Neo Fusion
获得战略资本支持的核聚变开发商,账面资本异常庞大,但仍缺少清晰市场定价,也没有客户背书的商业化证明。
Neo Fusion 在中国聚变路线图里具备战略重要性,但披露偏薄、商业化未验证、里程碑风险高,结论仍应停在继续研究, 置信度为中,估值立场未知。
封面要素
公司概况
Neo Fusion 是一家中国民营核聚变开发商,成立于 2023 年 5 月,核心任务是在合肥建设 BEST 紧凑型高场托卡马克。公开资料把 BEST 描述为燃烧等离子体、氘氚平台,目标是在 EAST 与未来中国聚变工程示范堆之间架桥,同时冲刺 2027 年首次等离子体里程碑和 2030 年前后的发电验证。公司当下的「客户」更应理解为战略出资方和未来电力侧利益方,而不是正在买电的客户。Neo Fusion 的融资故事账面规模异常大,但披露仍然偏薄,外部投资人还无法扎实核验实缴资本、电站经济性或客户变现。
- 成立时间
- 2023-05-19
- 创立地点
- Hefei, Anhui, China
- 总部
- Hefei, Anhui, China
- 产品
- Neo Fusion 在建设 BEST,这是一台紧凑型高场超导托卡马克,用来研究燃烧氘氚等离子体、测试材料和包层系统,并最终支撑未来聚变发电。
- 客户
- 当前利益相关方是国家与战略出资方;等国家支持的首台装置为技术降险后,未来客户最可能是中国电力公司、电网运营商,以及大型工业或数字负载买家。
- 商业模式
- 先靠出资方支持的资本建设并运行 BEST,把它作为里程碑桥梁;如果商业化成功,再通过未来聚变电站、场址合作伙伴关系,以及长期电力或基础设施合同变现。
- 阶段
- pre-commercial
- 融资情况
- 公开资料显示,2023 年 5 月 Neo Fusion 有一笔 US$723.37 million 的初始注册资本,2024 年 7 月又新增 US$1.306 billion 注册资本,但仍不能证明已经实缴或投入了多少。
执行摘要
主要优势
- Neo Fusion 嵌在全球最深的公共聚变生态之一里,与 EAST、CRAFT、Science Island 和 Hefei Comprehensive National Science Centre 相连。
- 以聚变初创公司看,公司公开资本故事异常大:2024 年增资后,名义注册资本约 US$2.03 billion。
- BEST 被包装成具体反应堆项目:目标 2027 年首次等离子体、2030 年代前后做发电示范,而不是一个模糊研究概念。
- CNPC、HFIPS 相关资本、安徽国资实体、NIO 等国有和战略支持方,比典型纯 VC 股权结构更能提供 sponsor 韧性。
- 如果 BEST 能把实验托卡马克积累接到可融资的聚变电力基础设施上,长期期权价值可能相当可观。
主要风险
- 核心逻辑仍依赖 Neo Fusion 尚未交付的燃烧等离子体、进度和工程里程碑。
- 注册资本不等于定价估值,也不等于已经到账、可部署的现金;公开资料没有披露实缴资本桥。
- Neo Fusion 没有公开 PPA、锚定场址协议、电价模型,或可与 CFS、Helion 等包装更成熟同行相比的客户经济性。
- 聚变仍然极其资本密集,而替代电源还在成本和部署速度上继续进步。
- 监管、安全、氚和电站级商业化细节仍明显披露不足。
未决问题
- 一份股权结构表和实缴资本时间表,说明名义注册资本中有多少是真正可部署的现金跑道。
- 一份整合里程碑预算,把当前状态接到 2027 年首次等离子体,再接到 2030 年代前后的示范叙事。
- Neo Fusion 专属的场址、购电、电价或电站经济性模型。
- BEST 的装置级安全、氚、废料和监管文件包。
- 足以判断外部资本真实入场经济性的治理、投资者权利和融资结构细节。
目录
01公司概况
1.1 身份定位、使命与成立理由
Neo Fusion 与其说是典型风投支持的初创公司,不如说是中国 BEST 托卡马克项目外层的商业化和工程化载体。公开资料一致将公司放在安徽合肥,并把其源头追溯到 2023 年 5 月 19 日:该实体成立,是为了把受控核聚变工作推向更接近产业建设的路径。它的一句话使命很直接:开发并最终商业化基于托卡马克的聚变能源系统。真正特殊的是资本基础的规模和构成。Neo Fusion 没有走传统种子轮到后续轮次的路径,而是带着国资关联注册资本起步;2024 年,随着 CNPC 和 Science Island 相关投资人加入,资本基础又迅速放大。因此,公司的身份与中国国家核聚变雄心、合肥综合性国家科学中心,以及 BEST 本身的实物建设进度绑在一起。[CO001, CO002, CO003, CO004, CO005, CO009]
| 指标 | 数值 / 状态 | 证据日期 | 置信度 | 限制说明 |
|---|---|---|---|---|
| 成立 | 2023-05-19 | 2023 / 2026 资料核查 | 高 | 不同来源对法人实体名称写法不一 |
| 总部 | 中国安徽合肥 | 2026 | 高 | 办公地址与登记地址到街道层面不一致 |
| 当前阶段 | 尚无收入的工程公司 | 2026 | 中 | 未披露收入 |
| 注册资本 | RMB 14.5B (~US$2.03B) | 2024 调整 / 2026 资料 | 高 | 不等同于有定价的 VC 估值 |
| 初始资本事件 | US$723.37M 注册资本 | 2023-05-19 | 中 | 来源是行业数据库,不是备案文件 |
| 追加资本事件 | US$1.306B 注册资本 | 2024-07-23 | 中 | 中文媒体以人民币口径报道了同一轮增资 |
| 主要战略支持方 | CNPC、HFIPS/Science Island、安徽国资、NIO 关联的 Weiju | 2024-2026 | 中 | 完整一手股权结构表未公开 |
| 旗舰资产 | BEST 紧凑型高场托卡马克 | 2025-2026 | 高 | 项目仍在建设中 |
| 首次等离子体目标 | 2027 | 2025-11 | 高 | 进度风险仍然重要 |
| 电力示范目标 | 约 2030 | 2026-01 | 高 | 取决于 D-T 燃烧等离子体能否跑通 |
| 公开客户 | 未披露 | 2026-07-19 复核 | 中 | 工程阶段公司 |
| 公开收入 | 未披露 | 2026-07-19 复核 | 中 | 无可用财务报表 |
注册资本数字来自行业数据库和中文商业媒体报道,不是有定价的风险融资备忘录。若干商业关键指标仍未披露。
[CO001, CO003, CO009, CO006, CO007, CO010]Neo Fusion 的可投资故事,压在短短几轮融资之后的一条 2025-2030 工程排期上。
Mid-2024 资本上调日期按约略处理,因为资料同时用备案时间和报道时间描述该注册变更。
[CO001, CO006, CO008, CO020, CO021, CO022]1.2 所有权结构、领导层与治理质量
所有权是 Neo Fusion 故事里最清晰的差异点。可获取的公开摘要显示,公司股权结构由国资背景资本主导,包括 CNPC 的 Kunlun Capital、Science Island 相关机构资本、合肥市级基金、安徽省能源资本,以及较小的 NIO 相关持股。这种组合给了 Neo Fusion 在核聚变初创公司中罕见的深层战略支持,但也意味着外部投资人拿不到西方硬科技融资轮里常见的标准化披露。Yan Jianwen 是最常见的具名领导者和公开发言人;Baidu Baike 将 Huang Suzhen 标为法定代表人,可访问媒体则引用副总经理 Liu Zhihong 对商业化模式的说法。除这些名字之外,公开治理细节仍很稀疏,领导层延续性和内部分权比融资堆栈本身更难尽调。该不透明度很关键,因为项目规模太大、离政策太近,不能只靠叙事承销;投资人需要看到更清楚的董事会监督、授权机制,以及谁负责把研究交接到电站执行。[CO010, CO011, CO012, CO013, CO014, CO015]
| 参与方 | 角色 | 已披露持股 / 出资 | 意义 |
|---|---|---|---|
| CNPC / Kunlun Capital | 国资背景战略投资者 | RMB 2.9B;36Kr 报道约 20% | 提供长期资本和能源行业政治背书 |
| Hefei Science Island / HFIPS 关联方 | 科研机构股东 | 公开摘要披露约 20% | 把公司直接接到中国领先的等离子体物理基地 |
| Hefei Chanto New Energy 基金 | 市级资本平台 | 36Kr 报道约 20.5% | 项目接入合肥产业政策和本地融资 |
| Anhui Wanneng 聚变基金 | 省级能源资本平台 | 36Kr 报道约 20.5% | 增加省级公用事业与能源系统协同 |
| NIO 关联的 Weiju Technology | 产业资本小股东 | Baike / 媒体摘要提到 5% | 证明非国资战略方也关注聚变期权价值 |
| 其他省级创新资本 | 剩余股东 | 可访问的一手文件未披露 | 股权结构表不止少数已点名方 |
本表刻意区分证据较充分的持有人和六股东结构中更不透明的剩余部分。2024 年之后的股权结构表,在可访问来源中只露出一部分。
[CO010, CO011, CO012, CO013, CO037]| 人物 / 角色 | 公开可见职能 | 主要证据 | 披露限制 |
|---|---|---|---|
| Yan Jianwen | 董事长兼主要公开发声人 | Baike、Anhui News 会议报道 | 董事长之外的运营职责披露不深 |
| Huang Suzhen | 法定代表人 / 运营高管 | Baidu Baike | 可访问英文来源未深入介绍其背景 |
| Liu Zhihong | 副总经理 / 商业化倡导者 | 36Kr 访谈摘录 | 该角色来自媒体,不是官方组织页面 |
| Song Yuntao | BEST 总工程师 / ASIPP 技术负责人 | Nuclear Engineering International、CAS 生态报道 | 技术项目负责人,是否为 Neo Fusion 高管并不清晰 |
| HFIPS / Science Island 机构 | 深度嵌入治理与技术影响 | Baike、CAS、Eurofusion、媒体 | 机构影响力比正式董事会结构更清楚 |
Neo Fusion 披露的管理团队名单比领先西方聚变初创公司更薄。这张表只保留出现足够一致、可引用的名字。
[CO014, CO015, CO016, CO018]公开披露显示,该联盟由国资和科研机构主导,另有一块较小的工业资本持仓与 NIO 相关。
若干持股比例来自中国商业媒体的二手梳理,而非可完整访问的一手股东文件。
[CO011, CO012, CO013, CO037]1.3 项目里程碑与合肥核聚变生态
支持 Neo Fusion 的最强证据不是客户名单或收入基础,而是周边技术生态。BEST 已从土建进入最终总装,2025 年 10 月完成大型 Dewar 底座安装,并在 2025 年 11 月与 EUROfusion 发布研究计划。材料称目标是 2027 年首次等离子体,机器预计将在 Q>5 下研究燃烧等离子体物理,同时作为氘氚运行、材料、包层模块和燃料循环系统的测试平台。战略上,BEST 被定位为 EAST 与未来中国示范堆之间的桥梁;合肥也已拥有 EAST、CRAFT,以及正在扩张的供应商和高校网络。资产集中给了 Neo Fusion 罕见的国家项目光环,但也制造了进度压力:从首次等离子体到 2030 年发电验证,窗口按核聚变标准看异常紧。[CO019, CO020, CO021, CO022, CO023, CO024]
| 日期 | 里程碑 | 变化 | 意义 |
|---|---|---|---|
| 2023-05-19 | 公司成立 | Neo Fusion / 聚变新能(安徽)有限公司在合肥成立 | 围绕 BEST 搭起商业化载体 |
| 2023-05-19 | 初始注册资本 | 约 US$723.37M,涉及 NIO、Anhui Province Energy 和另一名投资方 | 聚变公司成立首日资本金异乎寻常地高 |
| 2024-07 | 注册资本调整 | 注册资本从 RMB 5.0B 增至 RMB 14.5B,CNPC 和 Science Island 参与 | 标志资金结构从地方加产业资本转向国家级战略背书 |
| 2025-01 | EAST 等离子体运行创世界纪录 | HFIPS 在 EAST 上维持 1,066 秒高约束等离子体运行 | 强化 BEST 与合肥集群背后的技术底座 |
| 2025-10 | 关键部件安装 | BEST 杜瓦底座在最终装配阶段完成安装 | 公司从融资故事进入重工程执行 |
| 2025-11 | BEST 研究计划启动 | Neo Fusion 与 CAS 联合发布 BEST 计划和国际科学项目 | 增加国际科研可信度和合作网络深度 |
| 2026-01 | 聚变产业会议 | 董事长描述生态快速搭建和高校合作 | 公司不只是项目 SPV,更像集群组织者 |
| 2026-03 | 政策倡导升温 | 董事长公开呼吁为商业聚变完善立法、标准和人才建设 | 商业化障碍既是技术问题,也是制度问题 |
时间线把公司资本事件和项目执行里程碑放在一起,因为 Neo Fusion 基本离不开 BEST 的工程排期。
[CO001, CO006, CO008, CO020, CO005, CO030]| 资产 / 机构 | 当前角色 | 对 Neo Fusion 的价值 | 证据 |
|---|---|---|---|
| EAST 托卡马克 | 创世界纪录的超导等离子体实验装置 | 为 BEST 设计升级提供中国运行经验和可信度 | HFIPS / CAS 2025 纪录报道 |
| BEST 托卡马克 | 紧凑型燃烧等离子体工程平台 | Neo Fusion 的旗舰项目和商业化锚点 | CAS / HFIPS / Eurofusion / FEB |
| CRAFT | 子系统验证与部件研究设施 | 强化本地供应链认证和测试基础设施 | CAS 2024 设施报道 |
| Science Island / HFIPS | 科研园区和股东关联机构 | 提供科研人才、机构合法性,也可能影响治理 | Baike / FusionXInvest / HFIPS |
| Hefei Comprehensive National Science Centre 体系 | 国家科学城上位平台 | 把 BEST 周边地产、实验室和政策伞集中起来 | BEST 项目资料 / Xinhua 报道 |
| 聚变产业联盟 / 高校联系 | 供应商与人才管线 | 提升 Neo Fusion 为长期超大项目配人和找供应商的能力 | Anhui News / Baike |
这些生态资产之所以重要,是因为 Neo Fusion 与其说是独立初创公司,不如说是区域集中型国家聚变项目外层的商业化壳。
[CO026, CO027, CO028, CO029, CO031]1.4 反向信号与未解尽调问题
主要反向读数不是 Neo Fusion 缺少战略支持,而是公开记录相对于其声称的资本规模和雄心仍然偏薄。可获取资料没有披露收入、客户、员工数,也没有披露 RMB 14.5 billion 注册资本基础对应的现金使用。公司官网从国际网络访问时并不是可靠研究界面,甚至完整的 2024 年后股东表、现任法定代表人等基础治理信息,也更容易从二级镜像而不是主要公司备案中拼出。行业层面,Crunchbase 的 2024 年融资现实检查显示,如果项目离商业化仍有多年,核聚变资本周期可以很快降温。Neo Fusion 可能受国资支持保护,但其未来仍取决于能否在极压缩的时间表里,把国家项目声望转化为可重复的工程执行,再转化为可信商业运营模式。[CO033, CO034, CO035, CO036, CO037, CO038]
1.5 展示材料
02市场分析
2.1 市场边界:稳定清洁电力,而不是科研经费
Neo Fusion 竞争的不是实验室实验小市场;它想进入的是大得多的稳定低碳电力市场,以及建设首批核聚变电站所需的基础设施市场。这个区分很重要,因为许多已发布的核聚变市场报告会直接跳到数千亿美元结局,却没有分清示范支出和可融资售电。ITER 和其他官方来源解释了为什么奖品如此巨大:核聚变承诺提供高密度、可调度、燃料充足的电力,碳强度远低于化石发电。与此同时,数据中心用电需求上升、更广泛清洁能源需求增长,也支撑电网会持续需要新的稳定发电选项。Neo Fusion 的现实市场始于证明 BEST 能把科学推进为可融资的示范资产。换句话说,市场分析必须分阶段:先是经验证工程能力市场,再是示范电站市场,最后很久以后才是成熟的核聚变交付电力市场。[CM001, CM002, CM013, CM014, CM032]
| 细分领域 | 是否纳入 Neo Fusion 逻辑? | 意义 | 近期买方 |
|---|---|---|---|
| 电网级稳定电力 | 是 | 聚变电站长期主要变现路径 | 国有或受监管公用事业公司 |
| 示范电站资本开支 | 是 | 电力销售开始前的第一个真实市场 | 政府、国企、战略投资者 |
| 聚变供应链设备 | 是 | 电力销售出现前,BEST 阶段需求先从这里启动 | 研究机构和电站开发商 |
| 工业过程热 | 较晚 | 发电得到验证后的相邻市场 | 重工业和氢能开发商 |
| 消费端零售电力 | 否(初期) | 第一波聚变经济性还触不到这么下游 | 零售供电商,而非 Neo Fusion 直接面对 |
可服务市场是分阶段到来的:设备和示范支出先出现,电力销售则位于商业化曲线的最远端。
[CM001, CM017, CM026]核聚变市场应先从研发基础设施走向示范,再走向公用事业和工业购电,而不是直接进入大规模售电。
日期只是方向性,因为整个行业仍处商业化前,里程碑时间也不确定。
[CM017, CM034, CM038]2.2 规模测算视角与买方分层
第三方市场报告把长期核聚变机会放在 2030 至 2040 年的数千亿美元级别,但这些数字更适合作为方向感,而不是承销输入。它们依赖高度乐观的假设:技术成熟度、许可审批和电站部署都要以尚无任何核聚变公司证明过的速度推进。更有用的尽调视角是先识别谁会最早付款:先是政府和国资支持的电力公司为示范项目买单,再是受监管电力公司和部分工业买家购买早期电力。这个顺序尤其适合 Neo Fusion,因为公司深嵌于中国国资支持的核聚变生态。第一个市场因此是机构化、谈判型市场,不是大众市场,也没有透明价格。即便全球需求巨大,目前也只有极少数交易对手能吸收首台套技术、许可、建设和可用率风险。[CM003, CM004, CM005, CM006, CM007, CM010]
| 来源 | 参考时间范围 | 估算 | 解读 | 可靠性判断 |
|---|---|---|---|---|
| Precedence Research | 2030 | $471.99B | 宽口径核聚变市场 | 中;自上而下且偏乐观 |
| Maximize Market Research 估算 | 2032 | $572.74B | 聚变能源收入视角 | 中;方法不针对单家公司 |
| Industry Today | 2030 | $496.55B | 聚变市场总量口径 | 中低;转售型市场研究估算 |
| EIN News / Allied 式新闻稿 | 2040 | $840.3B | 长周期市场预测 | 低;偏宣传分发 |
| 保守投资测算视角 | 2030s | 远小于已发布 TAM | 首个市场是示范和采购,不是市场化售电 | 对尽调更有实际参考价值 |
这些估算可作为方向性背景,但过于乐观且口径不一致,不能直接作为 Neo Fusion 的投资测算基准情景。
[CM003, CM004, CM005, CM006, CM007, CM027]| 细分领域 | 买方 | 用户 | 付款方 / 赞助方 | 采用触发因素 |
|---|---|---|---|---|
| 国家 / 省级示范电站 | 国家能源规划部门 | 科研 + 运营联合体 | 国有资本和战略国企 | BEST 能燃烧 D-T 等离子体并支撑工程放大的证据 |
| 电网公用事业公司 | 区域公用事业公司或电网公司 | 公用事业发电部门 | 准许收益资产基数 / 战略资本开支 | 可融资的可靠性和许可路径 |
| 大型工业负荷 | 钢铁、化工、氢能、超大园区 | 电站运营方 | 企业或 PPA 赞助方 | 需要 24/7 稳定脱碳电力 |
| 数据中心集群 | 云厂商或园区运营方 | 电力与冷却运营 | 企业 PPA 和基础设施资本开支 | 电力紧缺和排放压力 |
| 国家实验室 / 研究机构 | 政府科研机构 | 研究人员 | 公共 R&D 预算 | 商业电站前需要下一步实验 |
Neo Fusion 最早的市场很可能是公共示范资金叠加战略产业或公用事业参与,而不是纯市场化发电。
[CM010, CM011, CM012, CM023, CM035]自上而下的市场报告指向很大的结果,但估算差异本身说明当前总可用市场(TAM)讨论仍然脆弱。
“实际投资判断视角”柱仅作示意,用来显示可融资的近端市场可能比长周期 TAM 小多少。
[CM003, CM005, CM004, CM006, CM027, CM008]实际市场会从全球电力需求迅速收窄到少数有能力赞助首批核聚变资产的交易对手。
数值是相对单位示意,不是美元金额;目的在于说明从广义需求到早期商业可融资性的收窄路径。
[CM023, CM027, CM038]2.3 中国特定的采用路径
中国给 Neo Fusion 提供了异常友好的早期市场,因为示范价值本身就有战略重要性。围绕 BEST 的官方报道称项目目标是在 2030 年进行发电验证;CAS 生态报道则描述合肥有意集中 EAST、CRAFT、BEST、供应商和新人才管线。这意味着 Neo Fusion 不需要一个常规商业市场来证明下一阶段支出合理;它需要继续站在国家工程路线图中心。如果 BEST 表现良好,最可能的下一个客户不是纯私营市场化电站,而是某种国资支持的示范项目或电力公司关联项目。这条路径在国内是优势,但如果地缘政治收紧,也可能限制公司的国际可移植性。因此,公司可能先享有特权本土市场,之后才出现类似全球开放出口市场的东西;这个先后顺序最终可能定义它的第一个十年。[CM015, CM016, CM017, CM024, CM031, CM034]
| 推动因素 | 证据 | 意义 | 剩余缺口 |
|---|---|---|---|
| BEST 2030 示范雄心 | 中国官方报道 | 打开国内里程碑驱动的市场窗口 | 需要按期实现首次等离子体 |
| 合肥聚变集群 | CAS 生态报道 | 集中人才、供应商和基础设施 | 集群深度不等于收入确定性 |
| EAST 和 CRAFT 积累 | HFIPS / CAS 报道 | 把运行和部件经验导入 BEST | 商业电站设计仍是另一件事 |
| 国资支持资本 | Neo Fusion 资料和媒体报道 | 支撑资本开支很重、收入前的执行 | 资本效率和投放节奏仍不透明 |
| 国际科研联系 | EUROfusion / CAS | 提升科研合法性和技术学习速度 | 不保证海外市场准入 |
中国提供了非常有利的示范环境,但从国资支持的工程项目走向有市场定价的电力,路径仍未跑通。
[CM015, CM016, CM017, CM031]2.4 约束与矛盾信号
核聚变市场的核心矛盾在于,估值叙事和市场规模报告已经按公用事业级数字讲故事,而行业仍在证明硬件里程碑。Crunchbase 2024 年的融资现实检查显示,大额融资一旦停下,投资人热情会很快降温;2026 年即便融资反弹,同一个结构性问题仍在。资本强度、变化中的许可框架、地缘敏感性和替代技术都会拖慢采用。落到 Neo Fusion,强劲宏观市场不会自动变成可融资的公司结果。公司仍要证明性能、定义经济性,并说明一个中国示范资产如何转化为可重复的商业需求。在此之前,市场热情应被读成未来品类的期权价值,而不是可融资客户市场已经存在的证据。这个区分是任何严肃核聚变尽调的核心,今天的 Neo Fusion 尤其如此。[CM019, CM020, CM021, CM018, CM027, CM037]
| 约束 | 证据显示 | 为什么拖慢采用 | 监测指标 |
|---|---|---|---|
| 资本强度 | 行业已融资数十亿美元,但仍需要一轮又一轮超大融资 | 首批电站对常规项目融资仍过于昂贵 | 追加国家资本或战略资本承诺 |
| 许可与标准 | 监管框架仍在成形 | 许可路径不清会拖慢公用事业采购 | 面向聚变的正式规则制定 |
| 时间表错配 | 商业需求叙事跑在硬件成熟度前面 | 买方合同可能先于工程证明 | 首次等离子体和 Q 里程碑按时达成 |
| 替代技术 | 裂变、储能和电网建设持续推进 | 公用事业公司可能更早在其他路径解决稳定电力需求 | 相对 LCOE 与许可对比 |
| 地缘政治 | 国际合作在增加,但敏感度高 | 中国项目的跨境客户市场可能收窄 | 出口管制、合作项目限制 |
市场机会确实存在,但从 BEST 走到任何可融资商业电站,每个主要约束都卡在关键路径上。
[CM019, CM018, CM020, CM021, CM031, CM037]2.5 展示材料
03竞争格局
3.1 格局:谁真正与 Neo Fusion 竞争
Neo Fusion 同时在两个相互重叠的赛场竞争。第一是民营核聚变竞赛,CFS、Helion、TAE、Zap 和 Tokamak Energy 占据主导,每家公司都试图把不同物理路线转化为可融资的电站故事。第二是公共项目赛场,托卡马克声望、科学学习和工程可信度由 ITER 以及更广泛的托卡马克生态塑造。BEST 夹在两者之间:它比纯国家实验室更像初创公司,却比传统民营初创公司更依赖国资支持、也更重基础设施。这种混合身份意味着 Neo Fusion 竞争的不只是技术,还包括政策支持、供应链准入,以及为中国拿下下一个重大里程碑的能力。未来买家和战略背书方可能会把国家项目可信度看得几乎和初创公司执行速度一样重。因此,竞争者集合比简单的风投支持初创公司名单更宽,Neo Fusion 必须同时争夺象征意义、工程可信度和最终可融资性。[CP001, CP002, CP033]
| 竞争对手 | 类别 | 规模 / 融资信号 | 目标细分市场 | 差异化 | 局限 |
|---|---|---|---|---|---|
| Commonwealth Fusion Systems | 私营托卡马克同业 | $863M Series B2;Google 合作关系 | 电网级公用事业电站 | 高场托卡马克,磁体项目强 | 仍处商业化前,资本开支重 |
| Helion | 替代磁约束同业 | $425M Series F;Microsoft PPA 证据 | 直接商业售电 | FRC 架构,主打直接发电概念 | 架构有风险,时间表激进 |
| TAE Technologies | 替代磁约束同业 | $150M 轮融资;路线图长 | 未来商业电站 / 工业用电 | 长期瞄准无中子聚变,采用束驱动 FRC | 商业化路径更靠后 |
| Zap Energy | 紧凑型替代路线同业 | $130M 新资金;DOE 里程碑 | 试点电站 / 低成本电站 | 不用大型超导磁体 | 公用事业规模的公开验证更少 |
| Tokamak Energy | 欧洲磁约束同业 | $125M 融资;聚焦磁体 | 聚变 + 磁体商业化 | HTS 磁体专长强 | 资本基础小于美国头部同业 |
| Neo Fusion / BEST | 中国国资支持的托卡马克入局者 | 注册资本折合 ~US$2.03B | 国资支持示范,随后进入公用事业市场 | HFIPS / EAST / CRAFT 生态和国家项目背书 | 缺少公开客户证明,披露薄 |
本表比较的是公开信号,而不是经审计的企业价值或收入,因为所有聚变同业几乎都仍处商业化前。
[CP001, CP004, CP009, CP013, CP016, CP020]同业会因公开客户牵引和公开技术透明度不同而聚成不同区间。
坐标轴为顺序评分:x = 公开客户 / 路线到市场证明,y = 公开技术透明度和里程碑可见度。
[CP029, CP022, CP023]3.2 同行画像与技术差异化
