Proxima Fusion
欧洲融资最充足的核聚变冠军 — 收入前 €2.4B 估值上的高信念登月赌注
纸面上,Proxima Fusion 是欧洲最强的核聚变资产:继承 IPP / W7-X,拥有同行评审的仿星器方案、一线战略投资人和欧洲最大核聚变资金池;但公司仍无收入,净能量尚未验证,电网收入至少十多年后才可能出现,€2.4B 估值只能按风险投资式期权纪律紧密跟踪,不能按基本面承销。
封面要素
公司概况
Proxima Fusion GmbH 是一家总部位于慕尼黑的商业核聚变能源公司,也是 Max Planck Institute for Plasma Physics (IPP) 历史上第一家拆分创业公司。它正在开发基于准等动力 (QI)仿星器的电站——这种磁约束设计以准稳态运行,避开困扰托卡马克的破裂问题, 但代价是极其复杂的 3D 磁体几何。Proxima 用高温超导(HTS)磁体、计算优化和 AI/ML 辅助设计来攻克这件事。公司直接承接 Wendelstein 7-X 实验,并发表了 Stellaris;该方案被描述为首个经同行评审的商业仿星器电站概念。路线图从 Stellarator Model Coil(2027 年)走向慕尼黑附近的 Alpha 净能量演示装置 (2030 年代初),再到巴伐利亚前 Gundremmingen 核电站址的商业 Stellaris 电站, 目标是在 2030 年代末接入德国电网。2026 年 7 月,公司以 €2.4B ($2.7B) 投后估值 完成 €411M ($468M) 融资,由 XTX Ventures 和 East X Ventures 领投,Google 与 RWE 作为战略投资方参投,使其成为欧洲融资最充足的核聚变公司。
- 成立时间
- 2023-04-01
- 创始人
- Francesco Sciortino, Lucio Milanese, Jorrit Lion, Jonathan Schilling, Martin Kubie
- 创立地点
- Munich, Germany
- 总部
- Munich, Germany
- 产品
- Proxima 还没有销售商业产品。它的“产品”是一个核聚变电站项目:QI 仿星器概念 Stellaris、HTS 磁体开发路线(包括 Stellarator Model Coil 里程碑以及与 PSI 的 框架协议),以及规划中的 Alpha 净能量演示装置。最终商业供给是向电网和大型能源 买家交付基荷、清洁、准稳态核聚变电力,首座商业电站计划与 RWE 一起落在前 Gundremmingen 裂变核电站址。
- 客户
- 面向未来购电方和战略能源买家,而不是今天的付费客户:公用事业公司(RWE 是场址和合作伙伴)、 巴伐利亚州和德国公共部门,以及超大规模 / AI 数据中心运营商等大型清洁稳定电力买家 (Google 的战略投资反映了这一点)。目前没有具约束力的商业购电协议。
- 商业模式
- 收入前深科技。公司目前靠风险股权、战略企业投资和公共补助融资。长期设想是在 2030 年代 证明净能量和商业可行性之后,通过长期购电协议或电站交付协议销售核聚变电力和/或电站。
- 阶段
- Series B / venture-backed private
- 融资情况
- 约 €7M 种子前轮(2023 年 5 月)、约 €20M(2024 年 4 月)、€130M A 轮 (2025 年 6 月,由 Cherry Ventures 与 Balderton Capital 共同领投),以及 2026 年 7 月以 €2.4B ($2.7B) 投后估值完成的 €411M ($468M) 融资,由 XTX Ventures 与 East X Ventures 领投,Google 和 RWE(约 €25M)作为战略投资方参投——已锁定总资本超过 €650M ($740M)。
执行摘要
主要优势
- 科学血统几乎无可匹敌:首家 Max Planck IPP spin-out,直接站在 Wendelstein 7-X 上,并用同行评审的 Stellaris 电站概念降低物理叙事风险,相比多数聚变创业公司更扎实。
