM028
CM010 QED-C reports 16,482 pure-play workers and 8,261 new quantum-related position openings in 2025. SM013, SM028
CM011 The Business Research Company estimates the global quantum computing market at $3.62 billion in 2025. SM015
CM012 The Business Research Company forecasts the quantum computing market will reach $16.27 billion in 2030 at a 33.7% CAGR. SM015
CM013 The ResearchAndMarkets summary says global quantum investments surpassed $1 billion in 2024 and average funding rounds regularly exceeded $50 million in 2025. SM014
CM014 ResearchAndMarkets describes the market as a multi-layered stack covering hardware platforms, software development tools, cloud services, and industry-specific applications. SM014
CM015 NIST says the Commerce Department signed 2026 letters of intent totaling about $2.013 billion for quantum foundries and quantum computing companies. SM019, SM029
CM016 The gap between QED-C’s $1.9 billion measured 2025 market and broader vendor-market estimates reflects definitional differences between observed revenue and broader stack/TAM framing. SM013, SM014, SM015
CM017 Broad TAM narratives overstate Quantinuum’s near-term SAM when they include future infrastructure build-out, generic application categories, or market layers without current buyer budgets. SM014, SM015, SM017
CM018 Quantinuum and HSBC publicly identified banking use cases in cybersecurity, fraud detection, and natural-language processing. SM005
CM019 The JPMorgan-Argonne-Quantinuum QAOA work highlights logistics, telecommunications, financial modeling, and materials science as relevant optimization/application areas. SM006
CM020 InQuanto is marketed for complex molecular and materials simulations and explicitly tied to drug discovery and next-generation material design. SM002, SM008
CM021 The Mitsui/QIDO launch positions hybrid quantum-classical chemistry workflows as a way to reduce the time and cost of drug and materials discovery. SM007
CM022 Airbus points to quantum computing use cases in fluid dynamics, finite-element simulation, aerodynamics, flight mechanics, fuel-cell chemistry, optimization, and navigation. SM027
CM023 Government demand is driven by economic and national-security strategy, with coordinated federal programs and direct incentives intended to accelerate domestic quantum capacity. SM019, SM020, SM029
CM024 DOE and Berkeley Lab position national laboratories and HPC-adjacent researchers as early users for quantum work in physics, chemistry, biology, open testbeds, and workforce development. SM020, SM021
CM025 IBM’s 300+ network members and the AWS/Azure cloud offers indicate many buyers access quantum through R&D, innovation, and cloud budgets rather than on-premises infrastructure purchases. SM024, SM025, SM026
CM026 The Business Research Company lists healthcare, BFSI, automotive, energy and utilities, chemicals, and manufacturing among the major end-user categories. SM015
CM027 Quantinuum’s all-to-all connectivity, mid-circuit measurement, conditional logic, and qubit reuse broaden the set of optimization and simulation workloads it can target relative to less flexible architectures. SM009, SM010, SM011, SM012
CM028 Cloud and hybrid workflow platforms reduce adoption friction by removing classical infrastructure management and offering priority access or multi-backend experimentation. SM004, SM024, SM025
CM029 The Business Research Company identifies rising startup investment and commercialisation momentum as primary growth drivers for the market. SM015
CM030 Government funding and strategic policy remain major tailwinds because public-sector payers are willing to finance capability build-out before broad enterprise ROI is proven. SM013, SM019, SM029
CM031 QED-C says the first useful quantum applications are expected in roughly three to five years, while fully fault-tolerant quantum computing is still years away. SM013
CM032 Talent shortages that bridge physics, hardware engineering, and software remain one of the biggest scaling bottlenecks for the industry. SM013, SM017, SM023
CM033 Export controls and deemed-export rules complicate international collaboration, hiring, and cross-border research in quantum computing. SM016, SM023
CM034 Regulatory fragmentation, data-sovereignty pressures, and national-security controls create heavier compliance burdens in finance, pharma, aerospace/defense, and materials-related deployments. SM016, SM017
CM035 NIST’s PQC program says quantum computers may be years or decades away from breaking current cryptography, but migration work has to start well before that date. SM001, SM018
CM036 Airbus’s own market framing reinforces that near-term aerospace demand is for quantum-assisted HPC workflows, which tempers compute-displacement TAM claims. SM027
CM037 Forecast spread should be treated as a set of boundary markers rather than a single underwriting base case. SM013, SM014, SM015
CM038 Quantinuum’s near-term serviceable demand is concentrated in chemistry/materials R&D, finance and optimization pilots, cybersecurity randomness, and public-sector research rather than general enterprise compute. SM001, SM002, SM005, SM006, SM007, SM019, SM020
CM039 Quantinuum competes with access marketplaces and network ecosystems as well as with other hardware vendors because buyers can choose AWS, Azure, IBM, or direct Nexus routes. SM004, SM024, SM025, SM026
CM040 The practical market boundary should exclude generic AI, semiconductor, and HPC budgets unless a buyer is procuring quantum-specific access, algorithms, software, or security products. SM014, SM015, SM027
CM041 In most early segments, the end user is a scientist, quantitative researcher, or security team while the payer sits in an innovation, R&D, security, or public-program budget. SM005, SM006, SM007, SM027, SM029
CM042 Quantinuum’s bundling of developer tools, cloud access, hardware, and applications supports a land-and-expand motion from experimentation into larger application or security deployments. SM001, SM003, SM004, SM009, SM010
CM043 The Business Research Company’s $3.62 billion 2025 market estimate conflicts with QED-C’s $1.9 billion 2025 market size because the publishers are using different boundaries. SM013, SM015
CM044 Workforce scarcity suggests commercialization pace can lag capital availability even when headline TAM narratives and funding rounds keep expanding. SM013, SM017
CP001 Quantinuum's relevant landscape spans trapped-ion, superconducting, neutral-atom, photonic, annealing, and classical-HPC substitutes rather than one narrow peer set. SP002, SP018, SP023, SP025, SP027
CP002 Quantinuum positions its systems around trapped-ion QCCD scaling, all-to-all connectivity, and mid-circuit measurement. SP002
CP003 Quantinuum markets direct subscriptions, Azure access, grants, and on-prem Helios deployment rather than a single cloud-only route. SP002, SP003
CP004 pytket is designed to build, compile, and run circuits across multiple formats, devices, and simulators, which reduces workflow lock-in at the compiler layer. SP004
CP005 IBM discloses Open, Pay-As-You-Go, Flex, Premium, and On-Prem quantum plans, with public minute-based prices for the cloud tiers. SP005
CP006 IBM publicly reports 30 or more quantum computers above 100 qubits, 2,300 or more available qubits, 97% availability, and 3.9 trillion or more circuits run. SP006
CP007 IBM couples open-source Qiskit tooling with IBM Quantum Platform services such as Runtime and the Functions catalog. SP005, SP007
CP008 Google's Willow program remains selective in 2026 because the company says the hardware is not yet available to the public. SP009
CP009 Google competes as an integrated hardware-and-software research stack rather than as a broadly packaged enterprise procurement offer. SP008, SP009, SP010
CP010 IonQ is Quantinuum's closest trapped-ion commercial overlap because it combines direct cloud access, broad SDK support, and a 2026 logical-qubit roadmap. SP011, SP012
CP011 IonQ also sells Forte Enterprise for data centers, giving it a direct on-prem route against Quantinuum's Helios packaging. SP013
CP012 IonQ reported $64.7 million of Q1 2026 revenue, with about 60% from commercial customers and remaining performance obligations of $470 million. SP014
CP013 Rigetti competes on a superconducting full stack that emphasizes hybrid quantum-classical coupling and deep control over hardware and software. SP015
CP014 Rigetti says its QCS platform supports less than one millisecond connectivity between customer classical resources and Rigetti QPUs. SP015
CP015 Rigetti's Novera page describes a commercially available 9-qubit on-prem QPU with immediate shipment and cites roughly $5.7 million of purchase orders for two systems. SP016
CP016 IQM markets superconducting systems at 5, 20, 54, and 150 qubits plus an accessible quantum cloud platform. SP017
CP017 AWS Braket lists IQM, IonQ, QuEra, and Rigetti side by side, which shows that cloud intermediaries can broker access across multiple modalities. SP018
CP018 QuEra's Aquila offers 256 neutral-atom qubits, AWS access over 100 hours per week, and a premium direct access tier. SP019
CP019 QuEra says it achieved 256 physical and more than 10 logical qubits on the path to fault tolerance. SP020
CP020 OQC markets commercial-grade quantum systems available in the cloud, colocated in sovereign data centres, or embedded through APIs. SP021
CP021 PsiQuantum frames its strategy around million-qubit, utility-scale photonic systems built with mainstream semiconductor manufacturing. SP025
CP022 PsiQuantum says its stack already uses 300mm wafers, more than one million tested devices, and standard fiber networking to enable scale. SP025
CP023 Nature corroborates that PsiQuantum has demonstrated a manufacturable photonic platform at the component level, but it does not establish broad commercial availability. SP026
CP024 AWS Braket reduces switching costs because buyers can sample IonQ, IQM, QuEra, and Rigetti through one procurement and execution surface. SP018, SP022
CP025 Amazon Braket pricing discloses a shared per-task model plus per-shot or reservation economics, including an IonQ Forte error-mitigation minimum of 2,500 shots. SP022
CP026 IBM's explicit pricing makes it easier to buy than Google or Quantinuum, but IBM-specific runtime and services can still create workflow stickiness. SP005, SP007
CP027 pytket lowers app-layer lock-in because it imports external formats and targets a range of devices and simulators rather than a single Quantinuum-only backend. SP004
CP028 Hardware switching remains hard because trapped ions, superconducting qubits, neutral atoms, photonics, and annealing expose different connectivity, timing, and workload behaviors. SP002, SP018, SP019, SP023, SP025, SP027
CP029 D-Wave is a real substitute for optimization buyers because Advantage2 and Leap are available now through cloud and on-prem routes. SP023, SP024