直接同行中,CFS 是最接近的托卡马克类比对象,因为它把高场磁体进展与 ARC 的并网电站叙事绑在一起。Helion、TAE 和 Zap 更像架构威胁,而不是直接复制者:如果它们的非托卡马克系统跑通,可能给出不同成本、复杂度和运行曲线。Tokamak Energy 重要,是因为它展示了美国之外磁约束商业化的另一条路。Neo Fusion 的差异化,不在于独特等离子体概念,而在于把中国 EAST 到 BEST 的工程传承,与异常强的国资资本结合起来。劣势是,这种差异化很大程度还只是上下文优势,尚未在面向客户的产品中得到证明。投资人能看见生态优势,却仍看不见该优势如何转化为客户偏好。相比之下,若干西方同行已经把差异化包装成明确面向场址方、电力公司和战略承购方的商业语言。[CP003, CP008, CP012, CP015, CP019, CP021]
| 能力 / 标准 | Neo Fusion | CFS | Helion | TAE | Zap | Tokamak Energy |
|---|---|---|---|---|---|---|
| 国家资本厚度 | 强 | 中 | 中 | 中 | 低至中 | 低至中 |
| 公开客户证明 | 弱 | 强 | 强 | 弱 | 弱 | 弱 |
| 英文技术透明度 | 弱 | 强 | 强 | 中 | 中 | 中 |
| 托卡马克谱系 / 磁约束连续性 | 强 | 强 | 低 | 低 | 低 | 强 |
| 可见监管进展 | 中 | 强 | 中 | 中 | 强 | 中 |
| 供应链生态优势 | 中国生态强 | 美国磁体生态强 | 中 | 中 | 中 | HTS 磁体强 |
评分是有证据支撑的序位判断,而不是数值打分。它们概括当前披露质量和里程碑可见度,而非最终技术成功概率。
[CP021, CP022, CP023, CP028, CP029]能力覆盖的差异更多来自商业化和披露,而不是宏观核聚变雄心。
定性矩阵仅基于引用的公开证据,不试图给物理成功概率打分。
[CP025, CP026, CP028, CP031]3.3 商业牵引、包装与监管姿态
CFS 和 Helion 目前在同行组中公开商业信号最强,因为它们披露了潜在客户关系;CFS 还展示出可见的并网和许可进展。Zap 和 TAE 在公开记录里仍更偏技术优先;Tokamak Energy 则继续呈现核聚变加磁体平台的混合叙事。Neo Fusion 在客户证明上落后于这一组,因为公开资料没有披露 PPA、场址承购或同等商业化协议。与此同时,没有任何同行真正发布成熟定价;多数公司仍在销售里程碑可信度,而不是标准产品包。因此,未来买家的比较选择很可能取决于信任、里程碑、监管清晰度,以及谁能承接首台套风险。Neo Fusion 可以在没有当前定价的情况下保持竞争力,但不能无限期缺少可见客户开发路径。这个缺口拖得越久,披露交易对手的同行就越会塑造买家对可信核聚变供应商的预期。[CP005, CP010, CP024, CP032, CP029, CP036]
| 公司 | 当前可见商业产品包 | 价格 / 单位可见度 | 包含能力 | 含义 |
|---|---|---|---|---|
| Neo Fusion | 尚无公开产品包 | Unknown | 仅 BEST 示范路径 | 投资人无法对标价格或可签约性 |
| CFS | 未来电力 / 战略合作叙事 | 未公布电价 | SPARC 验证 + ARC 电网电站概念 | 商业产品包仍取决于里程碑 |
| Helion | 购电协议叙事 | 未公开长期电价表 | 聚变电力交付承诺 | 即使定价不完全透明,已有客户证明 |
| TAE | 未来电站路线图 | Unknown | 电站和技术开发路径 | 仍未进入价格发现 |
| Zap | 试点电站 / 未来电力叙事 | Unknown | 示范系统和里程碑推进 | 可能靠简化架构竞争,而非当前价格 |
| Tokamak Energy | 技术平台加聚变电站愿景 | Unknown | 聚变和磁体技术栈 | 产品包仍早期且混合 |
没有可信同业公布成熟的聚变价目表;竞争对比因此要看合同形式、客户证明和里程碑可见度,而不是报价。
[CP024, CP005, CP010, CP032]Neo Fusion 在资本背书和生态入口上得分最高,但客户证明和披露透明度较弱。
1-5 顺序 KPI 分数是相对于具名竞争对手组的相对评分,仅基于公开证据。
[CP021, CP022, CP037, CP033]3.4 护城河耐久性与竞争风险
Neo Fusion 的护城河在国家能力重要的地方最强:资本深度、国家实验室准入,以及与合肥集中核聚变生态的整合。护城河在客户和投资人需要透明度的地方最弱:公开技术细节、客户开发证明,以及 BEST 之后公司到底卖什么的明确包装。短期商品化风险较低,因为行业仍分散在托卡马克、FRC 和 Z-pinch 路线之间。但随着时间推移,磁约束核聚变供应链的一部分可能标准化,任何单一托卡马克开发商的优势都会被压缩。因此,战略问题在于,Neo Fusion 能否在披露更充分的同行锁定首批严肃买家之前,把上下文优势转化为持久商业优势。否则,公司可能在技术上重要,却在商业上排到第二梯队。对中国来说这仍有战略意义,但会削弱公司对外部资本的独立吸引力。[CP025, CP026, CP027, CP023, CP030, CP031]
| 护城河主张 | 威胁 | 严重性 | 缓释措施 / 尽调问题 |
|---|---|---|---|
| 国家资本和政策支持 | 商业动作更灵活的西方同业可能更快拿下客户 | 高 | 跟踪国资支持带来的是执行速度,还是只是预算规模 |
| 合肥聚变生态 | 美国和欧洲的磁体、实验室与客户生态竞争 | 中 | 观察 BEST 能否有效吸引国际伙伴和供应商 |
| EAST 到 BEST 的托卡马克连续性 | 非托卡马克架构可能以更低成本或更快速度达成里程碑 | 高 | 持续对标 Helion、TAE 和 Zap 的里程碑质量 |
| 披露不透明可以保护 IP | 不透明也会削弱信任和客户开发 | 高 | 推动更多公开技术和商业披露 |
| 国家战略重要性 | 地缘政治可能限制国际市场进入 | 中 | 跟踪出口管制和合作规则 |
Neo Fusion 的护城河真实存在,但结构混杂:最强优势来自环境和政治,最弱环节在商业化和披露。
[CP021, CP023, CP028, CP035, CP037]3.5 展示材料
04财务情况
4.1 资本结构账面庞大,但难以承销
Neo Fusion 的财务故事从规模开始,却不是从清晰度开始。公开资料和中国商业媒体指向 2023 年初始注册资本事件,以及 2024 年更大的增资;两者合计意味着约 RMB 14.5 billion,或约 US$2.03 billion 的注册资本。对于核聚变初创公司,这个规模非常罕见,也给了公司多数同行艳羡的战略现金叙事。问题在于,注册资本不等于定价股权、账上现金或完全透明的资金用途表。可获取公开材料没有说明其中多少已经实缴、哪些里程碑控制资金投放,也没有说明现有资本基础能让 BEST 撑多久,才需要下一次融资。NIO 作为早期披露股东的存在,提供了一个有用背景信号:创始财团至少有一部分来自一家提交公开年报并进入公开资本市场的发行人,这不同于小型纯风投股权结构。但即便如此,Neo Fusion 今天到底能实际花多少钱这个更重要的问题仍未回答。[CI010, CI011, CI012, CI013, CI014, CI017]
| 账面现金 / 烧钱速度 / 现金跑道 / 资金用途 | 当前价值 / 状态 | 置信度 | 为何重要 | 缺口 / 备注 |
|---|---|---|---|---|
| 名义资本基础 | 注册资本 RMB 14.5B / ~US$2.03B | 高 | 划定支持规模 | 不等于定价股权或可动用现金 |
| 实缴资本 | 二手来源部分提及,尚未完全证明 | 中 | 决定真实资金能力 | 缺少一手证明 |
| 烧钱 | 未披露 | 低 | 计算现金跑道所需 | 无法从公开材料可靠推断 |
| 现金跑道 | 未披露 | 低 | 制定融资计划所需 | 取决于实缴现金和资本开支节奏 |
| 计划资金用途 | BEST 建设、生态搭建、人才、供应链 | 中 | 将资本与里程碑挂钩 | 缺少详细预算桥接 |
| 下一轮融资触发点 | 可能与 2027 年首次等离子体 / 后续示范需求挂钩 | 中 | 决定股权稀释时点 | 官方未披露融资路线图 |
| 债务 / 项目融资 | 公开资料未披露 | 高 | 关系到首座电站经济性 | 未来可能出现,但目前缺位 |
本表保留一个关键区别:注册资本口径的醒目数字,并不等于首创型聚变项目所需现金已经可用。
[CI010, CI011, CI012, CI013, CI014, CI015]最重的财务压力点都落在收入开始之前。
该矩阵用定性方式捕捉压力点,而非给出现金流数值,因为详细预算并未公开。
[CI007, CI014, CI041, CI035]4.2 变现仍停留在未来时
没有公开证据显示 Neo Fusion 今天已有收入。隐含终局足够清楚——等核聚变示范或商业电站出现后销售电力——但变现桥梁仍大多停留在概念层面。公开资料没有披露电价、客户条款、售电定价目标或付费工程工作。Helion 和 CFS 的可比 PPA 证明,核聚变客户可以在商业交付前签约,但它们不能填补 Neo Fusion 自己的定价缺口。现阶段,公司应被视为带有未来商业期权的工程项目,而不是已有可读收入模式的公司。这个区分很重要,因为投资人很容易过度解读资本堆栈的绝对规模,并假设商业化已经就绪。Neo Fusion 的披露不支持这个结论:BEST 仍是里程碑桥梁,不是收入机器,所有主要变现输入都被推迟到后续阶段。[CI001, CI002, CI003, CI004, CI026, CI033]
| 收入来源 | 机制 | 单位 | 当前价值 / 状态 | 质量 | 尽调问题 |
|---|---|---|---|---|---|
| 未来售电 | 通过示范或商业聚变电站售电 | MWh | 尚未开始 | 推测性强,但属主线 | 索取目标电价 / 承购模型 |
| 工程 / 项目开发支持 | 由发起方支持建设和示范工作 | 按项目 | 有所暗示,但未单独定价 | 可见度低 | 澄清是否存在付费工程服务 |
| 供应链或技术参与 | 围绕 BEST 的组件或合作关系获取价值 | 按项目 | 可能存在但未确认 | 推测性 | 询问这是否是真实计划收入来源 |
| 政府共同出资 / 战略支持 | 公共机构提供的非收入支持 | 按项目 | 可能重要,但未披露 | 非经常性 | 索取拨款 / 补贴计划 |
| 未来工业热 / 氢能邻近机会 | 向重工业出售稳定电力或热能 | 能源单位 | 未披露 | 推测性 | 澄清 BEST 之后的邻近业务路线图 |
上述收入来源都属前瞻性。Neo Fusion 尚未公开披露活跃收入、客户合同或正式变现时间表。
[CI001, CI002, CI003, CI033]| 价格 / 单位 / 合同 | 标价与实际成交价 | 包含能力 | 未知项 | 含义 |
|---|---|---|---|---|
| 电价 / PPA | 未披露 | 将包括电站出力和并网交付 | 无公开电价或 PPA 格式 | 无法建模收入 |
| 示范赞助 | 未披露 | 可能包括与里程碑挂钩的工程权益 | 看不到合同经济性 | 缺少客户证明 |
| 技术或供应协议 | 未披露 | 可能获得组件或技术诀窍 | 看不到产品包 | 战略价值可能早于定价价值出现 |
| 产能预留 / 期权结构 | 未披露 | 同业叙事中可见,Neo Fusion 尚无 | 未见 Neo Fusion 证据 | 同业无法填补 Neo Fusion 自身定价缺口 |
本章区分未来市场是否存在与已披露变现机制是否存在;目前只有前者可见。
[CI004, CI026, CI040]Neo Fusion 的收入桥从工程验证通向政府背书示范,之后才走到售电。
这是一条定性收入桥,因为今天没有公开定价或合同条款。
[CI002, CI005, CI042]4.3 资本强度是核心财务变量
即使没有公开项目预算,成本画像也很容易定性:Neo Fusion 正在建设首台套紧凑型燃烧等离子体平台,目标不只是撞上物理里程碑,还要测试材料、包层模块和燃料循环系统。这意味着商业收入到来前,需要高资本开支、长交付周期和专业劳动力。Lazard 电力成本研究等公开成本基准能提醒投资人,更广泛电力市场非常严苛,但它们不能替代 Neo Fusion 专属的 LCOE 或电站成本模型。缺少烧钱速度、毛利率或单站经济性时,真正的财务问题不是核聚变能否变得有价值,而是这家公司能否在不过度稀释或延误的情况下抵达下一个里程碑。更广泛行业资料强化了这种谨慎。Helion 和 Zap 等同行材料越来越围绕具名项目、里程碑和面向客户的叙事来包装核聚变,这只会让 Neo Fusion 仍然有限的商业化披露更显眼。[CI007, CI008, CI009, CI032, CI030, CI037]
| 指标 | 数值 / 空值 | 置信度 | 为何重要 | 尽调问题 |
|---|---|---|---|---|
| 每 MWh 收入 | 低 | 电力经济性的核心驱动 | 索取内部电价假设 | |
| 单座电站 / 每 MW 资本开支 | 低 | 决定项目融资可行性 | 索取技术经济模型 | |
| 毛利率 | 低 | 决定运营带来的股权价值 | 索取供电成本模型 | |
| 烧钱速度 | 低 | 决定现金跑道和资金需求 | 索取管理账 | |
| 现金跑道月数 | 低 | 决定融资紧迫性 | 索取实缴现金和预算节奏 | |
| 氚 / 燃料循环成本 | 低 | 潜在重大可变成本项 | 索取燃料循环计划 |
几乎所有单位经济性单元格仍为空值,因为 Neo Fusion 尚未发布可用于融资的成本或收入模型。
[CI008, CI009, CI030, CI015, CI016]Neo Fusion 未来的单位经济性取决于一长串仍未披露的输入项。
由于公开资料未披露 Neo Fusion 的资本开支、运营开支或电站可用率假设,这条桥只能定性呈现。
[CI009, CI032, CI030]关键财务输入仍未披露,公开层面只能给出很宽的范围。
数值混合了年份和百万美元风格的参考点,仅作方向性示意;它不是统一估值模型。
[CI013, CI016, CI018, CI025]4.4 公开承销结论:战略资金充足,常规披露不透明
因此,财务结论是混合的。Neo Fusion 看起来比许多民营同行资金更足,并受益于国资支持,短期融资危机概率应更低。但公开披露远不足以支撑常规承销。投资人仍缺少判断资金是否充足的最低材料包:实缴现金、详细项目预算、烧钱速度、定价假设、客户管线,以及债务或补贴结构。这些出现之前,Neo Fusion 的财务故事更应理解为押注中国核聚变路线图的战略期权,而不是可以透明建模的创投投资。SPARC、ITER 和其他领先项目给出的同行教训,不是 Neo Fusion 不可能成功,而是核聚变里长期、重资本开发本来就是常态;因此,不透明的融资结构应获得额外怀疑,而不是直接通行。[CI028, CI031, CI034, CI035, CI020, CI038]
4.5 展示材料
05产品与技术
5.1 BEST 是今天的产品,但仍是工程资产
Neo Fusion 的「产品」更应理解为下一步核聚变机器,而不是商业售电方案。公开材料把公司描述为在合肥建设 BEST 的主要组织之一;BEST 是一台紧凑型高场托卡马克,目标是从长脉冲实验工作走向燃烧氘氚等离子体,并最终走向净增益发电验证。这个定位很重要,因为它说明投资人今天真正承销的是什么:机器总装、子系统集成和物理执行。公司尚未披露电力技术供应商通常会拥有的商业界面,例如可用率保证、维护承诺、电价框架或并网交付包。可见技术栈反而是资产驱动的:EAST 提供运行履历,CRAFT 验证子系统,BEST 成为一体化验证场。因此,Neo Fusion 看起来不像一家已打包的电力公司,更像包裹在中国下一座重大托卡马克建设外层的商业化载体。[CE001, CE002, CE003, CE019, CE047]
| 模块 / 资产 | 主要用户 | 状态 / 成熟度 | 差异化 | 尽调缺口 |
|---|---|---|---|---|
| BEST 主机 | ASIPP / Neo Fusion 工程团队 | 正在组装;主厂房已完工 | 紧凑型高场燃烧等离子体托卡马克 | 需要详细的子系统就绪计划 |
| EAST 运行平台 | 中国和国际研究人员 | 正在运行,并创下世界纪录脉冲 | 长脉冲超导托卡马克经验积累 | EAST 运行经验有多少能直接迁移到 BEST |
| CRAFT 子系统试验台 | 聚变硬件工程师 | 子系统已建成 / 通过验收 | 反应堆部署前验证部件 | 需要把 CRAFT 输出直接映射到 BEST BOM |
| HFIPS / ASIPP 研究体系 | BEST 项目领导层 | 活跃的机构底座 | 深厚的等离子体物理和 ITER 经验 | 商业决策权与研究所控制权的边界仍不透明 |
| 未来示范电站路径 | 公用事业 / 工业承购生态 | BEST 之后的概念阶段 | 把研究能力接到发电雄心 | 2030 年后的商业化方案尚未明确 |
Neo Fusion 仍处于商业化前,产品栈因此以资产为中心;今天真正交付的是一条硬件和工程路径。
[CE001, CE014, CE019, CE017, CE047]| 用户任务 | 当前工作流 | Neo Fusion 方案 | 可衡量收益 | 限制 |
|---|---|---|---|---|
| 推进燃烧等离子体科学 | 运行 EAST 和国际托卡马克实验 | 把 BEST 作为下一步装置运行 | 从长脉冲物理推进到燃烧等离子体状态 | 仍不是商业电站 |
| 验证反应堆子系统 | 台架测试部件,或依赖海外项目 | 用 CRAFT + BEST 做集成部件学习 | 把偏滤器、包层和燃料循环经验本土化 | 集成细节仍稀疏 |
| 证明聚变发电路径 | 从概念上建模未来试点电站 | 目标是在 2030 年前实现净增益和发电示范 | 创造公用事业可承销的里程碑 | 公开成本和可用率假设缺位 |
| 建设国内聚变供应链 | 研发和采购分散 | 围绕 EAST / CRAFT / BEST 形成合肥集群 | 集中人才和制造学习 | 集群强度不能证明产品市场契合 |
本表用工作流视角看 BEST:今天它推进工程验证,多于推进客户部署。
[CE002, CE019, CE003, CE023]Neo Fusion 目前的流程从子系统验证走向燃烧等离子体验证,之后才面向公用事业示范。
这是从工程到市场的流程,不是商业时点的证明。
[CE014, CE019, CE013, CE006, CE003]5.2 机器架构雄心很大,但对商业化前核聚变公司来说异常具体
BEST 的公开技术描述足够具体,能勾勒出真实硬件路线图。HFIPS 和 CAS 描述了一台紧凑型高场托卡马克,Dewar 包裹的超导磁体在接近零下 269 摄氏度下运行,等离子体核心则超过 1 亿摄氏度。EUROfusion 补上关键物理框架:BEST 将研究 Q>5 以上的燃烧等离子体行为,同时支持对未来试点电站重要的材料、包层模块和燃料循环测试。这些说法仍是目标,不是已交付结果,但远比泛泛的「核聚变 AI」或「清洁能源」口号具体。2025 年 Dewar 底座安装、2026 年主厂房完工报道,以及反复出现的 2027 年首次等离子体或完工目标,共同说明项目已经进入精密总装。首次等离子体之后的路线图也很明确:中文公开来源称 BEST 寻求在 2030 年前后实现净增益和核聚变发电。因此,核心技术问题是进度执行,而不是使命定义。[CE005, CE011, CE006, CE009, CE008, CE012]
| 层级 / 部件 | 作用 | 依赖 | 风险 |
|---|---|---|---|
| 超导磁体系统 | 用高磁场约束等离子体 | 低温性能和导体质量 | 失超、可制造性和脉冲电源复杂度 |
| 杜瓦真空包壳 | 把磁体与高温等离子体环境热隔离 | 大型精密制造和装配 | 装配延误会影响整机进度 |
| 燃烧等离子体核心 | 达到 Q>5 并开展 D-T 运行 | 加热、控制和等离子体稳定性 | 磁约束聚变尚未证明净增益目标 |
| 偏滤器和面向等离子体部件 | 排出热量、杂质和粒子 | CRAFT 硬件就绪度和材料科学 | 热负荷失效或氚约束 |
| 包层 / 燃料循环测试系统 | 为未来电站学习增殖和燃料处理 | 材料、氚和集成测试 | 公开运行计划仍薄弱 |
架构被简化为几块关键工程模块;这些模块决定 BEST 能否从 EAST 式物理实验跨到试点电站学习。
[CE011, CE006, CE021, CE045, CE031]| 日期 / 阶段 | 里程碑 | 状态 | 含义 | 来源 |
|---|---|---|---|---|
| 2025-01 | EAST 达到 1,066 秒 H 模式脉冲 | 已完成 | 强化 BEST 的上游托卡马克谱系 | HFIPS / CAS |
| 2025-10 | BEST 杜瓦底座安装完成 | 已完成 | 确认进入主机组装阶段 | HFIPS / CAS |
| 2025 | CRAFIT 偏滤器原型通过验收 | 已完成 | 提升对高热流部件就绪度的信心 | CAS |
| 2027 | BEST 完工 / 首次等离子体目标 | 进行中 | 产品可信度的关键闸门里程碑 | HFIPS / EUROfusion |
| 2030 | 净增益和发电示范目标 | 已设定目标 | 若实现,将把 BEST 从研究资产变成能源示范证明 | HFIPS / People’s Daily 报道 |
路线图混合了已完成里程碑和前瞻目标;前瞻里程碑仍是主张,不是已交付成果。
[CE014, CE009, CE021, CE012, CE003]BEST 把机构传承、超导硬件和未来燃料循环学习叠在一个紧凑型托卡马克项目里。
分层简化了一台复杂得多的机器,但保留了公开资料中可见的主要工程依赖。
[CE017, CE011, CE006, CE021, CE007]BEST 依赖紧密相连的机构、硬件和燃料循环工作流。
该图聚焦公开可见依赖,省略了许多内部子系统。
[CE017, CE014, CE019, CE011, CE045, CE003]5.3 EAST 和 CRAFT 是 BEST 技术雄心周围真正的护城河
Neo Fusion 最强的技术边,不是单项已披露专利或软件秘密,而是 BEST 背后的生态。EAST 1,066 秒高约束脉冲给了项目少有同行能主张的长脉冲托卡马克传承;HFIPS 明确表示,千秒级稳定性是未来发电的先决条件。CRAFT 则补上部件层:它的存在就是为了建设并测试与反应堆相关的硬件;2025 年 CAS 称其偏滤器原型承受了每平方米 20 兆瓦,同时甚至改善了理论氚增殖性能。换句话说,BEST 并不是被要求从零发明每个子系统。它可以调用本地链条上的等离子体运行、反应堆部件测试,以及 ASIPP 围绕 EAST 和 ITER 积累的机构经验。这不能消除执行风险,但确实意味着公司的技术故事锚定在物理基础设施和人,而不只是资本或品牌。[CE014, CE015, CE016, CE019, CE021, CE022]
Neo Fusion 在物理传承上最强,在公开运行包披露上最弱。
定性评级总结证据强度,而非数值打分。
[CE014, CE009, CE007, CE042]5.4 全球同行背景同时凸显 Neo Fusion 技术故事的成熟度与披露缺口
全球比较让图景更清晰。ITER 更大、更国际化,但它作为科学优先托卡马克的角色,解释了为什么 BEST 被定位成通往后续示范堆的更紧凑桥梁。JET 最后的氘氚实验,以及 W7-X 的长脉冲仿星器工作都表明,领先项目仍要花多年时间,才能把物理和材料学习转化为与反应堆相关的运行能力。商业同行还给出另一课:MIT/CFS 和 CFS 自身会发布详细磁体里程碑,TAE 和 Zap 则维持公开研究资料库,让外部人能检查技术学习速度。Neo Fusion 在机构深度和机器雄心上表现不错,但公开透明度较弱。可获取来源仍未披露 BEST 的氚运行计划、公开质量框架或可靠性指标。即便董事长呼吁立法和标准,也最好被读作聚变治理层仍在建设中的证据。因此,产品技术结论是:物理谱系和总装动能很强,但运行包的可信度仍不完整。[CE018, CE025, CE028, CE029, CE031, CE033]
5.5 展示材料
06客户情况
6.1 利益相关方才是当前事实上的客户基础
Neo Fusion 尚未销售电力,所以当前「客户」更适合称为战略付款方和未来场址侧利益相关方。最清楚的直接证据是股权和项目支持关系:NIO 和 NIO Capital 在创立时入股,安徽政府相关能源实体持有主要控制权,2024 年再资本化又加入 CNPC 和 HFIPS 相关资本。这一点很重要,因为它界定了当前阶段买方 / 用户 / 付款方的分工。Neo Fusion 的工程师和科研伙伴造机器;公共机构和战略投资人支付建设资金;如果 BEST 成功,未来电力公司或工业场址方才是预期下游用户。简言之,公司已经有需求侧协同,但还没有商业客户采用。那仍然不是客户收入。这个区分是尽调核心,因为利益相关方承诺可以资助示范项目,却不能证明今天已有可融资的电力产品。[CU001, CU002, CU003, CU006, CU007, CU008]
| 客群 | 买方 / 用户 / 付款方 | 用例 | 规模 / 战略价值 | 缺口 |
|---|---|---|---|---|
| 国家赞助方 / 战略投资者 | 当前付款方;影响治理 | 资助 BEST 建设并推动政策对齐 | 当前战略价值最高 | 不等于经常性收入客户 |
| 全国 / 区域公用事业公司 | 未来买方和电网集成方 | 承接首批聚变电站或购买电力 | 可能是首个外部市场 | 尚无具名 Neo Fusion 承购协议 |
| 工业脱碳买方 | 未来买方 | 稳定电力、热、氢或自用清洁能源 | 潜在第二波市场 | 未公开商业化包装 |
| 数据中心 / 数字基础设施负荷 | 未来买方或 PPA 对手方 | 获取全天候清洁电力 | 长期需求信号在增强 | 取决于首座电站证明和价格 |
| 研究 / 公共项目生态 | 用户 / 赋能方 | 示范、验证和生态建设 | 重要的商业化前采用渠道 | 不等同于客户收入 |
客群定义区分两类人:现在谁为 Neo Fusion 出钱,以及未来谁可能购电或承接电站。
[CU002, CU016, CU017, CU041, CU015]Neo Fusion 的客户旅程始于战略赞助,之后才进入公用事业和工业购电。
这是阶段化买方旅程,不是当前商业部署证据。
[CU002, CU015, CU016, CU017, CU031]6.2 真实需求证明主要来自基准案例,而不是 Neo Fusion 合同
公开资料里最强的 Neo Fusion 专属需求证明,是利益相关方承诺,而不是客户部署。NIO 明确称核聚变具有战略意义,并承诺分阶段出资;安徽电力公司相关主体已经与 NIO 在电网灵活性和充电基础设施上合作;官方 BEST 报道也表明,国资支持的示范环境就是预期第一市场。但这些都不等同于承购协议。最近的高质量基准来自其他核聚变公司。CFS 公开披露了 200 MW Google 承购协议,Helion 则宣布 Microsoft 是其计划电站的首位客户。这些案例重要,因为它们展示了核聚变商业交付前可信客户证明可以长什么样。对照这个基准,Neo Fusion 仍处早期:没有披露 PPA、没有场址合同,也没有具名外部买家愿意在 BEST 之后承接输出。本章结论因此不是需求不存在,而是需求仍是隐含的、由政策居间的,并非合同化表达。[CU005, CU009, CU015, CU021, CU022, CU023]
| 指标 | 数值 | 日期 | 来源 | 置信度 | 含义 | 缺失分母 |
|---|---|---|---|---|---|---|
| 商业客户 | 未披露 | 2026-07-19 审阅 | 公开资料 + 官方报道 | 高 | 从客户维度看仍未产生收入 | 没有管线数量 |
| 战略赞助方连续性 | 2023 年创始赞助方在 2024 年扩容,2026 年仍可见 | 2023-2026 | CnEVPost + Fusion Energy Base + 媒体 | 中 | 赞助方层面存在持续性 | 没有资金提款计划 |
| 具名未来公用事业承购 | 未披露 | 2026-07-19 审阅 | 公开资料 + BEST 报道 | 高 | 缺少商业需求证明 | 没有场址候选名单 |
| 可比聚变承购基准 | Google-CFS 200 MW;Microsoft-Helion 首个客户 PPA | 2023-2025 | CFS + ANS + Helion | 高 | 行业可在交付前签下客户 | 不是 Neo Fusion 特有证据 |
| 电网需求背景 | 数据中心需求预计到 2030 年翻倍 | 2025-2026 | Gartner + S&P/IEA | 高 | 未来需求信号强 | 没有 Neo Fusion 份额估计 |
今天唯一持久的增长证据,是赞助方连续性和行业需求背景;Neo Fusion 自身客户数量仍未披露。