- 欧洲资金最厚的聚变公司,已锁定 €650M+,2026 年 7 月一轮融资 €411M;Google、RWE、XTX / East X 等一线战略方同时带来资本、包销意向和数据中心需求信号。
- 差异化 QI-stellarator 路线承诺准稳态运行,并避开 tokamak 破裂问题;HTS 磁体和 AI / ML 驱动的 3D 线圈优化构成真实技术护城河。
- 工业化路径具体:RWE 合作、Gundremmingen 商业场址、超过 50 家伙伴的 Alpha Alliance,以及分阶段路线图(Model Coil 2027 → Alpha 2030 年代初 → Stellaris 2030 年代末)。
- 创始团队强,治理层也在补工业化能力:CFO Sergei Galperin 和由欧洲资深产业人士组成的 Industrial Development Board,显示公司正从实验室转向工业执行。
主要风险
- 基础科学风险仍是最大变量:还没有仿星器(或任何装置)演示过商业净能量增益,核心价值主张尚未验证,存在彻底失败的可能。
- 资本强度和稀释都极端:仅 Alpha 项目就是数十亿欧元级工程,在产生任何收入前还要反复做巨额融资并依赖持续公共补贴。
- 收入周期很长(目标 2030 年代末并网),投资逻辑会暴露在时间表滑坡、技术替代和融资市场周期之下。
- €2.4B 的无收入估值在基本面上安全边际很薄,高度依赖期权价值和聚变赛道热度延续。
- 全球对手资金充足:Commonwealth Fusion Systems、Helion、TAE,以及 Type One、Thea、Gauss、Renaissance 等仿星器同行,可能率先抢到进度、关键人才或包销协议。
- 公司高度依赖创始团队、Max Planck IPP 资源,以及稀缺的仿星器工程人才。
未决问题
- 财务未披露:烧钱速度、现金跑道、资金用途细节,以及 €650M+ 中股权融资与公共补助的拆分都未公开。
- 2025 和 2026 轮的股权结构表、投资人持股比例、清算优先权、控制权 / 治理条款未披露。
- Stellaris 概念论文之外的技术里程碑证据(Model Coil 测试结果、HTS 磁体表现)尚未公开演示。
- 目前没有具约束力的商业包销、PPA 或电站交付合同;RWE / Gundremmingen 与数据中心需求仍是合作阶段信号。
- QI-stellarator 净能量路径和 Alpha 时间表缺少独立、经审计的验证;时间线仍由公司给出。
- 用作可比公司的精确同业估值标记(CFS、Helion 等)部分来自媒体报道估计,并非公司披露。
目录
01公司概览
1.1 身份、阶段与运营模式
Proxima Fusion 应被视为一家总部在慕尼黑、收入前的深科技基础设施公司,而不是传统能源卖方。法律实体为 Proxima Fusion GmbH,注册在慕尼黑地方法院,编号 HRB 283423;公司材料也将 Zurich 和 Oxford 列为运营地点。公开来源把公司创立时间指向 2023 年 4 月,并独立称其为 Max Planck Institute for Plasma Physics 的首家拆分创业公司。它的一句话模式在年轻核聚变创业公司里异常具体:建造 QI-HTS 仿星器,先做 2027 年的 Stellarator Model Coil,再在 2030 年代初于慕尼黑附近做 Alpha,最终在 2030 年代末做 Stellaris 电站。按通常意义,这还不是收入模式。已检索来源没有披露产品收入、ARR、付费客户数、 经审计财务或售电合同,因此运营阶段仍是收入前,面对的是公共伙伴和未来购电方,而不是商业客户。[CO001, CO002, CO003, CO004, CO005, CO006]
| 指标 | 数值 / 状态 | 日期 / 版本 | 置信度 | 缺口或尽调路径 |
|---|---|---|---|---|
| 法律实体 | Proxima Fusion GmbH;HRB 283423 | 当前登记册 / 网站法律声明 | 高 | 直接向管理层确认最新股东名册和章程。 |
| 总部 | 德国慕尼黑;运营地点还包括苏黎世和牛津 | 2026 年公司材料 | 高 | 按地点确认员工数,以及实验室 / 制造场地布局。 |