CP030 D-Wave says Leap provides 99.9% uptime, subsecond responses, and hybrid solving at up to two million variables and constraints. SP023, SP024
CP031 Google's current proposal-gated access means it exerts more long-term ecosystem pressure than short-term procurement pressure. SP008, SP009, SP010
CP032 IEEE Spectrum argues the quantum industry will not reach broadly useful commercial quantum computing in 2026. SP027
CP033 IEEE Spectrum frames the first 2026 error-corrected customer systems as scientific advantage rather than commercial advantage. SP027
CP034 IEEE Spectrum says neutral atoms have scalability benefits but slower operations than superconducting systems, which highlights modality trade-offs rather than a settled winner. SP027
CP035 Quantinuum's moat today is best described as high-fidelity trapped-ion performance plus enterprise deployment breadth, not exclusive distribution. SP002, SP003, SP004, SP017
CP036 The strongest commoditization force in this market is the app and procurement layer, where cloud aggregators and open SDKs standardize discovery and workflow setup. SP004, SP007, SP018, SP022
CP037 IBM's modular FTQC roadmap, QuEra's logical-qubit progress, and PsiQuantum's manufacturing thesis are the most credible long-term displacement threats in the retained source pack. SP006, SP020, SP025, SP026, SP027
CP038 Overall, Quantinuum looks better positioned today than selective-access Google and long-dated PsiQuantum, but it is not insulated from IBM, IonQ, and cloud-driven multi-homing pressure. SP003, SP005, SP009, SP012, SP018, SP025
CP039 Public price discovery is uneven because IBM and AWS publish explicit economics while Quantinuum, Google, OQC, and much of IQM remain quote-led or opaque on the cited pages. SP005, SP009, SP017, SP021, SP022
CP040 For many buyers, the practical status quo remains classical HPC, simulators, and limited pilot access because much of the universal quantum market is still immature or gated. SP004, SP009, SP024, SP027
CI001 Quantinuum reported $30.9 million of 2025 net revenue and a $192.6 million net loss, versus $23.0 million of revenue and a $144.1 million net loss in 2024. SI001, SI020, SI021
CI002 Quantinuum’s Q1 2026 net revenue fell to $5.2 million from $19.1 million in Q1 2025 while net loss widened to $136.6 million from $30.5 million. SI001, SI020, SI021
CI003 2025 bookings were $79.3 million, but Q1 2026 bookings were only $1.3 million versus $1.9 million in Q1 2025, underscoring lumpy contract timing. SI001, SI022
CI004 Cash and cash equivalents were $677.0 million at March 31, 2026, down from $762.6 million at December 31, 2025. SI001, SI022
CI005 Remaining performance obligations were $80.7 million at December 31, 2025, and only about 31% was expected to be recognized within 12 months. SI001, SI003
CI006 Honeywell’s filings show Quantinuum-related backlog of about $82 million at year-end 2025 and about $81 million at March 31, 2026 within Corporate and All Other. SI002, SI003
CI007 Honeywell’s 2025 10-K and 2026 10-Q describe Quantinuum as a majority-owned investment whose revenue in Corporate and All Other comes from integrated quantum-computing hardware and software solutions. SI002, SI003
CI008 Quantinuum says it derives revenue from specialized quantum-computing hardware sales, access contracts with maintenance and support services, and consulting services related to co-developing algorithms on its systems. SI001, SI007
CI009 Cloud-platform, research, and support contracts are generally sold as fixed-fee service periods with variable usage fees above contractual minimums and are recognized over time. SI001
CI010 Specialized hardware revenue is recognized when control transfers, usually on delivery and commissioning, and some arrangements qualify as sales-type leases. SI001
CI011 2025 revenue growth was driven by $16.5 million of specialized-hardware revenue from a sales-type lease transaction, partly offset by an $8.5 million decrease in cloud-platform, research, and support-services revenue. SI001
CI012 The 73% Q1 2026 revenue decline was primarily a comparability issue because Q1 2025 included the upfront recognition of a $16.5 million 45-month sales-type lease transaction. SI001
CI013 Public filings say cost of revenue is driven by operations, reliability and customer-support labor, depreciation of quantum systems, cloud and facility infrastructure, and third-party contractors. SI001
CI014 Quantinuum disclosed 2025 direct cost of revenue of $4.73 million and amortization expense of $11.36 million against $30.93 million of revenue, but it did not disclose a management-presented gross-margin figure or product-level margin split. SI001, SI020
CI015 Research and development expense was $165.4 million in 2025, about 5.35x reported revenue, showing that roadmap spending dominates the current cost base. SI001, SI020
CI016 Sales and marketing expense was $18.9 million and general and administrative expense was $29.9 million in 2025, so commercial and corporate overhead materially exceed direct cost of revenue. SI001
CI017 Operating cash use was $62.9 million and capex was $22.7 million in Q1 2026 versus only $5.2 million of revenue, implying a capital-consuming model even after the 2025 raise. SI001, SI021
CI018 Operating cash use was $160.3 million in 2025 and $120.9 million in 2024, indicating burn expanded before the IPO process began. SI001, SI020
CI019 Q1 2026 working-capital outflow included a $9.4 million prepayment to Honeywell under a strategic services and supply agreement, showing Quantinuum still relies on related-party support arrangements. SI001, SI002
CI020 Quantinuum says Honeywell has provided infrastructure, supply-chain relationships, management expertise, and early customer demand, and intends to remain a strategic customer and partner after the IPO. SI001, SI006
CI021 Honeywell announced a $300 million equity round for Quantinuum in January 2024 at a $5 billion pre-money valuation, lifting total capital raised since inception to about $625 million. SI004, SI019
CI022 Honeywell announced an approximately $600 million equity raise in September 2025 at a $10 billion pre-money valuation to fund Helios commercialization and the path to fault-tolerant computing. SI005, SI018, SI019
CI023 The 2025 raise roughly doubled Quantinuum’s private valuation in under two years while public 2025 revenue remained only $30.9 million, highlighting that capital access is being priced off roadmap expectations more than current scale. SI001, SI004, SI005, SI021
CI024 Quantinuum’s public S-1 filing does not yet set share count or price, so IPO proceeds are a prospective rather than current runway source. SI001, SI006, SI023
CI025 The S-1 says Quantinuum Holdings will use IPO proceeds for general corporate purposes and offering expenses, leaving the exact split across R&D, system build-out, and working capital undefined. SI001
CI026 Quantinuum frames monetization as layered: defend leadership in on-prem and cloud quantum systems first, then expand into higher-margin software, applications, and outcome-driven intellectual property. SI001, SI008
CI027 Nexus had more than 150 user organizations and more than 750 active users by March 9, 2026, providing a public software-adoption proxy even though software revenue is not broken out. SI001, SI011
CI028 Quantinuum’s Compute Platforms group owns the P&L for InQuanto, Quantum Systems (both cloud and HaaS), and Nexus, indicating these are managed as monetizable product lines rather than pure research tools. SI007
CI029 Quantinuum’s systems page offers direct subscriptions, InQuanto licenses, and Microsoft Azure subscriptions, showing the company sells both direct-enterprise access and partner-channel access rather than only bespoke projects. SI008, SI010
CI030 Azure Quantum publicly lists Quantinuum Standard and Premium monthly subscriptions at $125,000 and $175,000, plus a pay-as-you-go HQC formula, giving rare public list pricing for H2 access. SI015, SI016
CI031 Azure’s provider page says syntax checkers are free, emulator use is bundled with a hardware subscription, and queued sessions can get temporary exclusive hardware access, supporting an enterprise subscription or HaaS model. SI015, SI016
CI032 Amazon Braket’s pricing model is per-task, per-shot, or hourly reservation, but the fetched public page did not list Quantinuum-specific device pricing, so AWS appears to be a channel without clear public Quantinuum list prices. SI017
CI033 Quantum Origin is delivered as a software binary plus a pre-generated quantum seed, runs on standard CPUs, works in air-gapped environments, and needs no cloud connection or extra hardware, supporting a software-security licensing model. SI013, SI014
CI034 InQuanto is sold as a quantum-chemistry platform with access to Quantinuum systems, partner residency, and Nexus integration, but no public seat or contract pricing is disclosed. SI010, SI011
CI035 Nexus is a cloud platform with multi-backend orchestration, hosted JupyterHub, dedicated simulators, quotas, and collaboration controls, consistent with platform-software monetization and administrator-oriented enterprise packaging. SI011, SI012
CI036 Helios is available both through Quantinuum’s cloud service and on-premises deployments, meaning the hardware business spans remote access, dedicated installations, and related support obligations. SI009, SI024
CI037 Customer concentration is extreme: RIKEN was 60% of 2025 revenue, and another government-affiliated research institution plus the U.S. government together represented 71% of Q1 2026 revenue. SI001, SI020
CI038 The S-1 warns that many large customer relationships are pilot programs, research collaborations, or grant-funded projects rather than long-term production commitments, limiting recurrence visibility. SI001, SI020, SI021
CI039 RIKEN’s on-prem installation and BMW’s multi-year collaboration show enterprise traction and long sales cycles, but both are framed as research and co-development engagements rather than disclosed production-scale revenue contracts. SI024, SI025, SI021
CI040 The Next Web argues the IPO asks public investors to pay more than 600x revenue for Apollo, a fault-tolerant system planned for 2029, so valuation depends more on future technical success than current economics. SI021
CI041 Constellation Research says Quantinuum’s revenue is meager and lumpy and that no company is running quantum computing in production at a scale that materially affects its bottom line, challenging near-term commercialization claims. SI020
CI042 HPCwire notes that Quantinuum’s traditional IPO route may add credibility, but it also means the company will face greater public-market scrutiny than SPAC-era peers as it seeks new capital. SI019
CI043 The public record still withholds the core underwriting metrics: revenue mix by hardware versus software, realized prices and discounts, gross margin, renewal terms, and a clean post-IPO view of shared-services economics with Honeywell. SI001, SI002, SI003, SI020