[CU001, CU027, CU023, CU021, CU022, CU018]| 具名利益相关方 / 客户证明 | 客群 | 部署 / 用例 | 生产级还是试点 | 结果 / 信号 | 限制 |
|---|---|---|---|---|---|
| NIO | 战略赞助方 / 潜在用能方 | 股权投资和分阶段项目出资 | 商业化前赞助方 | 验证大型电动车生态参与者的战略兴趣 | 不是购电合同 |
| Anhui Province Energy / Wenergy | 省级公用事业生态 | 与 NIO 的电网灵活性合资公司,以及本地充电 / 储能基础设施 | 运营中的相邻能源部署 | 显示地方公用事业和电网集成对齐 | 证明不来自聚变发电 |
| CNPC / Kunlun Capital | 战略工业能源赞助方 | 参与 2024 年增资 | 商业化前赞助方 | 增加国家能源系统相关性 | 未披露承购或场址承诺 |
| CFS 基准中的 Google | 企业清洁电力买方 | 200 MW ARC 承购协议 | 交付前商业基准 | 显示聚变客户可在首座电站交付前签约 | 仅为基准,不是 Neo Fusion |
| Helion 基准中的 Microsoft | 企业清洁电力买方 | 首个聚变电站购电协议 | 交付前商业基准 | 证实超大规模买方愿意早期支持聚变 | 仅为基准,不是 Neo Fusion |
Neo Fusion 仍处于商业化前,最强的直接证明来自战略利益相关方;基准 PPA 则展示了真实客户证明最终应具备什么形态。
[CU003, CU009, CU007, CU021, CU022, CU024]当前漏斗从战略赞助走向有基准参照的购电逻辑,还没走到真实客户铺开。
Neo Fusion 尚未通过场址方或购电步骤;纳入这些步骤是为了展示必经路径。
[CU013, CU024, CU015, CU023, CU021, CU031]公开证明在赞助方承诺上最强,在商业部署和留存可见度上最弱。
序数标签概括公开证据强度,不表示数值化转换概率。
[CU003, CU009, CU007, CU021, CU022, CU023]6.3 未来买方地图从电力公司开始,之后才扩展到工业或数字负载买家
公开证据指向分阶段买方地图。首先是能吸收资本开支、许可复杂度和首台套风险的电力公司、电网和国资场址主体。SASAC 对 State Grid 的描述凸显了这个机会的规模;China Southern Power Grid 网站则显示,如果核聚变真正进入电网,它可能需要智能电网、储能、EV 和集成能力。中国新型能源体系规划进一步强化了机构优先逻辑:它把能源定位为战略基础设施,并强调源网荷储一体化和灵活性。只有经过这轮降险后,工业用户、氢项目或数据中心负载才会成为现实的第二波客户。Gartner 和 S&P/IEA 关于电力需求增长的数据让长期稳定清洁电力需求显而易见,但不会缩短首次采用路径。因此,Neo Fusion 的潜在客户理论上非常有吸引力,实践中却会高度滞后。[CU011, CU012, CU013, CU014, CU017, CU018]
该代理队列用公开关系的持续可见度替代真实收入留存,因为 Neo Fusion 尚未披露商业客户队列。
这张图只是一项披露代理:100 表示关系仍可公开看到,0 表示没有公开商业留存证据。
[CU027, CU026, CU040]6.4 耐久性只在出资方层面可见,集中度和采购风险很高
Neo Fusion 没有公开的经常性客户基础,因此 NRR、流失率、合同期限等经典耐久性指标尚不可用。唯一公开耐久性代理指标是出资方连续性:公司创始利益相关方组合在 2024 年再资本化和 2026 年公开表述中仍然可见。这有用,但不完整。它很少说明电力公司是否会真正签约、首个场址合同会持续多久,或第一个锚定买家会制造集中度风险而不是解决它。事实上,集中度正是今天的核心客户风险。Neo Fusion 依赖少数国家和战略主体,首个外部买家很可能也会是大型电力公司或场址主体,而不是分散长尾。采购摩擦同样很高,因为电力公司必须等待物理证明、监管清晰度、场址准备和可融资成本假设。行业报道看起来仍更像投资人记分牌,而不是客户记分牌。因此,客户结论是战略上对齐,商业上未证明。[CU026, CU027, CU028, CU029, CU030, CU034]
| 指标 | 值 / null | 客群 | 置信度 | 尽调要求 |
|---|---|---|---|---|
| NRR / GRR | 商业客户 | 低 | 销售启动后,要求留存收入和续约定义 | |
| 合同续约率 | 商业客户 | 低 | 要求场址或承购续约条款 | |
| 赞助方连续性 | 2023-2026 年公开记录可见 | 战略利益相关方 | 中 | 要求股权结构表和实缴资本时间线 |
| 用户满意度 / NPS | 电站承接方 / 运营方 | 低 | 部署落地后,要求合作伙伴或场址方背调名单 | |
| 合同期限 | 未来公用事业 / 工业买家 | 低 | 要求提供条款清单草案或备忘录摘要 |
多数留存字段为 null,是因为 Neo Fusion 还没有公开的经常性客户基础;出资方延续性最多只能当代理指标。
[CU026, CU027, CU040]| 扩张驱动因素 / 风险 | 影响 | 证据 | 尽调路径 |
|---|---|---|---|
| 国家支持示范优先 | 短期正向 / 集中度负面 | BEST 2030 目标和出资方基础 | 要求提供落地场址、公用事业和政策路线图 |
| 公用事业 / 电网买家集中 | 集中度风险高 | State Grid 和 CSG 是合乎逻辑、但体量极大的锚定渠道 | 要求提供买家开发策略和多场址管线 |
| 工业和数据中心扩张 | 潜在上行 | WRI、Gartner 和 S&P 指向长期需求增长 | 要求提供分细分市场商业化计划 |
| 采购摩擦和许可 | 负面 | ITER 式周期和基准 PPA 仍依赖重大验证节点 | 要求提供预期采购周期和监管里程碑 |
| 国际客户触达 | 短期负面 | 中国项目定位叠加资金背景不透明 | 要求提供国内验证后的出口 / 合作策略 |
| 目前没有客户指标 | 尽调清晰度受损 | 公开来源只到利益相关方和基准证据 | 要求提供管线、伙伴和部署 KPI |
扩张路径在细分市场层面可见,但集中度和采购摩擦仍是客户侧最主要的风险。
[CU015, CU028, CU017, CU030, CU032, CU025]6.5 展示材料
07风险
7.1 首要风险栈先是物理,其次是进度,最后才是融资和客户
Neo Fusion 的风险画像异常前置,集中在营销或普通软件迭代无法解决的问题上。公司的核心机器 BEST 仍要证明燃烧氘氚等离子体,维持足够久、足以支撑发电的性能,并把 2027 年首次等离子体目标变成 2030 年发电验证叙事。公开资料显示,即便更广泛核聚变行业也仍把 Q>1 和实用净发电视为未来里程碑,而不是当前事实。这意味着最严重风险仍是科学和工程执行。承认这一点后,其他风险会机械传导:进度滑坡制造融资缺口,融资缺口拖延客户信心,客户延迟削弱估值杠杆。Neo Fusion 的国资支持生态让它比普通初创公司更有韧性,但没有改变因果层级。因此,风险章节从一个简单判断开始:公司投资逻辑仍取决于里程碑交付,而不是运行输出。[CR001, CR003, CR004, CR005, CR046]
最高残余风险集中在物理、进度、监管和融资,而不是常见的软件创业问题。
序数值概括本章的风险判断,而不是数值评分模型。
[CR001, CR004, CR008, CR021, CR031, CR046]技术拖期不会只留在实验室里,而会级联到融资、客户和政策风险。
该图抓住公开证据可见的主要因果链;真实项目会有更多分支。
[CR001, CR004, CR025, CR033, CR037, CR046]7.2 即便核聚变更清洁,监管、法律和环境风险仍然真实
核聚变常被包装成比裂变天然更安全、更清洁,但公开法律和监管界面显示,这不等于可以免审。中国《核安全法》覆盖核设施、核材料和放射性废物,把责任放在运营者和相关供应商身上,并将核安全纳入正式监督框架。NNSA 的公开资源也强化了这一点:框架背后有机构化监测和报告体系。Neo Fusion 面临的问题不是缺少监管,而是缺少公开的机器级方案,证明 BEST 成熟为燃烧等离子体系统后,将如何处理氚、放射性副产物、废物、应急响应和质量保证。IP 和自由实施问题又加一层:全球和中国专利体系都存在,但本次审阅的公开记录没有让 Neo Fusion 自身专利位置变得可读。因此,监管风险在框架层面部分已知,在公司运行计划层面仍未解决。[CR007, CR008, CR009, CR011, CR012, CR016]
| 风险 | 可能性 | 影响 | 当前缓释 | 剩余暴露 | 投资含义 |
|---|---|---|---|---|---|
| 核安全许可与监管 | 中 | 高 | 中国法律框架已存在;NNSA 发布报告和法规 | 高 | 时间线和成本可能大幅波动 |
| 氚 / 燃料循环处理 | 高 | 高 | CRAFT 和路线图工作已有子系统研发 | 高 | 安全和运营成本风险仍未解决 |
| 放射性废物 / 污染控制 | 中 | 中 | 法律覆盖放射性废物和污染防治 | 中 | 电站走向成熟时,合规成本可能上升 |
| IP / 专利防御力 | 中 | 中 | 全球和中国都有专利制度 | 中 | IP 薄弱会削弱护城河,或增加争议 |
| 责任 / 保险披露缺口 | 高 | 中 | 未发现公开披露 | 高 | 投资人还无法给长尾法律暴露定价 |
这份台账区分两件事:框架存在,不等于 Neo Fusion 已经在机器层面满足框架要求。
[CR007, CR008, CR011, CR015, CR016, CR041]7.3 执行风险嵌在热负载、燃料循环复杂度、人才和机构依赖里
可见硬件栈让运营风险变得具体。BEST 依赖超导磁体、低温工程、长时等离子体控制、高热流材料、偏滤器性能,并最终依赖燃料循环学习。CRAFT 的偏滤器结果令人鼓舞,但它仍是受控条件下的子系统结果,不是连续电站级耐久性的证明。DOE 2026 年路线图从另一个角度得出同样的大结论:聚变燃料增殖和处理、材料、供应链、劳动力以及现实商业化,仍是行业的活跃挑战区。Neo Fusion 还带着集中的机构依赖,因为 ASIPP/HFIPS 和更广泛合肥生态提供了大量底层专业能力。公开呼吁立法、标准和人才发展,不只是政策修辞;它们说明核聚变周围的人和组织系统仍在建设。净结果是,Neo Fusion 的生态既是最强缓释项,也是最清晰的集中风险之一。[CR014, CR015, CR013, CR018, CR019, CR020]
| 风险 | 证据 | 重要性 | 当前缓释 | 剩余暴露 |
|---|---|---|---|---|
| 燃烧等离子体执行风险 | Q>5 和 2030 发电仍是目标 | 核心技术逻辑可能失败 | EAST 积累 + BEST 建设 | 高 |
| 热负荷和材料风险 | CRAFT 偏滤器里程碑仍只验证子系统 | 连续运行放大后可能失败 | CRAFT 测试进展 | 高 |
| 可靠性 / 可用时长风险 | 没有公开 RAM 指标或支持模式 | 客户需要可靠电力,不只是脉冲 | 公开不可见 | 高 |
| 进度滑坡风险 | 2027 首等离子体支撑 2030 叙事 | 延期会叠加融资和客户风险 | 国家支持和生态 | 高 |
| 安全披露缺口 | 没有公开 Neo Fusion 安全包 | 信任和审批都依赖它 | 行业层面法律和标准工作 | 高 |
Neo Fusion 拥有真实工程资产,但多数运营缓释来自基础设施,而不是结果证明。
[CR003, CR014, CR012, CR039, CR045]| 依赖方 | 角色 | 风险 | 缓释 | 剩余暴露 |
|---|---|---|---|---|
| ASIPP / HFIPS | 核心科研和工程骨干 | 机构或领导层扰动 | 生态深、公共支持强 | 中高 |
| 政府出资方 | 资本和政策支持 | 优先级变化或财政收紧 | 国家战略定位 | 高 |
| 专业部件供应商 | 磁体、低温、面向等离子体硬件 | 瓶颈或质量失效 | CRAFT 和本地产业集群支持 | 高 |
| 国际合作者 | 研究计划和科学交流 | 地缘政治限制或摩擦 | 正式合作渠道 | 中 |
| 首个公用事业 / 场址客户 | 初始商业化桥梁 | 一旦签约,集中度极高 | 潜在长期锚定价值 | 高 |
依赖风险重要,是因为 Neo Fusion 的生态优势本身也会形成集中机制。
[CR020, CR037, CR008, CR036, CR038]| 风险 | 信号 | 影响 | 缓释成熟度 | 尽调要求 |
|---|---|---|---|---|
| 关键科学家集中 | 项目依赖专业聚变人才 | 高 | 中 | 要求提供接班和板凳深度图谱 |
| 人才供给稀缺 | 公开推动立法、标准和人才 | 高 | 低中 | 要求提供招聘管线和留存数据 |
| 跨职能整合 | 物理、材料、低温和政策彼此咬合 | 高 | 中 | 要求提供整合项目治理模型 |
| 执行速度 | 2027 和 2030 目标几乎不留余量 | 高 | 中 | 要求提供关键路径和浮动时间分析 |
| 商业化转型能力 | 工程组织最终要转成面向客户的公用事业伙伴 | 中高 | 低 | 要求提供商业组织和落地场址计划 |
聚变执行风险一部分是人员系统问题,不只是物理问题。
[CR019, CR018, CR020, CR004, CR046]Neo Fusion 要跑通,机构、监管、组件学习曲线和支持方必须同时对齐。
依赖图经过简化,但抓住了主要非财务卡点。
[CR020, CR014, CR011, CR037, CR038]7.4 资本强度、替代电力竞争和淘汰标准定义下行空间
即便物理路径成立,Neo Fusion 仍面对残酷的经济和竞争环境。核聚变行业融资已经增长,但相对于首批电站的基础设施雄心,已披露总额仍然很小,头部民营公司还吸走了大量可得资本。融资波动是真实的:一些同行能融资数亿美元,另一些只能完成小得多的轮次,而 AI 资本如今让整个核聚变行业显得很小。与此同时,过去十年可再生能源成本大幅下降,迫使核聚变竞争的不只是理论能源稀缺,还包括快速改善的现实替代方案。客户风险也从这种压力中传导出来。如果电力公司和工业买家可以等待更便宜或风险更低的方案,BEST 的每一次延迟都会让商业化更难。投资人需要抓住的一点是,淘汰标准可以监测:如果 2027 年节点错过且没有可信恢复路径,如果真实安全和监管方案迟迟不出现,或如果无法展示通往 2027 年后示范的融资桥梁,投资逻辑就应迅速下调。[CR021, CR022, CR023, CR029, CR031, CR033]
| 风险 / 触发因素 | 当前缓释 | 监测指标 | 终止标准 |
|---|---|---|---|
| 2027 里程碑延误 | 国家支持建设和现有装配进展 | 建设更新、部件验收、首等离子体状态 | 2027 目标出现重大延误,且没有可信恢复计划 |
| 监管 / 安全包装缺口 | 法律框架和 NNSA 监管已存在 | 公开安全文件、面向监管方的里程碑、标准采纳 | 电站走向成熟时,仍没有可信机器级安全包 |
| 资金桥缺口 | 大额注册资本叙事和行业出资方基础 | 实缴资本信号、新融资、预算披露 | 看不到从当前资金接到后续示范的桥 |
| 替代电源经济性跑赢聚变 | 稳定电力需求仍存在 | 可再生成本曲线、储能经济性、公用事业采购行为 | 面对更便宜替代方案,公开经济性论证已站不住 |
| 伙伴集中 / 政策转向 | 国家战略定位和本地生态深度 | 领导层变化、预算优先级、合作变化 | 主要出资方或机构支持显著削弱 |
后续刷新这章时,终止标准都写成可监测的外部事件,仍能直接操作。
[CR042, CR043, CR044, CR031, CR037]7.5 展示材料
08估值
8.1 由于 Neo Fusion 有战略资本但没有清晰市场价格,建议仍是继续研究
Neo Fusion 并不轻量。它拥有核聚变领域最醒目的账面资本堆栈之一;公开材料显示,2024 年增资后注册资本约 RMB 14.5 billion,或约 $2.03 billion。单凭这个规模,公司就足以在中国核聚变路线图中具有战略重要性。但估值纪律始于一个区分:注册资本不等于定价轮次、投后估值,也不等于可以按已核验项目预算支出的已证明现金。公开来源仍没有说明资本基础中多少已经实缴、如何按里程碑释放,或外部投资人实际拥有的经济权益是什么。这些缺口是根本性的,不是装饰性的;在证据约束下,最佳姿态是继续研究,置信度中等,估值立场未知。Neo Fusion 最终可能远比今天证据能证明的更有价值,但当前公开记录不支持精确估值。[CV001, CV002, CV003, CV004, CV005, CV006]
| 建议 | 信心 | 风险评级 | 估值立场 | 决策含义 |
|---|---|---|---|---|
| 继续研究 | 中 | 高 | unknown | Neo Fusion 可能具备战略价值,但当前公开记录不足以支撑干净的市场价格承销判断。 |
该建议受两点约束:Neo Fusion 没有披露定价轮,战略资本头条与可融资的商业化证据之间仍有缺口。
[CV005, CV042, CV048, CV051]| 锚点 | 数值 / 状态 | 来源质量 | 含义 |
|---|---|---|---|
| 公开口径注册资本 | RMB 14.5B / ~$2.03B | 官方资料 + 中国商业媒体 | 战略支持锚很强,但不是市场出清的股权价格。 |
| 定价股权轮 | 未发现公开披露 | 缺少证据 | 目前没有干净的投后估值锚。 |
| 里程碑路线图 | 2027 年 BEST 完工;约 2030 年电力示范目标 | 官方 / 技术来源 | 当前价值很大程度取决于里程碑能否兑现。 |
| 同业商业化证明 | Helion 和 CFS 披露了电站和承购结构 | 官方同业来源 | 类似证明出现前,Neo Fusion 应打折。 |
| 行业资本环境 | 不含中国的私人聚变融资 $11.52B;更广口径披露融资约 $14.24B | 分析师 + 行业新闻 | 聚变已经能吸引大额资本,但阶段仍早、成本仍高。 |
这张表把最干净的公开锚点,与噪声大或不可比的数据点分开。
[CV004, CV006, CV009, CV010, CV025, CV019]这个案例起点是罕见的战略资本,但定价和商业化证据仍未解决,建议仍需谨慎。
[CV004, CV006, CV007, CV009, CV025, CV019]8.2 同行背景说明,Neo Fusion 在拿出可比 CFS、Helion 或 TAE 的商业化证明之前,应当打折
同业组有价值,不是因为哪家公司能完美对标,而是因为它说明投资者给聚变公司高估值时究竟奖励什么。CFS 把巨额融资、明确点名的 Google 购电、400 MW ARC 电站概念,以及 Virginia 私人合同所有权模型放在一起。Helion 则有清楚的数十亿美元估值标记、Microsoft 协议,以及已进入场址开发的 Orion 电站。TAE 的资本规模小于 Neo Fusion 披露的资本底座,但它披露的治理深度、资本市场包装和伙伴支持的供应链工作都比 Neo Fusion 更多。即便在行业低位,First Light 较小一轮融资也说明,商业化包装一旦变薄,估值会多快被压缩。相比之下,Neo Fusion 现在更像一个由战略资金推动的国家项目节点,而不是市场定价的私营公司。这个差异不代表它没有价值;它说明今天任何估值都应低于那些客户、治理和电站开发证据更清楚的同业。[CV012, CV008, CV017, CV018, CV019, CV020]
| 公司 / 锚点 | 最新披露资本 / 估值 | 商业化证明 | 估值含义 | 限制 |
|---|---|---|---|---|
| Neo Fusion | RMB 14.5B / ~$2.03B 公开口径资本基础;无公开定价轮 | BEST 目标 2027 年;约 2030 年电力示范目标;未公开承购 | 战略意义可能很大,但公开证据太不透明,难以给出有把握的定价。 | 注册资本不等于估值,也不等于实缴现金。 |
| CFS | 累计私人融资 ~$2.923B;$863M Series B2 轮 | Google 200 MW 承购;Virginia ARC 400 MW 电站计划 | 把托卡马克商业化包装的门槛拉得很高。 | 融资额不等于股权估值,SPARC 仍要交付。 |
| Helion | 2025 年投后 $5.425B;2026 年分析师更新口径投后 $15.5B | Microsoft 承购;Orion 电站在建 | 说明市场愿意为明确商业化证明付溢价。 | 估值标记仍是私人口径,且偏激进。 |
| TAE | >$1.3B 累计私人融资;最新一轮 >$150M | Copernicus / Da Vinci 路线图;CFO、CRO、IR 以及 UKAEA 供应链信号 | 即便周期很长,治理和伙伴深度也能支撑估值。 | 物理路线不同,且仍处商业化前。 |
| First Light Fusion | 2026 年首关 £25M | 替代路线技术野心强,但融资规模更小 | 说明不同聚变平台之间分化可以很大。 | 它不是托卡马克,也不能直接对标中国国家支持口径。 |
可比组有意筛得很窄:重点是最能服务决策的清晰锚点,而不是数据集里的每一家聚变公司。
[CV004, CV018, CV019, CV023, CV024, CV025]| 因素 | 支撑证据 | 缺失证明 | 估值影响 |
|---|---|---|---|
| 战略赞助 | 表面资本规模和公开市场相关支持方 | 实缴现金和所有权经济性 | 支撑下行韧性,但不能直接支撑某一估值点。 |
| 技术路线图 | 2027 BEST 与 ~2030 发电目标 | 里程碑交付证明和应急预算 | 带来期权价值,也带来很大的时间折价。 |
| 客户 / 场址方模式 | 同业已有购电和电站选址先例 | Neo Fusion 尚未披露同等安排 | 这是 Neo 应低于高溢价同业语境定价的主因。 |
| 治理 / 融资包装 | 行业同业披露了 CFO、IR、电站所有权和合作伙伴结构 | Neo Fusion 公开资料仍然单薄 | 压低置信度,也可能压缩外部投资者定价。 |
本表拆出一个原因:Neo Fusion 可以看起来很有战略重要性,但现阶段仍应给估值打折。
[CV008, CV009, CV010, CV021, CV026, CV031]8.3 用宽区间看期权价值,比给点估值更诚实
Neo Fusion 的估值更适合被看作里程碑期权,而不是建立在可见收入上的现值模型。BEST 仍在瞄准 2027 年建成或首次等离子体里程碑,以及 2030 年前后的电力示范目标。如果这些节点兑现,上行空间可能很大,因为稳定清洁电力在任何能源转型中都有战略价值。但当前现值必须为仍然缺失的东西大幅折价:没有已定价融资轮、没有实缴资本桥接、没有 Neo Fusion 特定的场址经济性、没有已披露客户结构,也没有可与快速改善的替代方案清楚对比的公开电站成本模型。行业背景也强化了这种谨慎。聚变行业确实融到真金白银,但资本仍集中在少数头部公司,AI 现在吸走的投资者注意力也大得多。与此同时,Lazard 2026 年摘要中,可再生能源仍是新建电源里最便宜的基准,过去十年还在继续大幅降本。因此,数字上应采用宽的合成区间,而不是装作精准的单点标记。[CV009, CV010, CV011, CV013, CV014, CV015]
| 情景 | 示意性估值区间(十亿美元) | 核心假设 | 情景失效条件 |
|---|---|---|---|
| 悲观 | 0.8 - 1.8 | 公开口径资本实际可用性低于暗示;2027 年延误;没有客户或落地场址证明出现。 | 实缴现金进一步不透明,或里程碑延误超出当前路线图。 |
| 基准 | 1.8 - 3.0 | 战略支持仍是真实的,BEST 继续推进,但 Neo Fusion 仍缺少干净商业化和披露证明。 | 国家支持如果不能转化成更清楚的预算、客户或监管包,基准情景就失效。 |
| 乐观 | 3.0 - 6.0 | BEST 击中关键里程碑,实缴资本桥开始可见,Neo 在商业化包装上开始接近更成熟的同业。 | 里程碑落空,或拿不出首座电站经济性。 |
这些是合成的期权价值区间,不是观测到的市场价格,也不是 DCF 输出。区间被战略资本锚框住:低于 Helion 式溢价案例,高于纯研究项目下行。
[CV042, CV043, CV044, CV045, CV046, CV047]Neo Fusion 的任何数值区间都必然很宽,因为最干净的公开锚点是战略资本,不是市场价格。
这些数值是按十亿美元计的合成期权价值区间,来自注册资本锚点、同行商业化信号和里程碑不确定性;不是实际融资定价点。
[CV042, CV045, CV046, CV047, CV043]最大摇摆因素是现金到位证明、里程碑交付、客户打包方式和替代电力压力。
数值是方向性估值影响评分,不是字面美元变动。
[CV007, CV049, CV050, CV036, CV037]Neo Fusion 在战略背书和长期期权价值上得分较好,但当前披露和商业化证明偏弱。
分数是基于保留下来的公开证据综合得出的 0-10 序数尽调判断,不是管理层提供的 KPI。
[CV004, CV043, CV044, CV031, CV048]8.4 建议可以改善,但要等公司经济账变得可读
真正会改变判断的东西很直接。第一,公开材料或尽调资料室要说明,披露的资本底座有多少已经实缴,以及这笔钱如何接上 2027 年,再接上 2030 示范叙事。第二,Neo Fusion 需要拿出同业现在使用的某种商业化证据:真实的场址逻辑、首个客户或购电结构,至少也要把从 BEST 走向一座可融资、可运营电站的路线定义得更清楚。第三,治理和监管包装必须清楚到足以让外部人相信,公司能从战略科研项目转成可融资基础设施。在此之前,区间会继续很宽,置信度也会被封顶。下调触发因素同样清楚:2027 明显延误、可用资本弱于假设,或替代清洁电力选择持续变便宜时,公司仍然不透明。投资者不应把国家重要性和即时可投资性混为一谈;二者有关联,但不是同一个概念。[CV044, CV049, CV050, CV051, CV021, CV026]
| 尽调事项 | 当前公开状态 | 重要性 | 估值影响 |
|---|---|---|---|
| 实缴资本和出资节奏 | 公开资料未证实 | 把名义注册资本和真实现金跑道区分开。 | 若披露,可收窄区间并提高信心。 |
| 通往 2027 和 2030 的整合里程碑预算 | 未公开 | 检验当前支持能否接到创造价值的里程碑。 | 会显著降低情景不确定性。 |
| 首个落地场址 / 承购 / 电价逻辑 | Neo Fusion 尚无证明 | 商业化证明最能把 Helion、CFS 与其他项目拉开。 | 相对纯研究项目折价,可支撑一定溢价。 |
| 机器级安全和监管包 | 框架可见;公司包未公开 | 客户、场址方和融资方的信心都需要它。 | 会提升商业化可信度。 |
| 电站经济性和替代电源敏感性 | 无公开模型 | 要同低价可再生能源和其他稳定电源竞争,这一项不可缺。 | 既可能支撑估值,也可能大幅压缩估值。 |
| 股权结构表、治理和投资者权利结构 | 不透明 | 决定表面锚点是否反映可投资经济性。 | 直接影响入场吸引力和下行保护。 |
剩余大部分估值不确定性,并不在于核聚变是否可能重要,而在于 Neo Fusion 的具体经济性和控制权能否经得起投资尽调。
[CV007, CV044, CV049, CV050, CV051]8.5 图表材料
免责声明
这份报告元数据工件仅来自截至 2026-07-19 在 Neo Fusion 章节 YAML 中审阅过的公开证据。Neo Fusion 仍是私营公司,披露存在重大缺口,若干醒目数字指向注册资本,而不是清晰的市场定价股权或已证明可动用现金。因此,建议和估值判断高度受未披露的股权结构条款、预算节奏、安全规划和首座电站经济性的影响。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Neo Fusion's public company profiles and registry mirrors identify the business as founded on 19 May 2023 in Hefei, Anhui. | 中 | SO001, SO004 |