| 成立 | 2023 年 4 月;IPP 来源称于 2023 年初启动 | 2023–2026 年来源 | 高 | 除非登记册摘录显示确切设立日期,否则以 2023 年 4 月为规范月份。 |
| 阶段 | 商业化前深科技 / 聚变示范堆建设 | 2026 年评估 | 中 | 索取经审计收入、已确认拨款、现金和现金续航期。 |
| 技术 | 基于 W7-X 的 QI-HTS 仿星器电站 | 当前公司材料 | 高 | 技术章节应尽调 TRL 和 Alpha 准备度。 |
| 员工 | 2026 年 7 月轮后约 200 人 | 2026-07 | 中 | 核实薪资名册人数、承包商和招聘计划。 |
| 最新轮次 | €411M ($468M) | 2026-07-07 | 高 | 确认一级 / 二级交易组合和交割条件。 |
| 最新估值 | €2.4B ($2.7B) 投后 | 2026-07-07 | 高 | 确认完全稀释后股数和优先权堆栈。 |
| 已锁定总资金 | €650M+ ($740M+),包括 €95M 公共拨款 | 2026-07-07 | 中 | 核对股权融资、拨款和非稀释承诺。 |
| 收入 / ARR | null — 未公开披露 | 当前缺口 | 低 | 索取 FY2024-FY2026 管理账和拨款会计政策。 |
| 当前客户 | null — 公开证据中只有合作伙伴 / 未来购电方 | 当前缺口 | 低 | 区分战略合作伙伴、付费试点、LOI 和有约束力的购电合同。 |
| 商业并网目标 | 2030 年代末,通过 Stellaris/Gundremmingen 路线图 | 2026 年路线图 | 中 | 验证场址、许可、并网和项目融资关键路径。 |
快照采用 runDate 当日的公开披露。null 值表示未找到公开证据,不代表经济价值为零。
[CO001, CO002, CO003, CO005, CO008, CO009]概览逻辑把 IPP/W7-X 传承、QI-HTS 技术、资本、合作伙伴和未解决的商业化风险串在一起。
流程图是分析框架;不表示所有合同连接都具约束力。
[CO004, CO005, CO006, CO013, CO026, CO027]公开 KPI 显示融资能力很强,但商业牵引披露很少;收入、ARR 和客户数仍为 null。
货币换算沿用公司提供的美元指引;null 表示公开披露不支持该值。
[CO008, CO009, CO010, CO013, CO043, CO047]1.2 创始人、领导层与关键人物依赖
领导层既是优势,也是依赖。公开创始人名单包括 Francesco Sciortino、Lucio Milanese、Jorrit Lion、Jonathan Schilling 和 Martin Kubie,这个创始团队直接连着 Max Planck IPP、MIT 和 Google X 式技术背景。Sciortino 在融资、Stellaris 和 IPP 合作材料中仍是核心公开发言人和 CEO; Milanese 也在公司法律声明中担任董事总经理;Lion 是 Stellaris 上可见的首席科学家声音。因此,2026 年 6 月 Sergei Galperin 加入担任 CFO 很重要,因为它在资本密集的 Alpha 和 Stellaris 阶段前补强了财务班底。 不过,公开治理面仍不完整。Industrial Development Board 加入了重量级产业顾问,但已检索材料没有披露完整法定董事会、 观察员权利、投票门槛、创始人归属或继任规划。后续尽调应测试:Proxima 与 IPP 绑定的创始人,究竟仍是执行加速器, 还是集中度风险。[CO019, CO020, CO021, CO022, CO023, CO024]
| 人物 | 角色 / 相关性 | 背景或公开证据 | 创始人-市场匹配 / 依赖 | 尽调请求 |
|---|---|---|---|---|
| Francesco Sciortino | 联合创始人兼 CEO | 在 IPP、Stellaris、CFO 和融资材料中均被引用 | 核心公开运营者;关键人依赖高 | 审查接班计划、归属期和从技术到产业化的授权分工。 |