CI044 Quantinuum announced a 2026 CHIPS R&D letter of intent for proposed federal support and domestic supplier partnerships, but it is not closed cash and therefore cannot be treated as funded runway yet. SI026
CI045 Quantinuum had approximately 700 employees by March 2026, including about 410 hardware experts and 105 software experts, so the company is supporting a large technical organization relative to current revenue. SI001, SI007
CE001 Quantinuum publicly positions itself as a full-stack quantum company spanning hardware, software, developer tools, and cybersecurity products. SE001, SE005
CE002 System Model H1 is Quantinuum's first-generation trapped-ion system and uses a single linear architecture with all-to-all connectivity, mid-circuit measurement, conditional logic, and qubit reuse. SE002, SE010
CE003 The H1 data sheet specifies 20 physical 171Yb+ qubits, five parallel gate zones, typical two-qubit gate infidelity of 1×10^-3, and typical single-qubit gate infidelity of 2×10^-5. SE002, SE010
CE004 System Model H2 is Quantinuum's second-generation trapped-ion system and moves from a linear layout to a racetrack-style architecture while keeping all-to-all connectivity, mid-circuit measurement, conditional logic, and qubit reuse. SE003, SE011
CE005 The H2 data sheet specifies 56 physical 171Yb+ qubits, four parallel two-qubit zones in two connected linear sections, typical two-qubit gate infidelity of 1×10^-3, and typical SPAM error of 1×10^-3. SE011, SE016
CE006 Azure Quantum exposes H2-1 and H2-2 syntax checkers, emulators, and QPUs, and states that Quantinuum targets support integrated hybrid circuits. SE016
CE007 Helios is marketed as a 98-physical-qubit flagship with logical qubits, Guppy, NVIDIA GPUs integrated into the control system, and both cloud and on-prem access. SE004, SE012
CE008 The Helios data sheet specifies 98 physical 137Ba+ qubits, eight gate zones, a circular storage ring with two linear gating sections, real-time arithmetic and control flow, and typical two-qubit infidelity of 8×10^-4. SE004, SE012
CE009 Data Center Dynamics reported at launch that Helios was positioned for general-purpose hybrid workloads and named Amgen, BlueQubit, BMQ Group, JPMorganChase, and SoftBank as early adopters or users. SE027
CE010 Quantinuum says an NVIDIA GPU-based decoder integrated in the Helios control engine improved logical fidelity by more than 3%. SE008, SE026
CE011 Quantinuum's developer-tools page describes TKET as an open-source toolkit that constructs programs, compiles and optimizes them for target hardware, and executes them through many extension modules. SE005, SE013
CE012 The TKET docs present the toolkit as a build-compile-run workflow: construct circuits in Python, convert to native gates, optimize for fidelity, and execute on devices or simulators. SE013
CE013 The public TKET repository states that the stack consists of a high-performance C++ core, Python bindings, separate extension packages, and an experimental C API. SE018
CE014 The TKET public repo shipped v2.17.0 on 2026-05-13 and v2.18.0 on 2026-05-15, indicating active release cadence in May 2026. SE019, SE022
CE015 PyPI shows pytket 2.18.0 supports Python 3.10+, cross-platform wheels, and a pytket-quantinuum extension path. SE022
CE016 Pepy reports about 4.5 million all-time pytket downloads and 176.6k downloads in the last 30 days, indicating materially broader public developer uptake than Quantinuum's other retained open-source tools. SE024
CE017 Quantinuum markets InQuanto as a quantum chemistry platform for complex molecular and materials simulations and pairs it with partner residency and direct system access. SE006, SE014
CE018 Quantinuum says the latest InQuanto release connects to Nexus, integrates NVIDIA cuTensorNet via pytket-cutensornet, and is up to 10× more accurate and resource-efficient than leading open-source competitors. SE006
CE019 InQuanto docs show the software covers chemical specification, circuit execution on simulators and hardware, and error-reduction methods such as symmetry verification. SE014
CE020 Quantum Origin is positioned as a software-only cryptographic product that uses a quantum-generated seed to enhance system randomness without extra hardware or cloud connectivity. SE007
CE021 Quantinuum says Quantum Origin verifies seed quality with a Bell test and already lists Honeywell, Fornetix, Thales, Keyfactor, and Senetas as integration examples or endorsements. SE007, SE001
CE022 Quantinuum's developer-tools page positions lambeq as an open-source QNLP library that designs end-to-end pipelines and integrates with TKET. SE005
CE023 The lambeq docs describe a pipeline that converts sentences into quantum circuits, provides tutorials and use cases, and supports extensibility for custom features. SE015, SE029
CE024 The public lambeq repo requires Python 3.10+, installs via pip, uses the Bobcat parser by default, is Apache 2.0 licensed, and warns that DepCCG support is no longer actively maintained. SE020, SE023
CE025 The lambeq 0.5.0 release on 2026-05-15 added DisCoCircReader, OncillaParser, PytorchQuantumModel, and made tket an optional dependency, showing continuing but research-oriented maintenance. SE021
CE026 Pepy reports about 96.9k all-time lambeq downloads and 3.7k downloads in the last 30 days, far smaller than pytket's public footprint. SE025
CE027 Quantinuum docs and OLCF guidance describe Nexus as the control plane for hardware and compilation services, with APIs, portal access, and machine-availability views. SE009, SE017
CE028 OLCF recommends a staged workflow of syntax checker to emulator to quantum computer and lists H2 and Helios endpoints plus quota controls, implying Quantinuum access remains mediated rather than purely self-serve. SE017
CE029 Azure documents support for mid-circuit measurement and reset, arbitrary-angle ZZ gates, optional SU(4) entanglers, emulator noise parameters, and TKET optimization settings in the Quantinuum stack. SE011, SE016
CE030 Azure and OLCF both describe hardware availability windows, queues, quotas, and upgrade or development periods rather than always-on commodity service levels. SE016, SE017
CE031 Both the H1 and H2 data sheets state that TKET is supported in the stack and automatically provides circuit optimization for submitted circuits. SE010, SE011
CE032 The H1 and H2 product pages advertise direct subscriptions, Azure Quantum access, and ORNL or QCUP access as the principal public entry points for H-Series usage. SE002, SE003, SE017
CE033 Quantinuum positions Helios as the first platform in its next-generation software stack with Guppy and GPU-linked control, available through cloud or on-prem deployment. SE004, SE008
CE034 Quantinuum and Microsoft reported four logical qubits with 800× lower logical error rates than physical error rates and 14,000 error-free circuit instances on Quantinuum hardware. SE028
CE035 The same Microsoft-Quantinuum release says the logical-qubit milestone used 30 of 32 physical qubits on an H2 processor and relied on 99.8% two-qubit gate fidelity, active syndrome extraction, and all-to-all connectivity. SE028, SE003
CE036 Public hardware materials show a clear scale-up path from H1's 20-qubit linear trap to H2's 56-qubit racetrack QCCD to Helios' 98-qubit circular-ring system with more gate zones and real-time control features. SE010, SE011, SE012
CE037 Quantinuum's public portfolio combines hardware, compiler/runtime, application software, and a separate cyber product, which is broader than a hardware-only quantum supplier model. SE001, SE005, SE006, SE007
CE038 Public access and deployment still depend on managed portals, emulators, partner channels, or on-prem engagements rather than a simple commodity API-only model. SE016, SE017, SE027
CE039 In retained public evidence, trust signals are strongest for technical validation and software provenance, while public disclosure of enterprise security certifications or formal trust-center materials remains thin. SE007, SE016, SE017, SE018, SE022
CE040 The main product risks are continuing hardware upgrade cycles, quota-gated access, limited public SLA data, and the integration complexity of hybrid classical-quantum workflows. SE016, SE017, SE026, SE027
CE041 The home page and Quantum Origin page tie the cyber product to named integrations and quantum-resilience messaging from Honeywell and Thales, suggesting the security line is commercially positioned now rather than only as research. SE001, SE007
CE042 The Helios product page identifies JPMorganChase, BMW Group, SoftBank, and Amgen as early public workloads or collaborators across finance, materials, telecom, and biologics. SE004
CE043 Quantinuum and NVIDIA describe their collaboration as spanning Helios hardware, Guppy and CUDA-Q developer tools, and AI-assisted chemistry workflows such as ADAPT-GQE. SE008, SE026
CU001 Quantinuum’s S-1 says active customer engagements are primarily focused across pharmaceuticals, materials science, financial services, government, and industrial markets and names JPMorgan Chase, Amgen, Mitsui & Co., and Honeywell as customers and innovation partners. SU019
CU002 Quantinuum’s Helios launch materials publicly name Amgen, BMW Group, JPMorganChase, and SoftBank as early users or collaborators and pair that launch with a Singapore in-country access partnership. SU001, SU003
CU003 Helios is sold through cloud and on-premise channels and the product page routes access through a contact form rather than a public price card. SU001, SU003
CU004 Quantinuum’s Q-Net page says the company has 2,500+ active Nexus users and 170+ research papers using Quantinuum systems. SU007
CU005 Quantinuum says Q-Net Connect 2026 drew over 170 attendees from commercial enterprises, startups, academia, research institutions, public-sector bodies, and non-profits. SU008
CU006 Quantinuum says JPMorganChase received the 2026 Guppy Adopter Award for exemplary adoption of the Guppy programming language in its research workflows. SU008
CU007 Quantinuum and BMW say their collaboration began in 2021 and was expanded into a multi-year partnership in May 2026. SU009, SU012, SU025
CU008 BMW will use the current Helios system and planned Sol and Apollo generations, tying account expansion to Quantinuum’s hardware roadmap. SU009, SU025
CU009 BMW, Airbus, and Quantinuum used Quantinuum’s H-Series quantum computer in an industrial workflow focused on oxygen-reduction-reaction and fuel-cell catalyst simulations. SU009, SU010, SU011
CU010 The BMW and Airbus proof is strong on industrial R&D specificity but does not disclose production contract value, deployment scale, or recurring commercial economics. SU009, SU010, SU011
CU011 JPMorganChase, Quantinuum, and U.S. national-lab partners announced a certified-randomness milestone on Quantinuum’s 56-qubit H2 system in March 2025. SU013, SU014
CU012 JPMorganChase’s own blog says the protocol certified at least 71,313 bits of entropy and treated the remote Quantinuum system as an untrusted server whose outputs were verified with supercomputers. SU014