| CO002 | The operating legal entity commonly associated with Neo Fusion is 聚变新能(安徽)有限公司, which English-language databases translate as Neo Fusion or Fusion New Energy (Anhui). | 中 | SO004, SO003 |
| CO003 | Across public profiles the company is consistently located in Hefei, Anhui, China, even though registry and office addresses vary in their street-level detail. | 中 | SO001, SO003, SO004 |
| CO004 | Neo Fusion presents itself as a controlled-fusion developer focused on designing and building a tokamak-based fusion device intended for eventual commercial energy use. | 中 | SO003, SO001 |
| CO005 | Neo Fusion is publicly described as one of the principal construction and operating entities for the BEST compact fusion project in Hefei. | 中 | SO004, SO012 |
| CO006 | Fusion Energy Base records an initial 19 May 2023 registration-capital round of about $723.37 million backed by NIO, Anhui Province Energy, and another investor. | 中 | SO001 |
| CO007 | Fusion Energy Base records a 23 July 2024 additional registration-capital round of about $1.306 billion funded by CNPC and HFIPS. | 中 | SO001 |
| CO008 | Chinese business media reported the same 2024 corporate-registration event as an increase in registered capital from RMB 5.0 billion to RMB 14.5 billion. | 中 | SO016, SO017 |
| CO009 | Taken together, the publicly disclosed registration-capital base amounts to roughly $2.03 billion or RMB 14.5 billion after the 2024 step-up. | 中 | SO001, SO016 |
| CO010 | Public profiles describe Neo Fusion as majority owned by state-linked entities including CNPC and Hefei Science Island / HFIPS affiliates. | 中 | SO003, SO016 |
| CO011 | 36Kr reported that CNPC's Kunlun Capital invested RMB 2.9 billion for a 20% stake in the 2024 ownership reset. | 中 | SO016 |
| CO012 | 36Kr reported that Hefei Science Island and Hefei Chanto New Energy each held roughly 20.5% after the 2024 capital increase. | 中 | SO016 |
| CO013 | Chinese business coverage describes NIO-linked Weiju Technology (Anhui) as a continuing minority shareholder in Neo Fusion. | 中 | SO016, SO017 |
| CO014 | By early 2026, public Chinese media identified Yan Jianwen as Neo Fusion's chairman and principal public spokesperson. | 中 | SO004, SO019 |
| CO015 | Baidu Baike identifies Huang Suzhen as the company's legal representative and a core operating executive. | 中 | SO004 |
| CO016 | 36Kr cited vice general manager Liu Zhihong as a public advocate of the company's “research institute plus commercial company” commercialization model. | 中 | SO016 |
| CO017 | The 2024 registration-capital increase was accompanied by multiple personnel changes according to Yicai and 36Kr. | 中 | SO017, SO016 |
| CO018 | Compared with Western fusion peers, Neo Fusion publishes much thinner board, executive, and organizational disclosure in accessible English-language materials. | 中 | SO001, SO003, SO020 |
| CO019 | BEST began construction in 2023 and by late 2025 had entered final assembly with the main building completed and major component pre-assembly underway. | 中 | SO002, SO015, SO005 |
| CO020 | A 400-ton Dewar base installation in October 2025 marked a major precision-assembly milestone for BEST. | 高 | SO005, SO006, SO007 |
| CO021 | EUROfusion and Chinese institutional sources place BEST's first plasma target in 2027. | 中 | SO011, SO002 |
| CO022 | Chinese official reporting says BEST is intended to demonstrate net fusion power gain and electricity generation around 2030. | 高 | SO009, SO010 |
| CO023 | BEST is repeatedly framed as the intermediate engineering bridge between EAST and a future Chinese fusion engineering demo reactor such as CFEDR. | 中 | SO002, SO015 |
| CO024 | The BEST research plan describes the machine as a compact high-field tokamak designed to study burning-plasma physics at fusion gain above Q>5. | 中 | SO011 |
| CO025 | The same research plan positions BEST as a deuterium-tritium burning-plasma platform and a testbed for materials, blanket modules, and fuel-cycle systems. | 中 | SO011 |
| CO026 | Neo Fusion sits inside a Hefei cluster that already includes EAST, CRAFT, Science Island laboratories, and the Hefei Comprehensive National Science Centre. | 中 | SO024, SO022, SO025 |
| CO027 | EAST set a 1,066-second high-confinement plasma record in January 2025, reinforcing the technical pedigree that BEST inherits. | 高 | SO022, SO023 |
| CO028 | CRAFT was described in 2024 as a national fusion-technology research facility under construction in Hefei to validate key subsystems for future reactors. | 中 | SO024 |
| CO029 | Public company and media profiles describe a Hefei-based fusion industrial ecosystem linking more than 200 supply-chain participants and nearly 60 local fusion firms. | 中 | SO004, SO019 |
| CO030 | At the 2026 Hefei fusion conference, Neo Fusion's chairman said participating exhibitors covered materials, components, and engineering equipment across the full chain. | 中 | SO019 |
| CO031 | Neo Fusion-linked conference remarks described plans to work with 10 to 20 universities after helping launch a fusion science and engineering school in Hefei. | 中 | SO019 |
| CO032 | Baidu Baike says the company entered Forbes China's 2024 innovation ranking and the 2026 Hurun Global Unicorn list. | 中 | SO004 |
| CO033 | Neo Fusion does not yet present the kind of robust English-language public website or investor-relations surface common among global private fusion peers. | 中 | SO001, SO003 |
| CO034 | No accessible source in this review disclosed Neo Fusion revenue, signed customers, or current employee count with the rigor expected for commercial diligence. | 中 | SO001, SO003, SO004 |
| CO035 | The gap between 2027 first plasma and a 2030 power-generation demonstration leaves Neo Fusion with an unusually compressed engineering schedule. | 中 | SO011, SO009, SO015 |
| CO036 | Neo Fusion's public “unicorn” narrative is grounded in registration-capital and state-backed ownership rather than a transparent priced venture round. | 中 | SO001, SO004, SO016 |
| CO037 | Accessible public sources provide only a partial and sometimes secondary-account breakdown of the six-shareholder structure, rather than a full primary cap-table filing. | 中 | SO004, SO016 |
| CO038 | Baidu Baike cites paid-in capital figures, but accessible primary project disclosures do not explain how much of the registered-capital base has actually been deployed into BEST. | 中 | SO004, SO012 |
| CO039 | Public materials emphasize engineering milestones and industrial ecosystem building, but they do not disclose a date for first commercial revenue. | 中 | SO001, SO019 |
| CM001 | For Neo Fusion, the economically relevant market is not current scientific spending but future purchases of firm low-carbon electricity and the enabling fusion-plant supply chain. | 中 | SM005, SM022 |
| CM002 | ITER frames fusion as a dense, dispatchable, carbon-free energy source with fuel abundance and fission-like power density, making it a potential substitute for firm generation rather than only intermittent renewables. | 中 | SM005 |
| CM003 | Precedence Research projects a global nuclear-fusion market of roughly $472 billion by 2030. | 中 | SM001 |
| CM004 | Maximize Market Research projects the fusion-energy market to approach $573 billion by 2032 from a 2025 base of about $347 billion. | 中 | SM002 |
| CM005 | Industry Today cites an estimate of roughly $497 billion for the fusion-energy market by 2030. | 中 | SM003 |
| CM006 | EIN News cites a 2040 fusion-market estimate of about $840 billion, illustrating how top-down forecasts expand sharply at longer horizons. | 中 | SM004 |
| CM007 | Published fusion market-size estimates vary widely because they blend different boundaries: some model electricity value, some include supply-chain and technology-platform value, and some extend assumptions out to 2040. | 中 | SM001, SM002, SM004 |
| CM008 | NucNet reported that total disclosed fusion-industry funding reached about $14.24 billion in 2026. | 中 | SM010 |
| CM009 | Fusion Industry Association reporting says more than $2.5 billion was invested into the fusion industry in the year before its 2025 report. | 中 | SM012 |
| CM010 | The most plausible first external buyers for successful fusion plants are national or regional utilities seeking firm clean generation for the grid. | 中 | SM005, SM007 |
| CM011 | Large industrial loads with decarbonization pressure, especially steel, chemicals, hydrogen, and data-center-adjacent users, are plausible second-wave buyers for fusion power. | 中 | SM007, SM008, SM009 |
| CM012 | Governments, state labs, and public-private programs remain critical early payers because first-of-a-kind fusion assets require policy support, licensing frameworks, and mission-driven capital. | 中 | SM020, SM018, SM016 |
| CM013 | Gartner expects data-center electricity demand to double by 2030, reinforcing the macro case for new firm generation sources. | 中 | SM008 |
| CM014 | S&P Global summarized IEA work showing data-center power demand could double by 2030, another sign that dispatchable low-carbon power could command premium value. | 中 | SM009 |
| CM015 | Chinese official reporting says BEST is targeting net power gain and electricity-generation demonstration around 2030, implying China is trying to create a domestic fusion demonstration market ahead of broad commercialization. | 高 | SM022, SM023 |
| CM016 | CAS reporting portrays Hefei as an emerging fusion-industrial cluster, which lowers commercialization frictions for Neo Fusion inside China even before international market access is proven. | 中 | SM029 |
| CM017 | The plausible adoption sequence for Neo Fusion is experiment and burning-plasma proof, then a state-backed demonstration plant, and only after that broader utility procurement. | 中 | SM024, SM022 |
| CM018 | AIP and FIA sources highlight that a distinct fusion regulatory framework matters because fusion projects face different licensing logic than fission plants. | 中 | SM020, SM019 |
| CM019 | Fusion remains capital-intensive because the sector needs bespoke facilities, superconducting systems, fuel-cycle engineering, and long-duration testing before revenue begins. | 中 | SM027, SM010 |
| CM020 | Crunchbase described fusion funding as cyclical and fragile in 2024, underscoring the mismatch between long commercialization timelines and shorter investor patience. | 中 | SM027 |
| CM021 | From a buyer perspective, fusion competes most directly with other firm clean power options such as advanced fission, geothermal, and long-duration storage rather than with solar alone. | 中 | SM021, SM005, SM007 |
| CM022 | Fusion is more likely to complement rather than replace renewables because intermittent resources plus storage will still set many grid economics through the 2030s. | 中 | SM007, SM021 |
| CM023 | Early fusion budget owners are likely to be governments, state-owned utilities, and strategic industrial buyers rather than merchant power developers. | 中 | SM022, SM020, SM016 |
| CM024 | The United States, China, the United Kingdom, Germany, and the wider ITER / EU ecosystem appear to be the most active regions in funding or licensing commercial fusion pathways. | 中 | SM023, SM017, SM016 |
| CM025 | Private-fusion narratives emphasize capital raised and prototype milestones, whereas public programs emphasize enabling infrastructure, regulation, and long-duration engineering capability. | 中 | SM028, SM026, SM012 |
| CM026 | Because fusion could eventually provide both electricity and high-quality heat, adjacent markets such as hydrogen, e-fuels, and large industrial-process heat sit inside the long-run opportunity set. | 中 | SM005, SM002 |
| CM027 | Top-down market reports likely overstate the investable near-term market because they assume technical success, regulatory approval, and broad grid adoption years before any private fusion company has commercial output. | 中 | SM001, SM002, SM027 |
| CM028 | No accessible public source in this review quantified Chinese fusion electricity demand specifically for BEST-scale plants or a first-wave domestic deployment fleet. | 中 | SM022, SM024 |
| CM029 | No accessible source published a reliable LCOE forecast specific to Neo Fusion’s post-BEST commercial pathway. | 中 | SM024, SM021 |
| CM030 | Accessible public sources discuss the fusion supply chain extensively, but they do not cleanly separate equipment value pools from long-run generated-electricity value for Neo Fusion. | 中 | SM014, SM029 |