| Lucio Milanese | 联合创始人;公开来源中的董事总经理 / 对外事务角色 | 公开证据列名为创始人和董事总经理 | 对机构 / 利益相关方接口很重要 | 确认当前头衔、职责和公共部门关系归属。 |
| Jorrit Lion | 联合创始人兼首席科学家 | 在 Stellaris 论文发布公告中被引用 | QI-HTS 概念的核心科学权威 | 评估物理专业知识集中度和留任计划。 |
| Jonathan Schilling | 联合创始人 / 公开创始人名单中的实验室负责人 | 技术创始团队成员 | 运营角色在已抓取来源中不太可见 | 确认当前职责、实验室里程碑和汇报线。 |
| Martin Kubie | 联合创始人 / 公开创始人名单中的工程负责人 | 技术创始团队成员 | 工程执行可能对 SMC 和 Alpha 至关重要 | 确认当前职责、硬件交付责任和招聘缺口。 |
| Sergei Galperin | 首席财务官 | 2026 年 6 月加入,此前有 J.P. Morgan、Alan 和 Ribbit 经历 | 在资本密集型 Alpha 阶段前补强财务班底 | 审查募资计划、内控、拨款会计和项目融资策略。 |
| 工业发展委员会 | 顾问委员会:Luc Rémont、Michael Bolle、Ann Mettler、Erich Clementi | 2026 年 5 月宣布,支持工业化扩张 | 增加工业网络,但不能替代法定治理 | 索取正式授权范围、会议频率、薪酬和董事会观察员链接。 |
公开领导层覆盖不完整:创始人和部分顾问可见,但完整董事会、观察员、委员会和接班细节未披露。
[CO019, CO020, CO021, CO022, CO023, CO024]1.3 资本形成与利益相关方地图
融资形成是最清晰的公开动能信号。Proxima 从 2023 年约 €7M 种子前轮,走到 2024 年 €20M 种子轮、 2025 年 6 月 €130M A 轮,再到 2026 年 7 月 €411M 融资,投后估值 €2.4B。最新融资由 XTX Ventures 和 East X Ventures 领投,Google 和 RWE 作为战略投资方参投;Proxima 称包括公共补助在内, 已锁定资本超过 €650M。这足以确立 Proxima 作为欧洲融资最充足核聚变公司的位置,但不足以支撑经济性判断。 来源没有披露新股/老股比例、清算优先栈、债务、董事会权利或内部人持股。利益相关方地图也关键,因为资金和执行绑在一起: RWE 贡献场址和监管经验,Bavaria 与公共补助帮助降低 Alpha 风险,IPP 提供科学传承,XTX/Google 则把叙事接到 AI 赋能工程和长期数据中心电力需求。[CO010, CO011, CO012, CO013, CO014, CO015]
| 利益相关方 | 角色 | 控制权 / 经济重要性 | 公开信号 | 尽调请求 |
|---|---|---|---|---|
| XTX Ventures | 2026 年 7 月共同领投方 | 资本加 AI/ML 技术验证叙事 | 列名为领投;XTX Ventures 宣传 AI/ML 技术支持 | 确认持股、董事会 / 观察员权利和技术支持承诺。 |
| East X Ventures | 2026 年 7 月共同领投方 | 共同定价最新估值,且可能获得治理权包 | 在 Proxima 和 CNBC 报道中列名为共同领投 | 确认基金身份、支票规模和保留事项。 |
| 战略投资者 | 潜在长期需求信号,指向稳定清洁电力和 AI 数据中心 | 列名战略投资者;CNBC 突出 Google 支持 | 判断是否存在任何购电、云、算力或技术协议。 | |
| RWE | 战略投资者和场址 / 购电合作伙伴 | €25M 投资者,具备 Gundremmingen 场址和审批经验 | RWE 新闻稿描述投资和场址合作 | 审查场址 MOU、排他性、购电条款、许可义务和退出权。 |
| Max Planck IPP | 科学传承和合作伙伴 | 核心技术依赖和可信度来源 | IPP 首个分拆公司和合作协议 | 审查 IP 许可、合作条款、人员依赖和冲突规则。 |