CU013 The filing and Helios materials together make JPMorganChase a named financial-services customer and innovation partner rather than only an investor. SU019, SU003, SU014
CU014 HSBC’s partnership with Quantinuum covers quantum-hardened cryptographic keys as well as longer-term work in risk management, fraud detection, and quantum machine learning. SU016
CU015 Quantinuum’s HSBC page frames the best public proof as a pilot that uses Quantum Origin QRNG to secure tokenized gold rather than as a broad production deployment. SU015, SU020
CU016 Quantinuum’s Helios materials describe Amgen as an investor and research collaborator applying hybrid quantum-machine learning to biologics discovery. SU001, SU003
CU017 Fierce Biotech says Quantinuum’s filing names Amgen’s partnership and pharmaceuticals as active customer engagement, but the public proof remains research-stage and does not disclose production workflow metrics. SU024, SU019
CU018 Quantinuum’s Singapore partnership says Helios will be hosted in-country in 2026 and targeted at computational biology, finance, advanced materials, and combinatorial optimization. SU005, SU006
CU019 The Singapore model separates payer and infrastructure sponsor from end user because NQO, NQCH, and related public bodies fund access while researchers and industry participants use the system. SU005, SU006
CU020 Quantinuum says RIKEN procured an H2 system in 2026 after previously using H1 in Reimei-Fugaku, creating clear repeat-procurement proof in public-sector research. SU022
CU021 RIKEN says its H1 platform was widely used by JHPC-quantum users and that the H2 upgrade should support larger workloads and help demonstrate quantum advantage in hybrid computing. SU022
CU022 Quantinuum’s QIDO launch says Mitsui distributed the platform in Japan and beta testing involved JSR Corporation, Panasonic Holdings Corporation, and Chugai Pharmaceutical before general release. SU021
CU023 The QIDO beta-user testimonials are useful customer proof because they cite concrete chemistry workflows, but they also openly describe early validation and remaining technical challenges. SU021
CU024 Quantinuum’s Synopsys collaboration and Microsoft integration show partner- and channel-led expansion into engineering and cloud ecosystems. SU023, SU017, SU018
CU025 Public Microsoft materials in the retained set describe Azure Quantum Elements integration and private preview access, not Microsoft as a disclosed direct end-customer of Quantinuum hardware. SU017, SU018
CU026 Quantinuum’s Helios technical blog says SoftBank and JPMorgan conducted commercially relevant research during a two-month early-access program before Helios became generally available. SU004, SU001
CU027 Across the retained public sources, Quantinuum does not disclose a total paying customer count or a product-level count of cloud, on-prem, or community users converted to contracts. SU019, SU001, SU007
CU028 Across the retained public sources, Quantinuum does not disclose NRR, GRR, churn, renewal rate, contract length, or a public price card. SU019, SU003, SU007
CU029 Quantinuum’s S-1 says one customer accounted for $18.7 million, or 60%, of 2025 revenue and $14.6 million, or 63%, of 2024 revenue. SU019
CU030 Quantinuum’s S-1 says one customer accounted for $2.5 million, or 47%, of Q1 2026 revenue, while a different customer accounted for $17.4 million, or 90%, of Q1 2025 revenue. SU019
CU031 Quantinuum’s S-1 says the U.S. Government represented approximately 24% of Q1 2026 revenue and 16% and 9% of 2025 and 2024 revenue, respectively. SU019
CU032 The filing’s concentration disclosures imply a revenue model driven by a few large accounts rather than a broadly diversified enterprise base. SU019
CU033 Quantinuum’s S-1 says Honeywell has served as both a testing ground and an early customer and intends to remain a strategic customer and partner after the offering. SU019
CU034 Q-Net user and event metrics indicate meaningful ecosystem breadth, but they are not equivalent to disclosed paying customers or retained ARR. SU007, SU008, SU019
CU035 Quantinuum’s route to market depends on direct enterprise selling, partner ecosystems, national hubs, and application-layer distribution rather than only a self-serve software motion. SU003, SU017, SU021, SU023
CU036 Procurement friction is visible because several public proofs rely on apply-now product access, private preview, beta testing, or bespoke co-development rather than open purchasing. SU003, SU018, SU021
CU037 The retained sources show geographic customer proof across North America, Europe, and Asia. SU009, SU010, SU011, SU018, SU021, SU022
CU038 Named enterprise proofs cluster in finance and cybersecurity, pharma and chemistry, mobility and materials, and industrial engineering. SU013, SU015, SU016, SU021, SU023
CU039 Public-sector and research proofs cluster in Singapore, Japan’s RIKEN platform, U.S. Government revenue disclosures, and national-lab participation in community events and research. SU005, SU006, SU008, SU019, SU022
CU040 BMW, RIKEN, JPMorgan, and Honeywell provide the clearest durability proxies because each shows repeat engagement, renewal, upgrade, or strategic-customer language beyond a single launch announcement. SU009, SU019, SU022, SU008
CR001 Quantinuum was established in December 2021 through the merger of Honeywell Quantum Solutions and Cambridge Quantum. SR005, SR006
CR002 Quantinuum's board is chaired by Honeywell CEO Vimal Kapur, keeping the parent company at the center of governance. SR005, SR027
CR003 Honeywell and Quantinuum said the company has publicly filed an S-1 for a Nasdaq listing under ticker QNT but the share count and price range remain undetermined. SR001, SR004
CR004 Honeywell described Quantinuum as majority owned in the April 2026 confidential filing announcement, and Reuters repeated that characterization. SR003, SR027
CR005 Honeywell's September 2025 financing round valued Quantinuum at $10 billion pre-money. SR002, SR026
CR006 The same round added Quanta Computer, NVentures, QED Investors and other backers, increasing external expectations for liquidity and scale. SR002
CR007 Public Quantinuum materials moved from over 630 employees in November 2025 to approximately 700 employees in multiple 2026 releases. SR008, SR013, SR007
CR008 Quantinuum publicly lists facilities across the United States, United Kingdom, Germany, Japan, and Singapore, underscoring a cross-border operating footprint. SR001, SR005, SR013
CR009 Quantinuum's public proof set is concentrated in a short list of named accounts and collaborators rather than a broad disclosed production customer base. SR008, SR013, SR014, SR015, SR016, SR017
CR010 Reuters listed Honeywell, Airbus, BMW Group, HSBC, and JPMorgan Chase as users of Quantinuum technology, reinforcing that public customer proof still revolves around a small set of flagship logos. SR027
CR011 BIS's September 2024 interim final rule imposed worldwide export controls on quantum computers, related equipment, components, materials, software, and technology. SR018, SR019
CR012 The BIS rule created an Implemented Export Controls exception for countries that adopt equivalent controls. SR018, SR020, SR021
CR013 The export-control package keeps deemed export and foreign-national disclosure obligations in scope for controlled quantum technology even though broad licensing was deferred. SR018, SR021, SR024
CR014 APS said the new reporting requirements impose a disproportionate compliance burden on younger or smaller quantum companies. SR024
CR015 Phillips Lytle said the new quantum controls can broaden mandatory CFIUS review because controlled quantum items become critical technologies. SR021
CR016 Quantinuum centralizes corporate governance, intellectual property, export control, privacy, and compliance oversight in one chief legal and government-relations executive. SR005
CR017 Global Legal Insights says the quantum ecosystem faces fragmented export controls, data-sovereignty pressure, supply-chain fragility, talent shortages, and compliance burdens. SR022
CR018 Global Legal Insights also argues that divergent national approaches to quantum regulation raise compliance complexity for private industry. SR022, SR023
CR019 Quantinuum's 2024 roadmap positions Apollo as a fifth-generation universal fault-tolerant system targeted for 2030, with Helios and Sol as intermediate steps. SR009, SR010
CR020 Quantinuum's technical roadmap targets thousands of physical qubits and hundreds of logical qubits in Apollo, but those are still forward-looking milestones rather than shipped capacity. SR010
CR021 The same technical roadmap said H2 offered 56 physical qubits while Helios was planned around 100 physical qubits. SR010
CR022 The Helios paper reports a 98-qubit trapped-ion processor with average two-qubit gate infidelity of 7.9×10^-4 and claims performance beyond the reach of classical simulation on random circuit sampling. SR012, SR008
CR023 Quantinuum's Helios launch says the system now exposes 94 error-detected logical qubits and 48 error-corrected logical qubits to customers, but still frames Helios as the beginning of enterprise adoption. SR008, SR032
CR024 MIT Technology Review said Helios is still not powerful enough to run the industry's hoped-for money-making algorithms in materials or finance. SR025
CR025 Google said its 105-qubit Willow chip achieved below-threshold error correction and reduced errors exponentially as qubit arrays scaled. SR029
CR026 IBM says it has delivered more than 30 quantum computers above 100 qubits since 2022 and now exposes over 2,300 available qubits across its fleet. SR028
CR027 IonQ markets Forte Enterprise as a data-center deployable trapped-ion computer with 36 physical qubits and roughly 0.4% two-qubit gate error. SR030
CR028 QuEra argues neutral-atom systems offer flexible large-array layouts, long-range interactions, and room-scale deployment without cryogenic cooling. SR031
CR029 Quantinuum's roadmap now depends on real-time error correction and hybrid execution built around NVIDIA GPUs and CUDA-Q. SR011, SR015
CR030 Quantinuum's CHIPS letter of intent names GlobalFoundries and Monarch Quantum as key component partners for scale-up. SR007
CR031 The same CHIPS initiative frames domestic semiconductor and photonics supply-chain strengthening as necessary to clear technical bottlenecks in trapped-ion scale-up. SR007
CR032 Quantinuum's Helios launch names Amgen, BMW Group, JPMorganChase, and SoftBank as early Helios users or collaborators. SR008, SR032
CR033 BMW's 2026 expansion commits the automaker to successive generations of Quantinuum systems through Helios, Sol, and Apollo. SR013
CR034 Quantinuum's first on-site deployment outside the U.S. is tied to RIKEN and a Japanese public-sector program, highlighting the strategic weight of government-backed anchor projects. SR014