| CM031 | Because BEST is embedded in a Chinese national program, Neo Fusion’s export and collaboration options are likely to face tighter geopolitical scrutiny than those of Western private peers. | 中 | SM026, SM016 |
| CM032 | Tokamaks remain the most institutionally supported fusion configuration globally, with ITER, EAST, BEST, and multiple private developers all pursuing variants of magnetic confinement. | 中 | SM006, SM024, SM013 |
| CM033 | Fusion received growing policy attention in 2025-2026 through reports, budget requests, and new action plans, but policy support is still more about de-risking than about near-term commercial demand. | 中 | SM015, SM030, SM017 |
| CM034 | If Neo Fusion reaches its 2030 demonstration target, it would be arriving just as major economies are searching for post-renewables firm power options, but long before a mature merchant fusion market exists. | 中 | SM022, SM008, SM007 |
| CM035 | The company’s first market is almost certainly institutional power buyers rather than retail consumers or distributed-energy customers. | 中 | SM005, SM022 |
| CM036 | Fusion market development will look more like heavy infrastructure deployment than software adoption because permitting, site preparation, components, and fuel-cycle engineering dominate the path to revenue. | 中 | SM027, SM020, SM025 |
| CM037 | The coexistence of multi-hundred-billion-dollar TAM projections with still-precommercial hardware programs is a central contradiction of the fusion market today. | 中 | SM001, SM027, SM010 |
| CM038 | No buyer segment can underwrite meaningful fusion demand without proof of sustained burning plasma, cost, licensing, and uptime performance. | 中 | SM024, SM020 |
| CP001 | Neo Fusion’s most relevant private comparators are Commonwealth Fusion Systems, Helion, TAE Technologies, Zap Energy, and Tokamak Energy. | 中 | SP001, SP011, SP017, SP021, SP026 |
| CP002 | On the public-program side, BEST competes for attention and engineering leadership with ITER-linked tokamak work and other national tokamak efforts rather than with small software-like startups. | 中 | SP029, SP030 |
| CP003 | Commonwealth Fusion Systems is the clearest private tokamak comparator because it is building SPARC and positioning ARC as a commercial grid plant. | 中 | SP001, SP003 |
| CP004 | CFS disclosed an $863 million Series B2 round in 2025 to accelerate SPARC and ARC. | 中 | SP002 |
| CP005 | CFS has already disclosed strategic customer-style traction through its Google partnership around future fusion power offtake. | 中 | SP004, SP010 |
| CP006 | CFS has secured radioactive-materials licensing progress for SPARC, giving it more visible regulatory traction than Neo Fusion currently discloses publicly. | 中 | SP005 |
| CP007 | Independent and company sources show CFS further along on visible prototype-construction milestones than Neo Fusion in terms of public technical transparency. | 中 | SP008, SP009, SP007 |
| CP008 | Helion is pursuing a field-reversed configuration and direct-electricity concept rather than a tokamak, making it the most visible architecture alternative to BEST among top-funded peers. | 中 | SP012, SP011 |
| CP009 | Helion disclosed a $425 million Series F in 2025 after earlier billion-dollar-scale financing commitments. | 中 | SP013 |
| CP010 | Helion has stronger public customer proof than Neo Fusion because it announced a Microsoft-linked fusion power purchase agreement. | 中 | SP016 |
| CP011 | Helion publicized 2026 Polaris milestones including deuterium-tritium operation and 150 million degree temperatures, which give it a clearer near-term technical narrative than Neo Fusion currently provides in English. | 中 | SP014, SP015 |
| CP012 | TAE represents the leading beam-driven FRC alternative, emphasizing an aneutronic long-run vision and a commercial roadmap distinct from tokamak confinement. | 中 | SP017, SP019 |
| CP013 | TAE disclosed a $150 million funding round, which is meaningful but still far smaller than Neo Fusion’s state-backed registered-capital base or the largest CFS and Helion financings. | 中 | SP018 |
| CP014 | TAE argues it has delivered physics and design improvements that could materially reduce future plant costs, positioning cost structure as its key differentiation. | 中 | SP020 |
| CP015 | Zap Energy competes on a lower-component-count Z-pinch architecture that avoids large superconducting magnets. | 中 | SP021, SP025 |
| CP016 | Zap disclosed $130 million of fresh capital tied to demo-system progress, far below Neo Fusion’s implied capital base but potentially far more capital-efficient per step. | 中 | SP023 |
| CP017 | Zap has public DOE milestone recognition around pilot-plant preconceptual design, giving it visible U.S. government validation. | 中 | SP022 |
| CP018 | Zap also publicized first-plasma and scientific-breakeven-oriented milestones on its next-generation device, maintaining competitive technical momentum. | 中 | SP024 |
| CP019 | Tokamak Energy is a relevant European magnetic-confinement peer combining spherical-tokamak ambitions with a strong high-temperature superconducting magnet narrative. | 中 | SP026, SP027 |
| CP020 | Tokamak Energy raised $125 million in late 2024, underscoring that even serious Western peers still operate on disclosed funding rounds far smaller than Neo Fusion’s registered-capital base. | 中 | SP028 |
| CP021 | Neo Fusion’s standout competitive advantage is the scale and strategic character of its state-backed capitalization. | 中 | SP002, SP013, SP018, SP023, SP028 |
| CP022 | Neo Fusion’s clearest weakness is the absence of any publicly disclosed customer contracts, PPAs, or host-site offtake relationships comparable to Helion or CFS. | 中 | SP016, SP004 |
| CP023 | Compared with Western peers that publish product pages, FAQs, milestone blogs, and funding releases, Neo Fusion remains much more opaque to outside investors. | 中 | SP001, SP011, SP017, SP021, SP026 |
| CP024 | No serious fusion peer currently publishes commercial pricing that can be used as a clean apples-to-apples comparison for Neo Fusion. | 中 | SP003, SP012, SP019, SP025 |
| CP025 | Switching costs are low today because buyers are still evaluating architectures before signing long-term build or power contracts. | 中 | SP004, SP016, SP029 |
| CP026 | Once a utility or host site commits to a specific fusion design, switching costs will become high due to licensing, siting, supply-chain, and integration lock-in. | 中 | SP005, SP003, SP022 |
| CP027 | In the 2020s and early 2030s, sophisticated buyers can effectively multi-home by running diligence across multiple fusion vendors in parallel because no platform is yet dominant. | 中 | SP004, SP016, SP030 |
| CP028 | Neo Fusion’s access to HFIPS, EAST, CRAFT, and the Hefei cluster is a real supply and talent advantage, while CFS’s HTS magnet ecosystem and Tokamak Energy’s magnet specialization are rival strengths. | 中 | SP007, SP027, SP030 |
| CP029 | CFS and Helion currently look strongest on disclosed commercial-readiness signals because each has both large financing and visible prospective customer relationships. | 中 | SP002, SP004, SP013, SP016 |
| CP030 | Near-term commoditization risk is low because fusion developers still compete on radically different physics, component stacks, and validation pathways. | 中 | SP003, SP012, SP019, SP025 |
| CP031 | Over a longer horizon, tokamak-adjacent supply chains could partially standardize around magnets, blankets, and balance-of-plant systems, which would narrow Neo Fusion’s moat if BEST succeeds. | 中 | SP006, SP027, SP029 |
| CP032 | Future buyers will likely select fusion vendors based on some mix of milestone proof, regulatory certainty, capital backing, and route-to-market partnerships rather than on published price lists. | 中 | SP004, SP016, SP005, SP022 |
| CP033 | BEST’s strongest competitive frame is as a state-backed, engineering-led tokamak bridge from EAST to a demo reactor, not as a copy of any single Western startup playbook. | 中 | SP029, SP003, SP031 |
| CP034 | The 2025-2026 landscape shifted most through CFS’s major financing and partnership announcements, Helion’s Polaris disclosures, Zap’s DOE-linked milestones, and Neo Fusion’s own BEST buildout. | 中 | SP002, SP014, SP022 |
| CP035 | Helion and TAE pose the biggest architecture threat to BEST because they promise different cost and operating profiles if their non-tokamak physics translate commercially. | 中 | SP012, SP019, SP020 |
| CP036 | Until Neo Fusion can show a customer-development path or sponsor-backed offtake logic, competitor comparisons will continue to favor better-disclosed peers. | 中 | SP016, SP004, SP032 |
| CP037 | Neo Fusion’s disclosure gap is not cosmetic; it weakens trust relative to peers that already expose milestones, customers, and technical pathways in public detail. | 中 | SP011, SP001, SP017, SP021 |
| CI001 | Neo Fusion is best treated as a pre-revenue engineering company rather than an operating power business. | 中 | SI002, SI003 |
| CI002 | The implied primary future revenue stream is electricity sales from a later fusion demonstration or commercial plant rather than software, services, or recurring subscriptions. | 中 | SI007, SI008 |
| CI003 | Public materials imply potential secondary monetization through project development, strategic partnerships, or supply-chain participation, but none of these streams are contractually defined. | 中 | SI006, SI003 |
| CI004 | No public source in this review disclosed Neo Fusion pricing, tariff, $/MWh target, or customer contract economics. | 中 | SI002, SI003 |
| CI005 | The current go-to-market motion appears to be state-backed engineering execution first, followed by demonstration-plant sponsorship and only later utility or industrial offtake. | 中 | SI007, SI006 |
| CI006 | No public sales-efficiency, CAC, payback, pipeline, or conversion metrics are available because Neo Fusion has not disclosed a commercial selling motion. | 中 | SI001, SI002 |
| CI007 | Neo Fusion’s cost structure should be expected to be dominated by heavy engineering, facilities, superconducting systems, plasma hardware, and highly specialized labor. | 中 | SI008, SI007 |
| CI008 | No public gross-margin, operating-margin, or EBITDA metrics are disclosed for Neo Fusion. | 中 | SI003, SI002 |
| CI009 | No public unit-economics model exists for Neo Fusion in terms of cost per shot, cost per megawatt, or future electricity margin. | 中 | SI003, SI008 |
| CI010 | Fusion Energy Base records an initial registration-capital event of about $723.37 million on 19 May 2023. | 中 | SI001 |
| CI011 | Fusion Energy Base records an additional registration-capital event of about $1.306 billion on 23 July 2024. | 中 | SI001 |
| CI012 | Chinese media described the same 2024 event as an increase in registered capital from RMB 5.0 billion to RMB 14.5 billion. | 中 | SI004, SI005 |
| CI013 | The combined registered-capital base after the 2024 increase is roughly RMB 14.5 billion or about $2.03 billion equivalent. | 中 | SI001, SI004 |
| CI014 | Accessible public sources do not prove how much of the registered-capital base is fully paid in, ring-fenced, and available to fund BEST construction. | 中 | SI003, SI001 |
| CI015 | Neo Fusion’s current burn rate is not publicly disclosed. | 中 | SI002, SI003 |
| CI016 | Because paid-in capital, burn, and capex timing are undisclosed, a reliable runway estimate cannot be derived from public information. | 中 | SI003, SI008 |
| CI017 | Registered capital should not be treated as identical to a priced venture valuation or to immediately deployable cash. | 中 | SI001, SI009 |
| CI018 | Public milestone timing implies that any meaningful revenue is more likely to arrive after, not before, a successful 2030-era power demonstration. | 中 | SI007, SI008 |
| CI019 | No public debt, project-finance, or government-loan structure specific to Neo Fusion has been disclosed. | 中 | SI001, SI003 |
| CI020 | The public record does not prove that Neo Fusion’s current capital base is sufficient through first commercial demonstration without further funding. | 中 | SI008, SI010 |
| CI021 | NucNet reported total sector funding of about $14.24 billion in 2026, highlighting how capital-intensive fusion remains even after record financing. | 中 | SI010 |
| CI022 | Crunchbase’s 2024 reality check showed that fusion funding can cool quickly when milestone narratives stop advancing. | 中 | SI009 |
| CI023 | Neo Fusion’s registered-capital base is larger than CFS’s disclosed 2025 Series B2 round but not directly comparable because one is a registration-capital stack and the other a priced financing. | 中 | SI001, SI014 |