| 巴伐利亚自由州 / 公共拨款方 | 公共资金和 Alpha 生态赞助方 | 帮助催化私人资本,并可能支持枢纽 / 场址 | Proxima 提到 €95M 公共拨款和巴伐利亚路线图贡献 | 验证拨款条件、里程碑、追回条款和采购义务。 |
| Series A 投资者 | Cherry、Balderton、UVC、Plural、DTCF、Lightspeed、redalpine 等 | 更早的优先权堆栈和跟投能力 | TechCrunch 和 Sifted 列出参与方 | 梳理持股、优先权、按比例跟投权和老股转让。 |
| 未来电力客户 / 购电方 | 尚未证明的付费客户 | 收入情形取决于未来 PPA 或战略需求 | 当前证据列名合作伙伴,而非付费客户 | 索取 LOI、有约束力的购电条款、定价假设和交易对手信用。 |
利益相关方地图有意保持不完整,因为公开来源识别了重要投资者和合作伙伴,但没有披露经济条款、治理权或有约束力的商业合同。
[CO011, CO012, CO014, CO016, CO026, CO030]1.4 里程碑、路线图与技术传承
里程碑序列最好理解为一场从机构科学走向工业执行的竞赛。IPP 的 Wendelstein 7-X 给 Proxima 一个可信的仿星器基础, 2025 年 Stellaris 论文则把这份传承转成公司专属电站概念。随后路线图转向硬件:2027 年完成 Stellarator Model Coil,完成 Alpha 设计,2030 年代初在慕尼黑附近运行 Alpha,并在 2030 年代末争取在 Gundremmingen 建成首座商业磁约束核聚变电站。RWE 2026 年 7 月公告让场址策略不只是幻灯片:公告称 Proxima 已选择 Gundremmingen,并将与巴伐利亚环境部启动审批。卡点在于所有重大价值拐点仍在前方。Stellaris 是经同行评审的概念, Alpha 是计划中的演示装置,商业并网电力仍是 2030 年代末的目标。因此,这条时间线足够支撑继续尽调,但还不足以假设执行成功。[CO006, CO007, CO027, CO028, CO029, CO030]
| 日期 | 事件 | 类型 | 金额 / 估值 / 状态 | 参与方 | 含义 |
|---|---|---|---|---|---|
| 2023-04 | Proxima 在慕尼黑成立 | 创立 | 公司设立 | 来自 IPP、MIT、Google X 背景的创始团队 | 确立规范成立日期和创始人-市场匹配。 |
| 2023-05-30 | IPP 合作协议和首个分拆公司公告 | 合作 | 合作协议 | Proxima 与 Max Planck IPP | 锚定 W7-X 传承和科学依赖。 |
| 2023-05 | Pre-seed 融资被报道 | 融资 | ~€7M | Max Planck Innovation、EU-Startups 报道的投资者 | 为分拆后的首次公司建设提供资金。 |
| 2024-04 | NucNet 报道 Seed 融资 | 融资 | €20M | 包括 redalpine 生态在内的 Seed 投资者 | 支持 QI 仿星器开发。 |
| 2024-06 | 面向 HTS 磁体的 PSI 框架协议 | 合作 | 框架协议 | Proxima 与 Paul Scherrer Institute | 为仿星器路线图增加磁体技术伙伴。 |
| 2025-02-26 | Stellaris 概念论文发布公告 | 产品 | 同行评审概念 | Proxima、IPP、KIT 和学术伙伴 | 形成首个重要公司专属技术里程碑。 |
| 2025-06 | Series A 交割 | 融资 | €130M;融资总额至 ~€185M | Cherry、Balderton、UVC、Plural、DTCF、Lightspeed 等 | 当时欧洲聚变初创公司最大轮次。 |
| 2026-05-13 | 成立工业发展委员会 | 治理 | 顾问委员会 | Rémont、Bolle、Mettler、Clementi 等顾问 | 标志从科学转向工业化扩张。 |