CR035 Microsoft and NVIDIA collaborations are embedded in Quantinuum's claims about logical-qubit credibility, cloud distribution, and hybrid workflow adoption. SR009, SR015, SR016
CR036 SoftBank and Quantinuum said the sector still lacks mature cost-recovery strategies, clarified revenue models, and enough clear commercial use cases. SR017
CR037 SoftBank also said service timing depends on the pace of technical progress and market need, reinforcing commercialization-timing risk. SR017
CR038 APS reported that the new export-control regime intentionally preserved research collaboration but still added non-trivial reporting and legal-advice costs. SR024
CR039 Global Legal Insights said the quantum sector experienced a private-investment surge followed by market consolidation beginning in late 2022. SR022
CR040 HPCwire said analysts view Quantinuum's traditional IPO as a credibility test for the broader quantum sector as public-market scrutiny rises. SR026
CR041 Honeywell said the proposed offering remains subject to market conditions and may not complete on any disclosed timetable or terms. SR001, SR003, SR004
CR042 Quantinuum replaced founder-CEO Ilyas Khan with Raj Hazra in 2023, but Khan remains vice chair and a senior product executive. SR005, SR006
CR043 Quantinuum's executive bench now includes multiple leaders with prior Honeywell backgrounds in strategy, operations, supply chain, and legal functions. SR005
CR044 The company's legal and government affairs responsibilities remain unusually concentrated in a single executive role. SR005
CR045 The combination of Hazra as scaling operator, Khan as founder-product architect, and centralized legal/government affairs keeps key-person exposure material even with a broader bench than in 2023. SR005, SR006
CR046 Quantinuum's strongest mitigants are deep capital support, continued Honeywell sponsorship, multi-year anchor partnerships, active government engagement, and a larger leadership bench. SR002, SR007, SR013, SR014, SR015, SR016, SR017
CR047 Residual exposure remains high because liquidity, export-control burden, scaling milestones, and repeatable commercial use cases all remain open as of May 2026. SR001, SR017, SR018, SR025, SR026
CR048 The thesis should be re-underwritten if the IPO path stalls, if Sol or later systems slip materially, or if regulation and compliance restrict talent, partners, or capital. SR003, SR017, SR021, SR024, SR026
CR049 The public record still does not disclose post-IPO cap table mechanics, customer revenue mix, export-license history, or supplier concentration by spend. SR001, SR017, SR022
CV001 Quantinuum announced that it closed a $300 million equity round at a $5 billion pre-money valuation on January 16, 2024. SV001, SV005
CV002 JPMorgan Chase, Mitsui, Amgen, and Honeywell participated in the January 2024 round, and Honeywell said it remained Quantinuum’s majority shareholder. SV001, SV005
CV003 The January 2024 financing established a $5 billion private-market benchmark roughly 20 months before the September 2025 step-up to a $10 billion pre-money valuation. SV001, SV005, SV002
CV004 Honeywell said in September 2025 that Quantinuum raised approximately $600 million at a $10 billion pre-money valuation. SV002, SV007
CV005 The September 2025 round added Quanta Computer, NVentures, and QED Investors while JPMorgan Chase, Mitsui, Amgen, and Honeywell reinvested. SV002, SV007
CV006 Quantinuum said the 2025 capital raise would fund Helios rollout, manufacturing and supply-chain scaling, and the path toward universal fault-tolerant quantum computing. SV002, SV007
CV007 The reviewed public evidence cleanly corroborates the original $600 million September 2025 raise but does not provide equally clear primary-source confirmation of a later expansion to $800 million total. SV002, SV006, SV029
CV008 Quantinuum confidentially submitted a draft S-1 on February 17, 2026 and publicly announced that fact on April 22, 2026. SV003, SV006
CV009 Quantinuum publicly filed an S-1 on May 8, 2026 but said neither the number of shares nor the price range had yet been determined. SV004, SV003
CV010 Reuters described Quantinuum as majority owned by Honeywell during the 2026 IPO preparation process, implying meaningful continuing control and overhang risk. SV006, SV032
CV011 Quantinuum launched Helios commercially on November 5, 2025 and named Amgen, BMW Group, JPMorganChase, and SoftBank as launch customers. SV011, SV013
CV012 Quantinuum said it became a founding collaborator for NVIDIA’s Accelerated Quantum Research Center in March 2025. SV010, SV002
CV013 Quantinuum’s public roadmap still points to a 100-logical-qubit system by 2027 and hundreds of logical qubits by decade end. SV010, SV011
CV014 BMW expanded its work with Quantinuum into a multi-year partnership in May 2026 that includes access to current Helios systems and planned Sol and Apollo generations. SV013, SV011
CV015 Quantinuum and bp escalated from a pilot to a broader May 2026 seismic-imaging project, showing at least one industrial workflow progressing beyond research-only framing. SV014
CV016 Quantum Origin is marketed as a software QRNG that can be deployed without additional hardware or a cloud connection. SV009, SV015
CV017 Quantinuum said Quantum Origin became the first software QRNG to receive NIST validation in April 2025. SV015, SV009
CV018 Quantinuum’s public materials consistently place operations across the United States, United Kingdom, Germany, Japan, and Singapore. SV008, SV013, SV014
CV019 Quantinuum’s public headcount references rose from almost 500 in early 2024 to roughly 550 in 2025, 630 in late 2025, and about 700 in May 2026, indicating rapid scaling but inconsistent disclosure baselines. SV001, SV010, SV011, SV013, SV014
CV020 IonQ traded at roughly $23.75 billion market capitalization on May 26, 2026 according to Stock Analysis. SV017
CV021 IonQ reported $130.02 million of FY2025 revenue and a $510.38 million FY2025 net loss. SV018, SV025
CV022 IonQ’s quoted market capitalization equaled about 126.9 times trailing-twelve-month revenue on May 26, 2026. SV017, SV018
CV023 Rigetti traded at roughly $8.78 billion market capitalization on May 26, 2026 according to Stock Analysis. SV019
CV024 Rigetti reported $7.09 million of FY2025 revenue and a $216.21 million FY2025 net loss. SV020, SV026
CV025 Rigetti’s quoted market capitalization equaled about 876.2 times trailing-twelve-month revenue on May 26, 2026. SV019, SV020
CV026 D-Wave traded at roughly $10.88 billion market capitalization on May 26, 2026 according to Stock Analysis. SV021
CV027 D-Wave reported $24.59 million of FY2025 revenue and a $355.06 million FY2025 net loss. SV022, SV027
CV028 D-Wave’s quoted market capitalization equaled about 874.6 times trailing-twelve-month revenue on May 26, 2026. SV021, SV022
CV029 Quantum Computing Inc. traded at roughly $2.78 billion market capitalization on May 26, 2026 according to Stock Analysis. SV023
CV030 Quantum Computing Inc. reported only $0.68 million of FY2025 revenue and a $18.67 million FY2025 net loss. SV024, SV028
CV031 Quantum Computing Inc.’s quoted market capitalization equaled about 642.0 times trailing-twelve-month revenue on May 26, 2026. SV023, SV024
CV032 The quoted public quantum peer set remains overwhelmingly loss-making despite multi-billion-dollar market capitalizations. SV018, SV020, SV022, SV024, SV026, SV027, SV028
CV033 Against the public peer cluster, Quantinuum’s $10 billion private mark sits below IonQ, near D-Wave, above Rigetti, and well above Quantum Computing Inc. SV002, SV017, SV019, SV021, SV023
CV034 Because reviewed public sources still do not disclose Quantinuum revenue, gross margin, burn, or customer concentration, public evidence does not yet fundamentally prove that the $10 billion mark is cheap. SV003, SV004, SV006, SV032
CV035 The reviewed IPO and financing materials do not disclose price range, share count, liquidation preferences, secondary mix, or Honeywell sell-down terms, limiting visibility into dilution and overhang. SV003, SV004, SV029
CV036 Quantinuum’s investor base now spans finance, pharma, industrials, AI accelerators, and Asian distribution channels, giving it a broader strategic syndicate than most listed quantum peers. SV001, SV002, SV007
CV037 Named public customer and partner proof spans finance, pharma, autos, energy, cybersecurity, and government-facing use cases. SV011, SV013, SV014, SV015
CV038 24/7 Wall St. wrote that QTUM was up 73% over one year and 23% over one month, while warning that pure-play quantum names can swing 20% on a press release. SV030
CV039 Hot sector sentiment can help an IPO window, but it also raises the risk of fast multiple compression if commercialization milestones disappoint. SV030, SV021, SV023
CV040 The Motley Fool described Quantinuum’s September 2025 financing as valuing the company at roughly $10.6 billion after the raise, consistent with a $10 billion pre-money and $600 million primary round. SV031, SV002
CV041 At the last known private mark, upside is difficult to underwrite unless the eventual S-1 shows materially stronger revenue and margin than the current public record reveals. SV004, SV017, SV018, SV021, SV022, SV023, SV024
CV042 A reasonable bull case is roughly $16-22 billion if the S-1 reveals meaningful commercial scale, Helios converts marquee partners into expanding revenue, and the hot quantum IPO window stays open. SV011, SV013, SV030, SV017, SV018
CV043 A reasonable base case is roughly $8-12 billion if the S-1 shows progress but limited financial depth and public quantum sentiment remains volatile rather than euphoric. SV004, SV030, SV021, SV022
CV044 A reasonable bear case is roughly $3-6 billion if disclosed revenue is small, Honeywell overhang or financing terms are investor-unfriendly, or public quantum multiples fall sharply from spring-2026 levels. SV004, SV017, SV021, SV023, SV030
CV045 The probability-weighted payoff profile around a $10 billion entry is too flat and downside-heavy to justify an aggressive buy recommendation before fuller disclosure, so the prudent stance is track. SV004, SV017, SV018, SV021, SV022, SV023, SV024, SV030
CV046 The most decision-relevant diligence asks are S-1 financial disclosure, final 2025 financing size and terms, Honeywell ownership sell-down mechanics, and proof that Helios and Quantum Origin convert into recurring revenue. SV004, SV011, SV015, SV029
CV047 Quantinuum has a longer technical milestone record than its 2025-2026 fundraising headlines imply, including a 2022 announcement that logical qubits had begun outperforming physical qubits in a fault-tolerant circuit. SV012, SV010
CV048 Even supportive outside commentary still frames Quantinuum mainly as optional exposure through Honeywell or other strategic investors, reinforcing how thin standalone public financial disclosure remains. SV029, SV031
来源
编号出版方标题引文
SO001 Quantinuum Quantinuum Homepage Our best-in-class quantum computers, software, and developer toolkits are designed to work seamlessly.