| CI024 | Neo Fusion’s registered-capital base also exceeds Helion’s disclosed 2025 Series F round, but Helion has clearer customer and financing disclosures. | 中 | SI001, SI016, SI017 |
| CI025 | TAE, Zap, and Tokamak Energy each disclosed materially smaller financing rounds than Neo Fusion’s public capital base. | 中 | SI020, SI022, SI023 |
| CI026 | Even the best-known fusion PPAs from Helion and CFS do not reveal enough detail to build a credible pricing benchmark for Neo Fusion. | 中 | SI017, SI015, SI024 |
| CI027 | Helion’s Orion and CFS’s ARC show that leading peers are already packaging future power plants as customer-facing assets, whereas Neo Fusion still packages BEST primarily as an engineering bridge. | 中 | SI018, SI015 |
| CI028 | State-backed capital lowers financing-gap risk for Neo Fusion relative to a normal venture startup, but it does not eliminate execution or capital-efficiency risk. | 中 | SI004, SI009 |
| CI029 | NIO is a publicly listed company that filed a 2024 Form 20-F, reinforcing that one early Neo Fusion backer has public-capital-markets access unlike a pure venture syndicate. | 中 | SI027, SI001 |
| CI030 | Lazard’s power-cost work is useful for framing alternatives but does not substitute for a Neo Fusion-specific cost model or LCOE view. | 中 | SI012 |
| CI031 | The most important missing private metrics are cash on hand, paid-in capital, burn, project budget, pricing assumptions, customer pipeline, and first-plant economics. | 中 | SI003, SI001 |
| CI032 | Because BEST is a first-of-a-kind compact burning-plasma device with materials, blanket, and fuel-cycle ambitions, its development path inherently implies large pre-revenue capex. | 中 | SI008 |
| CI033 | BEST itself is primarily a demonstration and engineering platform, not yet a revenue-generating commercial product. | 中 | SI008, SI007 |
| CI034 | Neo Fusion’s financial disclosure is materially thinner than even opaque Western peers because public information stops at registration-capital headlines and shareholder summaries. | 中 | SI001, SI002, SI019 |
| CI035 | The correct financial verdict is that Neo Fusion has unusual strategic financial support but remains impossible to underwrite conventionally on public data alone. | 中 | SI001, SI009, SI008 |
| CI036 | ANS coverage of industry funding growth still implies that first plants will require very large additional capital pools beyond current startup rounds. | 中 | SI013 |
| CI037 | Comparator sources from Helion and Zap show that leading peers increasingly package their businesses around named plants, program milestones, and customer-facing commercialization stories rather than around undifferentiated research spend. | 中 | SI028, SI029, SI030 |
| CI038 | World Nuclear News coverage of SPARC and ITER progress shows that even leading fusion programs continue to consume capital for long periods before customer revenue appears. | 中 | SI026 |
| CI039 | Comparator power-purchase announcements from CFS and Helion show that customer interest in firm fusion power can exist before commercial delivery, even if Neo Fusion has not yet disclosed equivalent demand signals. | 中 | SI015, SI017, SI025 |
| CI040 | Neo Fusion has not publicly disclosed site-hosting economics, revenue-sharing terms, or a utility settlement structure for any future plant. | 中 | SI003, SI001 |
| CI041 | No public source discloses a milestone budget bridging today’s capital base to 2027 first plasma and then to a 2030 demonstration. | 中 | SI008, SI001 |
| CI042 | The financial model likely depends on government or state-enterprise sponsorship before it can depend on ordinary merchant power revenue. | 中 | SI007, SI006 |
| CE001 | Neo Fusion’s present product is not a sellable power contract but the BEST compact fusion device under construction in Hefei. | 中 | SE001, SE002 |
| CE002 | BEST is intended to bridge from experimental fusion research into a power-oriented engineering demonstrator rather than remain a pure physics test stand. | 中 | SE005, SE007 |
| CE003 | Official Chinese reporting says BEST aims to demonstrate net fusion power gain and electricity generation around 2030. | 高 | SE005, SE006 |
| CE004 | BEST is designed to demonstrate actual burning deuterium-tritium plasma rather than only non-burning plasma operation. | 高 | SE003, SE005, SE007 |
| CE005 | EUROfusion describes BEST as a compact high-field tokamak under rapid construction at ASIPP in Hefei. | 中 | SE007 |
| CE006 | The BEST research plan says the machine is designed to explore burning-plasma physics at fusion gain above Q>5. | 高 | SE007, SE001 |
| CE007 | The same research plan positions BEST as a platform to test materials, blanket modules, and fuel-cycle systems needed for later pilot plants. | 中 | SE007 |
| CE008 | By April 2026 CAS reported the BEST main building completed and pre-assembly under way, confirming the project had moved beyond conceptual planning. | 中 | SE009 |
| CE009 | The installation of BEST’s Dewar base in October 2025 marked the first major main-machine assembly milestone. | 高 | SE003, SE004 |
| CE010 | HFIPS described the Dewar base as over 400 tons, about 18 meters in diameter, and 5 meters in height. | 中 | SE003 |
| CE011 | The Dewar acts as a high-vacuum thermal enclosure that insulates superconducting magnets operating near minus 269°C while confining plasma above 100 million°C. | 中 | SE003 |
| CE012 | HFIPS said the BEST device is scheduled for completion by the end of 2027. | 中 | SE003 |
| CE013 | EUROfusion separately said first plasma is targeted for 2027. | 中 | SE007, SE001 |
| CE014 | EAST sustained high-confinement plasma for 1,066 seconds in January 2025, setting a world record. | 高 | SE010, SE011 |
| CE015 | HFIPS said future fusion plants require stable high-efficiency operation for thousands of seconds, which is why EAST’s 1,066-second result matters. | 中 | SE010 |
| CE016 | Since 2006 EAST has served as an open testing platform for Chinese and international fusion researchers. | 中 | SE010 |
| CE017 | China joined ITER in 2006 and ASIPP is the primary institution for the Chinese mission, tying BEST to a larger international tokamak engineering base. | 中 | SE010, SE015 |
| CE018 | ITER is the world’s largest experimental tokamak and a magnetic-confinement plasma-physics experiment rather than a compact demonstration device like BEST. | 中 | SE015 |
| CE019 | CRAFT is a Hefei-based platform for developing and testing key components for fusion energy reactors. | 中 | SE012 |
| CE020 | CAS expected CRAFT to be finished in 2025, making it a near-term feeder asset for China’s fusion hardware stack. | 中 | SE012 |
| CE021 | In 2025 CAS said the CRAFT divertor prototype demonstrated a steady-state thermal load of 20 megawatts per square meter. | 中 | SE013 |
| CE022 | CAS also said the divertor coating design could theoretically raise tritium breeding ratio by more than 3 percent. | 中 | SE013 |
| CE023 | CAS reporting portrays Hefei as a frontier-industries cluster where EAST, CRAFT, and fusion manufacturing infrastructure are co-located. | 中 | SE014, SE012 |
| CE024 | Official 2026 reporting tied BEST to a planned Hefei “Fusion City” science and industrial zone, reinforcing that Neo Fusion is embedded in a place-based engineering buildout. | 中 | SE005, SE014 |
| CE025 | JET ended operations in December 2023 after deuterium-tritium campaigns that EUROfusion says laid groundwork for ITER and future plants. | 中 | SE016 |
| CE026 | EUROfusion says JET trained multiple generations of fusion scientists and engineers, illustrating the long human-capital cycle behind tokamak commercialization. | 中 | SE016 |
| CE027 | Wendelstein 7-X uses modular superconducting coils and targets plasma discharges up to 30 minutes to demonstrate steady-state stellarator operation. | 中 | SE017 |
| CE028 | W7-X is designed to avoid the large plasma current of a tokamak, making it a useful contrast to BEST’s tokamak path rather than a direct substitute. | 中 | SE017 |
| CE029 | MIT and CFS reported a 20-tesla large-scale HTS magnet test, a major proof point for compact high-field tokamak economics. | 中 | SE018 |
| CE030 | MIT researchers said the 20-tesla magnet breakthrough changed estimated fusion-reactor cost per watt by roughly a factor of 40 by enabling much smaller devices. | 中 | SE018 |
| CE031 | CFS’s PIT VIPER cable shows that commercial tokamak peers are solving pulsed HTS-cable challenges alongside steady-state magnet problems. | 中 | SE019 |
| CE032 | CFS says its ARC plant is intended to put power on the grid in the early 2030s, providing a concrete commercial benchmark that BEST has not yet matched in public packaging. | 中 | SE020, SE021 |
| CE033 | CFS disclosed a 200 MW Google offtake agreement for its first ARC plant, showing a level of product packaging and buyer signaling Neo Fusion does not yet publish. | 中 | SE021 |
| CE034 | CFS has also publicized radioactive-materials licensing progress for SPARC, highlighting how much more visible some peer trust and compliance surfaces already are. | 中 | SE022 |
| CE035 | Helion is pursuing a direct-electricity field-reversed-configuration path rather than a tokamak, showing that BEST competes within a broader architecture race. | 中 | SE023 |
| CE036 | TAE says it has published more than 350 peer-reviewed posters and papers, signaling a much heavier public research surface than Neo Fusion currently exposes. | 中 | SE024 |
| CE037 | Zap’s public research library enumerates multiple peer-reviewed 2023-2025 papers, again giving outside observers more direct technical visibility than Neo Fusion offers. | 中 | SE026 |
| CE038 | Zap’s DOE-certified Century platform shows another peer publicly exposing repetitive pulsed-power, liquid-metal-wall, and reliability-testing milestones that Neo Fusion does not yet publish in similar detail. | 中 | SE027 |
| CE039 | The 2026 DOE fusion science and technology roadmap still treats materials, tritium breeding and handling, supply chains, workforce, and practical adoption as open commercialization challenges. | 中 | SE029 |
| CE040 | BusinessCraft’s 2026 comparison of major private fusion companies still framed every leading peer as pre-commercial and timeline-sensitive, despite stronger public packaging than Neo Fusion. | 中 | SE028 |
| CE041 | Neo Fusion’s accessible English-language technical disclosure remains thinner than that of Western private peers that publish detailed research libraries, magnet papers, and product pages. | 中 | SE002, SE024, SE026, SE018 |
| CE042 | Accessible public sources do not disclose uptime, availability, maintenance-cycle, or support-model metrics for BEST as an operating asset. | 中 | SE001, SE002, SE007 |
| CE043 | Accessible public materials reviewed here do not show a Neo Fusion trust-center equivalent, named quality certification set, or public safety-operating manual for BEST. | 中 | SE002, SE001, SE030 |
| CE044 | Neo Fusion chairman Yan Jianwen publicly argued in 2026 that fusion needs synchronized legislation, standards, and talent development. | 中 | SE030 |
| CE045 | BEST’s public technical materials confirm fuel-cycle and blanket-module testing ambition, but they do not disclose a detailed operating plan for those systems. | 中 | SE007, SE005 |
| CE046 | No accessible source in this review disclosed a plant-level tritium handling, storage, or waste-management operating plan for BEST. | 中 | SE007, SE005, SE002 |
| CE047 | The visible product today is an engineering program with major hardware dependencies rather than a deployable commercial plant package. | 中 | SE009, SE012, SE007 |
| CU001 | No accessible public source in this review shows Neo Fusion already serving commercial electricity customers. | 中 | SU001, SU002, SU012 |
| CU002 | At Neo Fusion’s current stage, the closest thing to a customer base is its sponsor-and-stakeholder set: state investors, industrial backers, and future utility channels. | 中 | SU001, SU002, SU006 |
| CU003 | CnEVPost and a Reuters-syndicated report say NIO acquired a 19.9 percent stake in Neo Fusion while NIO Capital took 10.1 percent. | 中 | SU003, SU004 |
| CU004 | NIO told CnEVPost that the project would be funded in phases and attract additional strategic and financial investors over time. | 中 | SU003 |
| CU005 | NIO publicly framed fusion as strategically significant for carbon neutrality, energy security, and its long-term industry influence. | 中 | SU003 |