SO002 Quantinuum Quantinuum | About Quantinuum’s headquarters is in Broomfield, Colorado, with additional facilities across the United States, United Kingdom, Germany, Japan, and Singapore.
SO003 Quantinuum Our Trapped Ion Quantum Computers Purchase a subscription directly with Quantinuum to access our trapped-ion quantum computers.
SO004 Quantinuum Quantinuum Documentation - Systems The world's most accurate quantum computer.
SO005 Quantinuum Quantum Origin | Quantum Random Number Generator Deploy anywhere, with no extra hardware, no cloud connection, and proven security.
SO006 Quantinuum InQuanto | Quantum Computational Chemistry InQuanto is our state-of-the-art quantum chemistry platform for complex molecular and materials simulations.
SO007 Quantinuum Developer Tools | TKET | Qermit | Lambeq Tackle the challenge of implementing quantum algorithms with TKET, our open-source quantum computing toolkit.
SO008 Quantinuum Introducing Quantinuum: The World's Largest Integrated Quantum Computing Company Honeywell is the largest shareholder of Quantinuum with an approximately 54 percent ownership stake in the new company.
SO009 Honeywell / Quantinuum Honeywell Announces the Closing of $300 Million Equity Investment Round for Quantinuum at $5B pre-money valuation The round is anchored by Quantinuum’s strategic partner JPMorgan Chase, with additional participation from Mitsui & Co., Amgen and Honeywell.
SO010 Honeywell / Quantinuum Honeywell Announces $600 Million Capital Raise For Quantinuum at $10b Pre-Money Equity Valuation to Advance Quantum Computing at Scale Quanta Computer, NVentures (NVIDIA’s venture capital arm) and QED Investors have joined existing shareholders JPMorganChase, Mitsui, Amgen...
SO011 Quantinuum Quantinuum Names Rajeeb (Raj) Hazra Chief Executive Officer Current Quantinuum CEO Ilyas Khan will remain a member of the board of directors and will now become Vice Chairman.
SO012 Quantinuum Nitesh Sharan Joins Quantinuum as Chief Financial Officer The company has a global workforce of approximately 700 employees.
SO013 Quantinuum Quantinuum Announces Commercial Launch of New Helios Quantum Computer Helios launches with customers Amgen, BMW Group, JPMorganChase, and SoftBank Corp.
SO014 Quantinuum Quantinuum Launches Industry-First, Trapped-Ion 56-Qubit Quantum Computer Quantinuum today unveiled the industry’s first quantum computer with 56 trapped-ion qubits.
SO015 Quantinuum Quantinuum Announces Updates to Quantum Natural Language Processing Toolkit λambeq λambeq is the first quantum NLP and computational linguistics toolkit.
SO016 Quantinuum Quantinuum selected as a founding collaborator for NVIDIA Accelerated Quantum Research Center Quantinuum will work with NVIDIA as a founding collaborator on breakthroughs at the NVIDIA Accelerated Quantum Research Center.
SO017 Quantinuum EAGLYS, Mitsui and Quantinuum Collaborate to Build a Quantum-Resistant Data Analytics (AI) Platform EAGLYS has now integrated Quantum Origin... strengthening the platform against the quantum threat to encrypted data.
SO018 Quantinuum Quantum Origin Becomes First Software QRNG to Achieve NIST Validation Quantum Origin is the first software QRNG to achieve this validation.
SO019 Quantinuum Quantinuum Introduces InQuanto InQuanto is available for the first time as a standalone platform to commercial organizations.
SO020 Quantinuum Quantinuum Unveils Accelerated Roadmap to Achieve Universal, Fully Fault-Tolerant Quantum Computing by 2030 Quantinuum unveiled its roadmap to universal, fully fault-tolerant quantum computing by 2030.
SO021 Honeywell Honeywell Announces $600 Million Capital Raise for Quantinuum at $10B Pre-Money Equity Valuation Quantinuum's over 630 employees... across the US, UK, Germany, and Japan, are driving the quantum computing revolution.
SO022 Honeywell Honeywell Announces Quantinuum's Filing of Registration Statement for Proposed Initial Public Offering Quantinuum intends to list its Class A common stock on the Nasdaq Global Select Market under the ticker symbol "QNT."
SO023 Reuters Honeywell's Quantinuum raises funds from Nvidia, others at $10 billion valuation Honeywell said on Thursday its quantum computing company, Quantinuum, had raised about $600 million... at a valuation of $10 billion.
SO024 Reuters Honeywell's Quantinuum confidentially files for US IPO The firm, majority owned by Honeywell, had submitted papers with the U.S. Securities and Exchange Commission in February.
SO025 Reuters Honeywell's Quantinuum moves closer to public markets with US IPO filing The Broomfield, Colorado-based company reported a net loss of $192.6 million on revenue of $30.9 million in the year ended December 31.
SO026 Mitsui & Co., Ltd. Mitsui to invest in Quantinuum and sign a distributorship agreement for marketing in Japan and Asia-Pacific Mitsui made a US$50M investment in Quantinuum... This investment is part of Quantinuum's US$300M equity fundraise.
SO027 Microsoft Azure Quantum Blog Microsoft and Quantinuum create 12 logical qubits and demonstrate a hybrid, end-to-end chemistry simulation Microsoft and Quantinuum applied Azure Quantum’s qubit-virtualization system to Quantinuum’s H2 trapped-ion quantum computer to create 12 highly reliable logical qubits.
SO028 Microsoft Learn Quantinuum provider - Azure Quantum Quantinuum provides access to trapped-ion systems with high-fidelity, fully connected qubits, and the ability to perform mid-circuit measurement.
SO029 JPMorganChase JPMorganChase, Quantinuum, Argonne National Laboratory, Oak Ridge National Laboratory and University of Texas at Austin advance the application of quantum computing The 56-qubit Quantinuum System Model H2 trapped-ion quantum computer... was used for this study.
SO030 U.S. Securities and Exchange Commission Quantinuum Inc. Form S-1 Registration Statement Our net revenue and net loss for the year ended December 31, 2025 was $30.9 million and $192.6 million, respectively.
SO031 HPCwire Honeywell Confirms Quantinuum IPO Filing as Quantum Firms Face Market Scrutiny Analysts view that distinction as notable... as it faces greater public market scrutiny.
SO032 The Quantum Insider Honeywell Announces Quantinuum's Confidential Submission of Draft Registration Statement For Proposed IPO The company develops quantum computing hardware, software and cryptographic technologies... in a sector still largely defined by long-term research investment rather than near-term profitability.
SO033 Constellation Research Quantinuum’s IPO: What you need to know Quantinuum's financials look like the rest of the quantum computing field... The financials are lumpy.
SO034 Tech Funding News Honeywell-backed Quantinuum files for US IPO at up to $20B valuation Despite investor interest, the company remains loss-making.
SO035 TradingKey Quantum Computing Welcomes Largest IPO. Quantinuum Files IPO Documents Target Valuation 20 Billion Based on its $20 billion IPO valuation target, Quantinuum's price-to-sales ratio exceeds 640x.
SM001 Quantinuum Quantum Origin | Quantum Random Number Generator
SM002 Quantinuum InQuanto | Quantum Computational Chemistry
SM003 Quantinuum Developer Tools | TKET | Qermit | Lambeq
SM004 Quantinuum Quantinuum Nexus | Access the world’s most powerful quantum computing stack
SM005 Quantinuum HSBC and Quantinuum Explore Real World Use Cases of Quantum Computing in Financial Services
SM006 Quantinuum JPMorgan Chase, Argonne National Laboratory and Quantinuum Show Theoretical Quantum Speedup with the Quantum Approximate Optimization Algorithm
SM007 Quantinuum Mitsui, QSimulate, and Quantinuum Launch “QIDO”: A Quantum-Integrated Chemistry Platform Targeting Faster Drug and Materials Discovery
SM008 Quantinuum Quantinuum Introduces InQuanto to Explore Industrially Relevant Chemistry Problems on Today’s Quantum Computers
SM009 Quantinuum Our Trapped Ion Quantum Computers | System Model H1
SM010 Quantinuum Our Trapped Ion Quantum Computers | System Model H2
SM011 Quantinuum Quantinuum System Model H1 Product Data Sheet
SM012 Quantinuum Quantinuum System Model H2 Product Data Sheet
SM013 QED-C State of the Global Quantum Industry 2026 Key findings list a 2025 market size of $1.9B, 556 pure-play companies, and 8,261 new quantum-related position openings.
SM014 Business Wire / ResearchAndMarkets Quantum Computing Market Report 2026-2046: Quantum Funding Rounds Surpass $50M Average as Commercial Viability Accelerates
SM015 The Business Research Company Global Quantum Computing Market Report 2026
SM016 U.S. Bureau of Industry and Security Department of Commerce Implements Controls on Quantum Computing and Other Advanced Technologies
SM017 Global Legal Insights Regulatory Regimes and Quantum Computing
SM018 NIST Post-Quantum Cryptography | CSRC
SM019 National Quantum Coordination Office National Quantum Coordination Office (NQCO)
SM020 U.S. DOE Office of Science Quantum Information Science (QIS)
SM021 Lawrence Berkeley National Laboratory At the Leading Edge of the Quantum Frontier
SM023 PostQuantum The Border Around Quantum: Export Controls, Deemed Exports, and “Research as a Controlled Flow”
SM024 Amazon Web Services Cloud Quantum Computing Service - Amazon Braket
SM025 Microsoft Azure Azure Quantum Computing
SM026 IBM IBM Quantum Computing | Home
SM027 Airbus Quantum technologies
SM028 QED-C Reports | QED-C
SM029 NIST Department of Commerce Announces Letters of Intent With 9 Companies for $2 Billion to Accelerate U.S. Leadership in Quantum Computing
SP001 Quantinuum Quantinuum | Accelerating Quantum Computing
SP002 Quantinuum Our Trapped Ion Quantum Computers
SP003 Quantinuum Helios | Quantinuum's Quantum Computers
SP004 Quantinuum Quantinuum Documentation
SP005 IBM IBM Quantum Computing | Products and services
SP006 IBM IBM Quantum Computing | Hardware and roadmap
SP007 IBM Introduction | IBM Quantum Documentation
SP008 Google Quantum AI Quantum Computer | Google Quantum AI
SP009 Google Quantum AI Willow Early Access Program | Google Quantum AI
SP010 Google Quantum AI Cirq | Google Quantum AI
SP011 IonQ Quantum Cloud Services - IonQ Quantum Cloud
SP012 IonQ IonQ | Roadmap
SP013 IonQ IonQ Forte Enterprise: Quantum Computer for Data Centers
SP014 IonQ IonQ Posts Q1 2026 Earnings with Record Revenue
SP015 Rigetti Computing Building scalable, innovative quantum systems
SP016 Rigetti Computing Novera
SP017 IQM Quantum Computers IQM Quantum Computers - Built for Real Impact
SP018 Amazon Web Services Quantum computers | Amazon Braket
SP019 QuEra Aquila | 256-qubit Quantum Computer
SP020 QuEra Our Quantum Roadmap
SP021 Oxford Quantum Circuits OQC – Enterprise ready quantum solutions
SP022 Amazon Web Services Amazon Braket Pricing
SP023 D-Wave The Advantage2 Quantum Computer | D-Wave
SP024 D-Wave The Leap Quantum Cloud Service | D-Wave
SP025 PsiQuantum Technology — PsiQuantum
SP026 Nature A manufacturable platform for photonic quantum computing
SP027 IEEE Spectrum Neutral Atom Quantum Computing: 2026's Big Leap If someone says quantum computers are commercially useful today, I say I want to have what they're having.
SI001 Securities and Exchange Commission Quantinuum Inc. Form S-1 Our net revenue and net loss for the year ended December 31, 2025 was $30.9 million and $192.6 million, respectively.
SI002 Securities and Exchange Commission Honeywell International Inc. Form 10-Q Corporate and All Other includes revenue from Honeywell's majority-owned investment in Quantinuum.
SI003 Securities and Exchange Commission Honeywell International Inc. Form 10-K The backlog within Corporate and All Other relates to the Quantinuum business.
SI004 Honeywell Honeywell Announces the Closing of $300 Million Equity Investment Round for Quantinuum at $5B Pre-Money Valuation Honeywell announced the closing of a $300 million equity fundraise for Quantinuum ... at a pre-money valuation of $5 billion.
SI005 Honeywell Honeywell Announces $600 Million Capital Raise for Quantinuum at $10B Pre-Money Equity Valuation Honeywell today announced an approximately $600 million equity capital raise for Quantinuum ... at a pre-money equity valuation of $10 billion.
SI006 Honeywell Honeywell Announces Quantinuum’s Filing of Registration Statement for Proposed Initial Public Offering Honeywell ... announced that Quantinuum ... has publicly filed a registration statement on Form S-1.
SI007 Quantinuum About Quantinuum Jenni is responsible for the profit and loss (P&L) ... of our compute products which include InQuanto, Quantum Systems (both cloud and HaaS) and Quantinuum Nexus.
SI008 Quantinuum Quantinuum Systems Purchase a subscription directly with Quantinuum to access our trapped-ion quantum computers.
SI009 Quantinuum Quantinuum Helios The Helios platform is now available to customers through Quantinuum’s cloud service and on-premises offering.
SI010 Quantinuum InQuanto Immerse your team with our experts via our Partner Residency Program ... Access to Quantinuum systems.
SI011 Quantinuum Quantinuum Nexus Quantinuum Nexus is our cloud-based platform that enables users to seamlessly run, review, and collaborate on quantum computing projects.
SI012 Quantinuum Introducing Quantinuum Nexus: Our All-in-one Quantum Computing Platform Nexus allows users to run, track, and manage resources across multiple quantum backends.
SI013 Quantinuum Quantum Origin Protect your most critical data with proven quantum randomness from Quantum Origin. Deploy anywhere, with no extra hardware, no cloud connection.
SI014 Quantinuum Documentation How Quantum Origin Works Quantum Origin leverages ... a software binary in combination with a pre-generated output from Quantinuum’s quantum computer.
SI015 Microsoft Learn Quantinuum provider - Azure Quantum System Model H2 Emulator usage is offered free-of-charge with a hardware subscription.
SI016 Microsoft Learn Pricing plans for Azure Quantum providers Standard Plan: USD125,000/Month ... Premium Plan: USD175,000/Month.
SI017 Amazon Web Services Amazon Braket Pricing Amazon Braket offers three pricing components for on-demand use of a quantum computer (QPU): a per-shot fee and a per-task fee or a single hourly reservation fee.
SI018 Reuters via U.S. News & World Report Honeywell’s Quantinuum Confidentially Files for US IPO The firm, majority owned by Honeywell ... Quantinuum was valued at $10 billion in a fundraising round in September.
SI019 HPCwire Honeywell Confirms Quantinuum IPO Filing as Quantum Firms Face Market Scrutiny Analysts view that distinction as notable ... as it faces greater public market scrutiny.
SI020 Constellation Research Quantinuum’s IPO: What you need to know Quantinuum's meager revenue looks similar to other pure plays in quantum computing. The financials are lumpy.
SI021 The Next Web Quantinuum filed for an IPO worth 20 billion dollars. It has 31 million in revenue and a quantum computer that does not exist yet. The company is asking public market investors to pay a premium of more than 600 times revenue for a quantum computer that does not yet exist in its final form.
SI022 Quantum Computing Report Quantinuum Has Submitted a Preliminary S-1 Filing to the U.S. SEC for a Proposed Initial Public Offering For the fiscal year ended December 31, 2025, Quantinuum reported $30.9 million in net revenue and a net loss of $192.6 million.
SI023 The Quantum Insider Honeywell Announces Quantinuum’s Filing of Registration Statement for IPO Honeywell announced that Quantinuum has publicly filed an S-1 registration statement with the U.S. Securities and Exchange Commission.
SI024 Quantinuum Quantinuum’s Reimei Quantum Computer Now Fully Operational at RIKEN This state-of-the-art machine is now fully operational ... Japanese researchers now have direct, on-site access to Quantinuum’s cutting-edge quantum computing technology.
SI025 Quantinuum Quantinuum and BMW Group Expand Landmark Quantum Computing Collaboration with New Multi-Year Partnership BMW Group will leverage successive generations of Quantinuum’s quantum computers. This includes the current Helios system and upcoming generations, Sol and Apollo.
SI026 Quantinuum Quantinuum Enters into Letter of Intent with the U.S. Department of Commerce for Funding Opportunity The letter of intent proposes that Quantinuum would receive federal funding to enable the development of large-scale, fault-tolerant trapped-ion quantum computers.
SE001 Quantinuum Quantinuum | Accelerating Quantum Computing
SE002 Quantinuum Our Trapped Ion Quantum Computers | System Model H1
SE003 Quantinuum Our Trapped Ion Quantum Computers | System Model H2
SE004 Quantinuum Helios | Quantinuum's Quantum Computers
SE005 Quantinuum Developer Tools | TKET | Qermit | Lambeq
SE006 Quantinuum InQuanto | Quantum Computational Chemistry
SE007 Quantinuum Quantum Origin | Quantum Random Number Generator
SE008 Quantinuum Quantinuum Powering Hybrid Quantum AI Supercomputing with NVIDIA
SE009 Quantinuum Quantinuum Documentation
SE010 Quantinuum Quantinuum System Model H1 Product Data Sheet
SE011 Quantinuum Quantinuum System Model H2 Product Data Sheet
SE012 Quantinuum Quantinuum Helios Product Data Sheet
SE013 Quantinuum Quantinuum Documentation — TKET
SE014 Quantinuum Quantinuum Documentation — InQuanto
SE015 Quantinuum Quantinuum Documentation — lambeq
SE016 Microsoft Learn Quantinuum provider - Azure Quantum
SE017 Oak Ridge Leadership Computing Facility Quantinuum — OLCF User Documentation
SE018 Quantinuum GitHub - Quantinuum/tket
SE019 Quantinuum Releases · Quantinuum/tket
SE020 Quantinuum GitHub - Quantinuum/lambeq
SE021 Quantinuum Releases · Quantinuum/lambeq
SE022 PyPI pytket
SE023 PyPI lambeq
SE024 Pepy pytket · 4.5M downloads on PyPI
SE025 Pepy lambeq · 96.9k downloads on PyPI
SE026 NVIDIA NVIDIA Partners Accelerate Quantum Breakthroughs with AI Supercomputing
SE027 Data Center Dynamics Quantinuum launches Helios for general-purpose hybrid quantum workloads
SE028 PR Newswire Quantinuum Partners with Microsoft in New Phase of Reliable Quantum Computing with Breakthrough Demonstration of Reliable Logical Qubits
SE029 arXiv lambeq: An Efficient High-Level Python Library for Quantum NLP
SU001 Quantinuum Quantinuum Announces Commercial Launch of New Helios Quantum Computer That Offers Unprecedented Accuracy to Enable Generative Quantum AI
SU002 Constellation Research Quantinuum launches Helios quantum computer, touts fidelity, enterprise customers
SU003 Quantinuum Helios | Quantinuum's Quantum Computers
SU004 Quantinuum Introducing Helios: The Most Accurate Quantum Computer in the World
SU005 Quantinuum Singapore's National Quantum Office and Quantinuum Forge Strategic Partnership to Accelerate Quantum Computing
SU006 Quantinuum Quantinuum Expands Global Footprint to Singapore with the Establishment of a New R&D Centre
SU007 Quantinuum Q-Net
SU008 Quantinuum Celebrating Our First Annual Q-Net Connect!