| CU006 | The Reuters-syndicated coverage said Anhui government-owned energy companies and investment arms controlled about 50 percent of Neo Fusion at founding. | 中 | SU004, SU003 |
| CU007 | Fusion Energy Base and Chinese business reporting show CNPC-linked capital joined Neo Fusion in the 2024 recapitalization. | 中 | SU001, SU006 |
| CU008 | The same 2024 recapitalization brought HFIPS-linked capital into the ownership structure, tightening the link between sponsor base and technical institution. | 中 | SU001, SU006 |
| CU009 | TechNode reported that NIO and Anhui Province Energy Group created a joint venture to build 1,000 battery-storage-integrated charging facilities in Anhui. | 中 | SU005 |
| CU010 | That same JV was described as a way to improve the flexibility and resilience of the power grid by balancing electricity demand and supply. | 中 | SU005 |
| CU011 | SASAC describes State Grid as the world’s largest utility enterprise, serving over 1.1 billion people across 88 percent of China’s land area. | 中 | SU009 |
| CU012 | China Southern Power Grid’s English site emphasizes power-grid development, smart-grid, energy-storage, EV, and international-cooperation functions. | 中 | SU010 |
| CU013 | China’s 2026-2030 new-type energy-system plan treats energy as national infrastructure tied to industrial competitiveness and national security. | 中 | SU011 |
| CU014 | The same plan emphasizes integrated generation, transmission, storage, distribution, and demand-side flexibility, favoring large institutional grid buyers over retail end users. | 中 | SU011 |
| CU015 | Official Chinese reporting on BEST implies the first “market” is a state-backed demonstration environment that can absorb engineering risk before merchant sales begin. | 高 | SU012, SU013 |
| CU016 | The most plausible first paying external buyers for successful fusion plants are large utilities or grid-linked public entities rather than households. | 中 | SU014, SU009, SU010 |
| CU017 | Heavy industry, hydrogen projects, and data-center operators are more plausible second-wave buyers after utility-backed first plants de-risk the technology. | 中 | SU015, SU016, SU017 |
| CU018 | Gartner expects data-center electricity demand to double by 2030, reinforcing the future value of firm clean power buyers. | 中 | SU016 |
| CU019 | S&P Global summarized IEA work showing data-center power demand could double by 2030 on AI growth. | 中 | SU017 |
| CU020 | WRI’s clean-energy charts show persistent growth in clean-power demand, but they also imply future customers will compare fusion against many lower-risk alternatives. | 中 | SU015 |
| CU021 | CFS publicly disclosed a 200 MW Google offtake agreement for its first ARC plant, proving that future fusion customers can sign before commercial delivery. | 高 | SU018, SU019, SU020 |
| CU022 | Helion announced Microsoft as the first customer for its planned fusion power plant in 2023, another benchmark for pre-delivery customer proof. | 中 | SU021 |
| CU023 | Neo Fusion has not publicly disclosed any comparable PPA, offtake framework, or anchor-host agreement. | 中 | SU001, SU002, SU012 |
| CU024 | Neo Fusion’s strongest public demand proof today is stakeholder alignment and state-capital commitment, not end-customer deployment. | 中 | SU003, SU006, SU012 |
| CU025 | No accessible public source disclosed active-customer counts, deployment counts, utilization, or revenue-bearing site numbers for Neo Fusion. | 中 | SU001, SU002, SU012 |
| CU026 | No public source disclosed NRR, GRR, churn, contract renewal rates, or contract lengths for Neo Fusion. | 中 | SU001, SU002 |
| CU027 | The only usable durability proxy today is continued sponsor continuity from 2023 through the 2024 recapitalization and into 2026 public reporting. | 中 | SU003, SU001, SU007 |
| CU028 | Neo Fusion’s current demand-side support base is highly concentrated because the company relies on a small number of state and strategic backers rather than a diversified customer book. | 中 | SU003, SU006, SU001 |
| CU029 | Because first demand likely comes through state-backed demonstration pathways, Neo Fusion is unusually dependent on policy continuity and sponsor priorities. | 中 | SU011, SU012, SU026 |
| CU030 | Fusion buyer procurement will be slow because utilities must weigh licensing, capex, host-site integration, and very long delivery timelines before signing. | 中 | SU014, SU020, SU024 |
| CU031 | A realistic land-and-expand path for Neo Fusion runs from sponsor-funded demonstration to utility-hosted first plant, then to industrial or data-center-adjacent offtake. | 中 | SU012, SU015, SU020 |
| CU032 | Neo Fusion is more likely to target domestic Chinese customers first than international buyers because BEST is embedded in a national program and local industrial cluster. | 中 | SU012, SU008, SU026 |
| CU033 | The Fusion Report explicitly excludes Chinese fusion-company private funding from its 2026 commercial-fusion tally because the market is opaque and public/private lines are blurry. | 中 | SU024 |
| CU034 | The Fusion Report’s broader funding coverage still frames the sector around capital raised and not around deployed customer fleets, underscoring how early the commercial market remains. | 中 | SU025 |
| CU035 | Crunchbase’s fusion-funding critique underscores that long timelines can outlast investor patience, which also affects prospective anchor customers waiting for proof. | 中 | SU023 |
| CU036 | NIO’s 2024 annual report confirms that one early Neo Fusion backer is itself a public-market issuer, which increases visibility into at least one sponsor’s capital-market access. | 中 | SU022 |
| CU037 | The scale and infrastructure role of State Grid suggest Neo Fusion’s future buyers will be institutional utilities, not consumer-facing retail energy accounts. | 中 | SU009, SU014 |
| CU038 | China Southern Power Grid’s emphasis on smart grid, storage, EV, and international cooperation makes it a plausible future integration channel if fusion reaches grid relevance in southern China. | 中 | SU010, SU011 |
| CU039 | Public evidence can identify likely buyer segments and strategic stakeholders, but not a credible named list of future paying Neo Fusion offtakers. | 中 | SU001, SU002, SU012 |
| CU040 | No review, procurement, or case-study source in this review exposed user-satisfaction data for Neo Fusion because the company has not yet deployed a commercial energy product. | 中 | SU001, SU002 |
| CU041 | Neo Fusion’s buyer map is therefore layered: state sponsors pay now, utilities and grids are likely first external offtakers later, and industrial or digital-load buyers follow after de-risking. | 中 | SU001, SU009, SU015, SU016 |
| CR001 | The Fusion Report said in mid-2026 that Q>1 had not yet been demonstrated in any of the commercial fusion systems then being built and probably would not be for at least another two and a half years. | 中 | SR015 |
| CR002 | POWER noted that the National Ignition Facility achieved a celebrated fusion-energy milestone in 2022, but that event does not remove the challenge of making magnetic-confinement fusion commercially viable. | 中 | SR029 |
| CR003 | BEST’s public roadmap still treats Q>5, burning-plasma operation, and 2030 electricity generation as targets rather than delivered results. | 高 | SR002, SR004 |
| CR004 | Public sources place BEST completion or first plasma around 2027, making any construction or subsystem slippage immediately material to the 2030 narrative. | 高 | SR003, SR002 |
| CR005 | HFIPS said future fusion plants need stable high-efficiency operation for thousands of seconds, underscoring how far even EAST’s 1,066-second record remains from a power-plant operating regime. | 中 | SR005 |
| CR006 | EAST’s 1,066-second pulse is a powerful pedigree signal, but it also reminds investors that long-duration plasma stability remains a core gating variable. | 中 | SR005, SR002 |
| CR007 | China’s Nuclear Safety Law covers nuclear facilities, nuclear materials, and radioactive waste, showing that a fusion operator cannot assume a regulation-free path. | 中 | SR007 |
| CR008 | The Nuclear Safety Law says the operating unit bears full responsibility for nuclear safety, while equipment and service suppliers bear corresponding responsibilities. | 中 | SR007 |
| CR009 | The same law says the State Council’s nuclear-safety supervision department is responsible for oversight and coordinates with nuclear-industry and energy departments. | 中 | SR007 |
| CR010 | Chinese law requires a nuclear-safety standards system that is mandatory and updated with social and technological progress. | 中 | SR007 |
| CR011 | NNSA’s English site shows a standing regulatory surface with laws and regulations, annual reports, monitoring data, and national safety reports. | 中 | SR008, SR009 |
| CR012 | Accessible public sources did not surface a Neo Fusion-specific plant safety plan, tritium-handling manual, or emergency-response package. | 中 | SR001, SR004, SR009 |
| CR013 | DOE’s 2026 fusion science and technology roadmap still lists breeding and handling of fusion fuels, materials, supply chains, workforce, and practical adoption as major challenge areas. | 中 | SR012 |
| CR014 | CRAFT’s 20 MW per square meter divertor result proves progress but also highlights the extreme heat-load engineering challenge that future devices must survive continuously. | 中 | SR006 |
| CR015 | CAS said the divertor coating could theoretically improve tritium breeding ratio, which implies tritium management is central enough to appear inside subsystem design itself. | 中 | SR006 |
| CR016 | WIPO Patentscope and CNIPA provide the patent infrastructure relevant to fusion IP, but accessible public sources in this review did not quantify Neo Fusion’s own patent estate or freedom-to-operate position. | 中 | SR010, SR011 |
| CR017 | AIP’s review of new U.S. fusion regulation shows that the industry still needs dedicated regulatory treatment distinct from fission, reinforcing that licensing is a real strategic variable rather than paperwork. | 中 | SR013 |
| CR018 | Neo Fusion chairman Yan Jianwen publicly argued in 2026 for synchronized legislation, standards, and talent construction, implying the enabling system is still incomplete. | 中 | SR030 |
| CR019 | Conference reporting around Hefei’s fusion ecosystem and new education initiatives suggests talent formation is an active workstream rather than a solved bottleneck. | 中 | SR031 |
| CR020 | Neo Fusion remains heavily dependent on ASIPP/HFIPS institutional capacity, making partner and key-person continuity a real dependency risk. | 中 | SR003, SR030 |
| CR021 | Fusion remains unusually capital intensive because specialized facilities, superconducting systems, materials programs, and long test cycles all precede revenue. | 中 | SR016, SR012 |
| CR022 | The Fusion Report said the top three private fusion companies still account for roughly half of all lifetime commercial-fusion funding. | 中 | SR015 |
| CR023 | The same report said AI funding in the first half of 2026 was roughly 100 times the amount raised for fusion and nuclear fission combined. | 中 | SR015 |
| CR024 | Crunchbase’s 2024 fusion-funding critique warned that milestone narratives can cool quickly and leave the sector exposed to funding slowdowns. | 中 | SR014 |
| CR025 | NucNet reported disclosed fusion-industry funding of about $14.24 billion in 2026, which is large in startup terms but still small against the sector’s total infrastructure ambition. | 中 | SR016 |
| CR026 | CFS’s $863 million Series B2 round shows even the best-capitalized private tokamak peers still need repeated large financings. | 中 | SR019 |
| CR027 | Helion’s 2025/2026 financing path likewise shows that commercially ambitious fusion peers still rely on very large private raises before delivering power. | 中 | SR020 |
| CR028 | TAE’s $150 million round suggests meaningful progress can still attract capital, but at a smaller scale than the biggest tokamak or direct-electricity narratives. | 中 | SR021 |
| CR029 | First Light Fusion’s £25 million first close in 2026 illustrates how sharply capital access can diverge by technical path and country even inside fusion. | 中 | SR022, SR023, SR024 |
| CR030 | Zap’s 2026 high-pressure and 37-million-degree announcements show peers continue to post technical milestones while remaining pre-commercial. | 中 | SR025, SR026 |