SU009 Quantinuum Quantinuum and BMW Group Expand Landmark Quantum Computing Collaboration with New Multi-Year Partnership
SU010 BMW Group BMW Group, Airbus and Quantinuum Collaborate to Fast-Track Sustainable Mobility Research Using Cutting-Edge Quantum Computers
SU011 Airbus BMW Group, Airbus and Quantinuum Collaborate to Fast-Track Sustainable Mobility Research Using Cutting-Edge Quantum Computers
SU012 The Quantum Insider Quantinuum and BMW Group Expand Quantum Computing Collaboration with New Multi-Year Partnership
SU013 Business Wire JPMorganChase, Quantinuum, Argonne National Laboratory, Oak Ridge National Laboratory and University of Texas at Austin Advance the Application of Quantum Computing to Potential Real-World Use Cases Beyond the Capabilities of Classical Computing
SU014 JPMorganChase Harnessing Quantum Computing for Certified Randomness
SU015 Quantinuum HSBC uses our Quantum Origin QRNG to pilot quantum-safe technology for tokenized gold
SU016 TechMonitor HSBC partners with Quantinuum on quantum banking
SU017 Quantinuum Quantinuum Partners with Microsoft in New Phase of Reliable Quantum Computing with Breakthrough Demonstration of Reliable Logical Qubits
SU018 Data Center Dynamics Quantinuum updates product roadmap; announces two industry firsts in collaboration with Microsoft
SU019 Securities and Exchange Commission Quantinuum, Inc. Form S-1 (quantinuum-sx1.htm)
SU020 Quantinuum Quantinuum’s Quantum Origin Becomes First Software Quantum Random Number Generator to Achieve NIST Validation
SU021 Quantinuum Mitsui, QSimulate, and Quantinuum Launch QIDO, a Quantum-Integrated Chemistry Platform Targeting Faster Drug and Materials Discovery
SU022 Quantinuum RIKEN Scales Quantum-Supercomputing in Japan with Quantinuum System Upgrade
SU023 Quantinuum Quantinuum Announces Collaboration with Synopsys Toward Advancing Industrial Design with Quantum Computing
SU024 Fierce Biotech Quantum firm hailed by Amgen tech leader as extraordinary platform plans IPO
SU025 Automotive World BMW expands quantum computing tie with Quantinuum
SR001 Honeywell Honeywell Announces Quantinuum's Filing of Registration Statement for Proposed Initial Public Offering
SR002 Honeywell Honeywell Announces $600 Million Capital Raise For Quantinuum at $10b Pre-Money Equity Valuation to Advance Quantum Computing at Scale
SR003 Honeywell Honeywell Announces Quantinuum's Confidential Submission of Draft Registration Statement for Proposed Initial Public Offering
SR004 Honeywell Honeywell Announces Quantinuum's Filing of Registration Statement for Proposed Initial Public Offering
SR005 Quantinuum Quantinuum | About
SR006 Quantinuum Quantinuum Names Rajeeb (Raj) Hazra Chief Executive Officer
SR007 Quantinuum Quantinuum Enters into Letter of Intent with the U.S. Department of Commerce for Funding Opportunity to Accelerate U.S. Leadership in Quantum Computing
SR008 Quantinuum Quantinuum Announces Commercial Launch of New Helios Quantum Computer that Offers Unprecedented Accuracy to Enable Generative Quantum AI (GenQAI)
SR009 Quantinuum Quantinuum Unveils Accelerated Roadmap to Achieve Universal, Fully Fault-Tolerant Quantum Computing by 2030
SR010 Quantinuum Technical perspective: By the end of the decade, we will deliver universal, fully fault-tolerant quantum computing
SR011 Quantinuum Real Time Error Correction at Increased Scale
SR012 Quantinuum / arXiv Helios: A 98-qubit trapped-ion quantum computer
SR013 Quantinuum Quantinuum and BMW Group Expand Landmark Quantum Computing Collaboration with New Multi-Year Partnership
SR014 Quantinuum Quantinuum's “Reimei” Quantum Computer Now Fully Operational at RIKEN, Ushering in a New Era of Hybrid Quantum High-Performance Computing
SR015 Quantinuum Quantinuum selected as a founding collaborator for NVIDIA Accelerated Quantum Research Center
SR016 Quantinuum Quantinuum Partners with Microsoft in New Phase of Reliable Quantum Computing with Breakthrough Demonstration of Reliable Logical Qubits
SR017 SoftBank Corp. SoftBank Corp. and Quantinuum Announce Groundbreaking Partnership Toward Practical Application of Quantum Computing
SR018 Bureau of Industry and Security Department of Commerce Implements Controls on Quantum Computing and Other Advanced Technologies Alongside International Partners
SR019 U.S. National Quantum Coordination Office Department of Commerce Releases Export Controls on Quantum Technologies
SR020 Kelley Drye & Warren LLP BIS Publishes New Controls on Quantum Computing and Other Advanced Technologies in Accordance with International Partners
SR021 Phillips Lytle LLP U.S. Commerce Dept Seeks New Quantum Tech Export Controls
SR022 Global Legal Insights Regulatory regimes and quantum computing
SR023 Global Legal Insights Preparing for the quantum future: challenges and solutions
SR024 APS News US Puts Export Controls on Quantum Computers
SR025 MIT Technology Review A new ion-based quantum computer makes error correction simpler
SR026 HPCwire Honeywell Confirms Quantinuum IPO Filing as Quantum Firms Face Market Scrutiny
SR027 Reuters via U.S. News & World Report Honeywell's Quantinuum Confidentially Files for US IPO
SR028 IBM IBM Quantum Computing | Hardware and roadmap
SR029 Google Meet Willow, our state-of-the-art quantum chip
SR030 IonQ IonQ Forte Enterprise: Quantum Computer for Data Centers
SR031 QuEra Building Quantum Computers with Neutral Atoms | QuEra
SR032 Quantinuum Helios | Quantinuum's Quantum Computers
SV001 Quantinuum Honeywell Announces the Closing of $300 Million Equity Investment Round for Quantinuum at $5B pre-money valuation Honeywell today announced the closing of a $300 million equity fundraise for Quantinuum ... at a pre-money valuation of $5 billion.
SV002 Quantinuum Honeywell Announces $600 Million Capital Raise For Quantinuum at $10b Pre-Money Equity Valuation to Advance Quantum Computing at Scale Honeywell today announced an approximately $600 million equity capital raise for Quantinuum ... at a pre-money equity valuation of $10 billion.
SV003 Quantinuum Honeywell Announces Quantinuum’s Confidential Submission of Draft Registration Statement for Proposed Initial Public Offering Quantinuum ... confidentially submitted a draft registration statement on Form S-1 ... on February 17, 2026.
SV004 Quantinuum Honeywell Announces Quantinuum’s Filing of Registration Statement for Proposed Initial Public Offering The number of shares to be offered and the price range for the proposed offering have not yet been determined.
SV005 Reuters Honeywell's quantum computing firm valued at $5 billion after latest fundraise
SV006 Reuters Honeywell's Quantinuum confidentially files for US IPO
SV007 Tech Funding News UK-based Quantinuum closes $600M at $10B valuation to build next-gen quantum computer — TFN
SV008 Quantinuum Quantinuum | About
SV009 Quantinuum Quantum Origin | Quantum Random Number Generator
SV010 Quantinuum Quantinuum selected as a founding collaborator for NVIDIA Accelerated Quantum Research Center
SV011 Quantinuum Quantinuum Announces Commercial Launch of New Helios Quantum Computer that Offers Unprecedented Accuracy to Enable Generative Quantum AI (GenQAI)
SV012 Quantinuum Logical qubits start outperforming physical qubits
SV013 Quantinuum Quantinuum and BMW Group Expand Landmark Quantum Computing Collaboration with New Multi-Year Partnership
SV014 Quantinuum Quantinuum and bp Collaborate Towards Solving Fundamental Wave Physics Challenges with Quantum Computing
SV015 Quantinuum Quantinuum’s ‘Quantum Origin’ Becomes First Software Quantum Random Number Generator to Achieve NIST Validation
SV016 Honeywell Quantum
SV017 Stock Analysis IonQ, Inc. (IONQ) Stock Price & Overview
SV018 Stock Analysis IonQ, Inc. (IONQ) Financials & Income Statement
SV019 Stock Analysis Rigetti Computing (RGTI) Stock Price & Overview
SV020 Stock Analysis Rigetti Computing (RGTI) Financials & Income Statement
SV021 Stock Analysis D-Wave Quantum (QBTS) Stock Price & Overview
SV022 Stock Analysis D-Wave Quantum (QBTS) Financials & Income Statement
SV023 Stock Analysis Quantum Computing (QUBT) Stock Price & Overview
SV024 Stock Analysis Quantum Computing (QUBT) Financials & Income Statement
SV025 U.S. Securities and Exchange Commission IonQ, Inc. 2025 Form 10-K
SV026 U.S. Securities and Exchange Commission Rigetti Computing, Inc. 2025 Form 10-K
SV027 U.S. Securities and Exchange Commission D-Wave Quantum Inc. 2025 Form 10-K
SV028 U.S. Securities and Exchange Commission Quantum Computing Inc. 2025 Form 10-K
SV029 Yahoo Finance Quantinuum IPO Plan and Dividend Move Could Be A Game Changer For Honeywell International (HON)
SV030 24/7 Wall St. Quantinuum's IPO Could Reshape QTUM in 2026. Here's What to Watch
SV031 The Motley Fool A Once-in-a-Decade Opportunity: 10 Billion Reasons to Pay Attention to This Monster Quantum Computing Company (Hint: Not IonQ) | The Motley Fool
SV032 U.S. News & World Report Honeywell's Quantinuum Confidentially Files for US IPO