| CR031 | Our World in Data says utility-scale solar power costs fell about 88 percent from 2009 to 2024, intensifying the economic bar fusion must clear. | 中 | SR017 |
| CR032 | Lazard’s LCOE work reinforces that buyers will compare fusion against cheaper and more mature generation options rather than against a blank slate. | 中 | SR018 |
| CR033 | Because fusion customers can choose lower-risk alternatives, every schedule slip or cost overrun directly increases customer-delay risk for Neo Fusion. | 中 | SR017, SR018, SR029 |
| CR034 | Even Google’s 200 MW CFS offtake was tied to SPARC achieving Q>1, showing that sophisticated buyers still condition commitments on technical proof. | 中 | SR029, SR028 |
| CR035 | The Fusion Report excludes Chinese fusion-company private funding from its main tally because the line between public and private support is opaque. | 中 | SR015 |
| CR036 | BEST’s EUROfusion research plan shows international collaboration is valuable, but it also means cross-border cooperation is part of the project’s risk surface. | 中 | SR002 |
| CR037 | Neo Fusion’s dependence on state sponsors means a change in national priorities, local fiscal stress, or institutional leadership could slow the program abruptly. | 中 | SR030, SR031, SR001 |
| CR038 | A first utility or host-site customer would likely de-risk the business technically while simultaneously creating acute concentration risk because there are no other paying customers behind it. | 中 | SR029, SR004 |
| CR039 | Accessible public sources do not show a detailed schedule buffer between 2027 first plasma and the 2030 electricity-demonstration target. | 中 | SR003, SR004 |
| CR040 | No accessible public source published a Neo Fusion-specific cost model or plant-economics case strong enough to rebut the falling-cost alternative-power threat. | 中 | SR001, SR018, SR017 |
| CR041 | No accessible source disclosed how Neo Fusion would finance accident, liability, or other long-tail nuclear-risk coverage once operations mature. | 中 | SR001, SR007 |
| CR042 | A missed 2027 completion or first-plasma milestone would be a direct thesis-break trigger because the 2030 narrative depends on it. | 中 | SR003, SR002 |
| CR043 | Failure to surface a credible safety, tritium, and regulatory package as BEST matures would be another thesis-break trigger. | 中 | SR009, SR007, SR001 |
| CR044 | A funding bridge that does not clearly cover the path from 2027 assembly to later demonstration would also threaten the thesis. | 中 | SR015, SR014, SR016 |
| CR045 | Neo Fusion’s real mitigants today are ecosystem depth, state backing, EAST pedigree, and CRAFT subsystem progress—not public proof that the hard risks are solved. | 中 | SR005, SR006, SR001 |
| CR046 | Residual exposure remains high because nearly every major risk—physics, schedule, regulation, financing, and customer adoption—still depends on future milestones rather than current outputs. | 中 | SR002, SR015, SR007 |
| CV001 | Fusion Energy Base records an initial registration-capital event of about $723.37 million on 19 May 2023. | 中 | SV001 |
| CV002 | Fusion Energy Base records an additional registration-capital event of about $1.306 billion on 23 July 2024. | 中 | SV001 |
| CV003 | Chinese business media described the 2024 event as an increase in registered capital from RMB 5.0 billion to RMB 14.5 billion. | 中 | SV002, SV003 |
| CV004 | The resulting public headline capital base is roughly RMB 14.5 billion or about $2.03 billion equivalent. | 中 | SV001, SV002 |
| CV005 | That RMB 14.5 billion / ~$2.03 billion figure should not be treated as a clean priced venture valuation or as proven deployable cash. | 中 | SV001, SV009 |
| CV006 | Accessible public materials in this review do not disclose a market-clearing equity round, share price, or post-money valuation for Neo Fusion itself. | 中 | SV001, SV003 |
| CV007 | Public sources still do not prove how much of the headline capital base is fully paid in, ring-fenced, and spendable against BEST milestones. | 中 | SV001, SV003 |
| CV008 | NIO’s 2024 annual report confirms that one early Neo Fusion backer has public-capital-markets access, which supports the strategic-sponsor story without solving Neo Fusion’s own pricing opacity. | 中 | SV004, SV001 |
| CV009 | Public technical sources place BEST completion or first plasma around 2027. | 高 | SV001, SV005 |
| CV010 | Official Chinese reporting says BEST aims toward net fusion gain and electricity-generation milestones around 2030. | 高 | SV006, SV005 |
| CV011 | Meaningful commercial revenue is therefore more plausibly a post-2030 outcome than a present operating fact. | 中 | SV006, SV008 |
| CV012 | The Fusion Report explicitly excludes Chinese fusion funding from its private-capital league tables because that market is opaque and the line between private and public funding is hard to parse. | 中 | SV009 |
| CV013 | As of 30 June 2026, The Fusion Report said lifetime private funding across commercial fusion companies totaled about $11.52 billion, excluding Chinese companies. | 中 | SV009 |
| CV014 | The same report said the top three private fusion companies still account for roughly half of all lifetime commercial private funding. | 中 | SV009 |
| CV015 | NucNet reported broader disclosed fusion-industry funding of about $14.24 billion in 2026, reinforcing how much capital the sector already absorbs before commercialization. | 中 | SV011 |
| CV016 | The Fusion Report said AI funding in the first half of 2026 was roughly 100 times the capital raised for fusion and fission combined. | 中 | SV009 |
| CV017 | CFS raised $863 million in an August 2025 Series B2 and said it had raised close to $3 billion in capital to date. | 中 | SV014 |
| CV018 | The Fusion Report later quantified CFS at about $2.923 billion in lifetime private funding, the highest in the private fusion sector. | 中 | SV009, SV010 |
| CV019 | CFS disclosed a 200 MW future offtake agreement with Google for its first ARC plant. | 高 | SV015, SV017 |
| CV020 | CFS says ARC is designed around roughly 400 MW of net electricity and early-2030s grid delivery. | 高 | SV015, SV018 |
| CV021 | Chesterfield County said CFS expects ARC power to be sold to large industrial or commercial customers via private contracts rather than through ratepayer funding. | 中 | SV018 |
| CV022 | Google’s own announcement acknowledged that no private company has yet achieved Q>1 even as it made a second capital investment in CFS and signed the offtake agreement. | 中 | SV017 |
| CV023 | Helion’s January 2025 Series F announcement said the $425 million round valued the company at $5.425 billion post-money and brought total invested capital to over $1 billion. | 中 | SV019 |
| CV024 | The Fusion Report later described a June 2026 Helion Series G of $465 million at a $15.5 billion post-money valuation, with roughly $1.5 billion in lifetime private funding. | 中 | SV010, SV009 |
| CV025 | Helion disclosed the world’s first fusion power purchase agreement with Microsoft for at least 50 MW beginning initial operations in 2028. | 高 | SV020, SV021 |
| CV026 | Helion’s Orion page says the company has completed environmental review and begun site construction for its first fusion power plant in Washington. | 中 | SV021 |
| CV027 | Helion’s public publishing cadence in 2025-2026 spans licensing, diagnostics, and plant-construction topics, signaling a more mature commercialization narrative than Neo Fusion’s public record currently provides. | 中 | SV022, SV021 |
| CV028 | TAE announced a latest funding round of more than $150 million and said it had raised more than $1.3 billion in equity capital since inception. | 中 | SV023 |
| CV029 | TAE said Copernicus is aimed at a net-energy milestone before the end of the decade and that its first prototype power plant, Da Vinci, is targeted for the early 2030s. | 中 | SV023, SV024 |
| CV030 | TAE appointed a CFO with large-cap energy and public-company experience in 2026 and explicitly framed the role around capital markets, governance, and public-company readiness. | 中 | SV025 |
| CV031 | TAE’s public leadership roster includes dedicated finance, revenue, investor-relations, and intellectual-property functions, indicating a more institutionally packaged commercialization story. | 中 | SV026, SV025 |
| CV032 | TAE’s UKAEA partnership adds supply-chain and neutral-beam commercialization evidence that Neo Fusion has not yet matched in public. | 中 | SV027 |
| CV033 | First Light Fusion’s £25 million first close in 2026 shows how sharply funding scale can fall for fusion platforms that remain further from broad commercial packaging. | 中 | SV028, SV029 |
| CV034 | Zap’s Century platform milestone shows that alternative architectures are also building commercialization narratives around repetitive operation, heat extraction, and systems integration. | 中 | SV031 |
| CV035 | Zap’s 2026 strategy note argues that fusion still faces real industrialization and regulatory complexity even if timelines are improving. | 中 | SV030 |
| CV036 | Lazard’s 2026 LCOE+ summary says renewables remain the most cost-competitive form of new-build generation on an unsubsidized basis. | 中 | SV012 |
| CV037 | Our World in Data said the cost of electricity from solar fell about 90% over the last decade, demonstrating how fast alternative clean-power benchmarks can move. | 中 | SV013 |
| CV038 | DOE’s 2026 fusion roadmap still treats fuels, materials, supply chains, workforce, and practical deployment as active challenge areas, which justifies a large present-value discount for pre-commercial programs. | 中 | SV032 |
| CV039 | Neo Fusion’s ~$2.03 billion headline capital base is larger than TAE’s disclosed $1.3 billion lifetime private funding, but TAE offers more legible governance and commercialization proof. | 中 | SV001, SV023, SV025 |
| CV040 | Neo Fusion’s headline capital base exceeds Helion’s disclosed lifetime private funding in 2025-2026 but still sits far below Helion’s explicit multi-billion-dollar market valuations and cleaner commercialization package. | 中 | SV001, SV019, SV010 |
| CV041 | Neo Fusion’s headline capital base is below CFS’s roughly $2.9 billion lifetime private funding and also lacks CFS’s named offtake and first-plant ownership model. | 中 | SV001, SV009, SV018 |
| CV042 | The right way to think about Neo Fusion today is as strategically sponsored capital committed to a national fusion roadmap, not as a conventionally underwritten pre-revenue startup with a clean market price. | 中 | SV001, SV009, SV005 |
| CV043 | If BEST hits its 2027 and 2030 milestones, the strategic option value could increase materially because firm clean power would sit in a huge future market. | 中 | SV006, SV007, SV017 |
| CV044 | Today that option value still deserves a steep discount because Neo Fusion has no public priced round, no disclosed paid-in cash bridge, no host-site economics, and no customer contract proof. | 中 | SV001, SV005, SV018 |
| CV045 | A bearish outcome is justified if Neo Fusion remains only a strategic research program with slippage into or beyond 2027 and no proof of a financeable commercialization model. | 中 | SV005, SV009, SV012 |
| CV046 | The base case is a wide range centered near the strategic-capital anchor, assuming BEST progress continues but public proof still lags better-packaged peers. | 中 | SV001, SV010, SV032 |
| CV047 | A bullish case requires milestone delivery, a clearer paid-in-capital and budget bridge, and visible offtake or host-site logic that starts to resemble Helion or CFS commercialization proof. | 中 | SV006, SV021, SV018 |
| CV048 | Under present public evidence, the right posture is research-more with medium confidence and an unknown valuation stance rather than buy-grade conviction. | 中 | SV001, SV009, SV032 |
| CV049 | The recommendation should improve only if Neo Fusion discloses a paid-in-capital bridge, a detailed milestone budget, a real first-customer or host-site model, and stronger regulatory-packaging evidence. | 中 | SV018, SV021, SV025 |
| CV050 | The view should worsen quickly if 2027 slips materially, if capital availability proves weaker than the headline registration-capital figure suggests, or if cheaper alternative-power options keep moving ahead while Neo remains opaque. | 中 | SV005, SV001, SV013 |
| CV051 | The main reason confidence remains capped at medium is that almost every decisive valuation variable—cap table, paid-in cash, plant economics, customer terms, and budget cadence—remains private. | 中 | SV001, SV003, SV032 |