| 2026-06-01 | Sergei Galperin 加入担任 CFO | 治理 | 领导层补强 | Proxima | 在 Alpha 资本需求前补强财务领导力。 |
| 2026-07-07 | 宣布 €411M 融资 | 融资 | €411M,投后估值 €2.4B | XTX、East X、Google、RWE、老股东 | 使 Proxima 成为欧洲资金最充足的聚变公司。 |
| 2026-07-07 | RWE 披露 €25M 投资和 Gundremmingen 计划 | 合作 | €25M;场址合作 | RWE、Proxima、巴伐利亚、IPP | 把商业电站计划推进到一个具名场址。 |
| 2031 目标 | Alpha 净能量示范装置目标 | 产品 | 计划中,尚未建成 | Proxima、巴伐利亚、IPP、RWE 生态 | 重大技术证明点仍在前方。 |
| 2030 年代末目标 | Stellaris 商业并网雄心 | 产品 | 计划中,尚未建成 | Proxima 和场址 / 公用事业伙伴 | 商业价值兑现仍偏长期。 |
里程碑时间线混合了已完成事件和明确未来目标;未来目标是路线图主张,不是已达成里程碑。
[CO003, CO004, CO007, CO010, CO014, CO015]Proxima 的公司历程从 2023 年 IPP 拆分孵化到 2026 年 €411M 融资;关键的 Alpha 和 Stellaris 证明点仍在未来。
未来路线图日期是公司目标,不是已达成里程碑。
[CO003, CO004, CO010, CO014, CO015, CO017]1.5 反向背景与尽调路径
反向观点不是 Proxima 不严肃,而是投资逻辑依赖证明一些尚无私人核聚变公司在商业上证明过的事情。CNBC 指出核聚变尚未商业部署, 并引用 Google 的保留意见:商业化极其艰难,成功没有保证。Sifted 对 A 轮的报道在净能量风险上更直接,提醒没有任何核聚变装置 已经做到产出能量超过消耗。即便纳入国家实验室点火里程碑,Proxima 自己的计划仍需要数个未融资或部分融资步骤:SMC、Alpha、许可、 场址改造、供应链放大,以及后续电站项目融资。公司有可观公共和私人支持,但广泛股东基础和战略伙伴无法消除稀释、治理或执行风险。 因此,本章对收入、ARR、客户数和股权结构经济性明确留空,尽调路径聚焦管理层财务、伙伴合同、董事会材料和 Alpha 技术就绪度评审。[CO009, CO039, CO040, CO041, CO042, CO045]
1.6 图表
02市场分析
2.1 市场边界:清洁稳定电力,而不是泛能源转型支出
Proxima 的市场规模应对标清洁、稳定、可调度电力和高温能源选项,而不是所有可再生能源、所有核能或全部气候科技资本。 纳入范围包括未来核聚变发电、电站开发、电网互联、HTS 磁体和热管理系统等使能组件,以及公用事业或大型企业负荷买家的战略购电。 相邻但不能完全替代的市场包括小型模块化裂变反应堆、增强型地热、长时储能、配碳捕集的天然气、氢能,以及传统可再生能源 PPA。 这个边界重要,因为 Proxima 仍无收入,也因为今天没有商业核聚变电站。当前市场信号因此不是销售转化,而是政府、公用事业、 超大规模云厂商和供应商愿不愿意为可能在 2030 年代变成售电的里程碑出钱。Proxima 获 RWE 和 Google 参与支撑这一边界: 二者都是电力需求相关角色,但都不能证明核聚变已进入可银行融资的发电市场。[CM001, CM002, CM003, CM004, CM005, CM006]
| 细分 / 类别 | 纳入支出 | 排除支出 | 买方 / 付款方 | 与 Proxima 的相关性 |
|---|---|---|---|---|
| 商业聚变电力 | 未来聚变电站售电、电网服务、购电合同 | 未绑定聚变的普通可再生 PPA 和常规核电输出 | 公用事业、超大规模云厂商、公共电力买方 | 如果 Stellaris 达到并网运营,这是核心长期收入池 |
| 聚变电站开发和部件 | HTS 磁体、仿星器工程、热管理、燃料循环系统、EPC 准备 | 没有商业化路径的学术研究 | 聚变开发商、政府、战略供应商 | 电力收入出现前的近期支出池 |
| 清洁稳定电力邻近市场 | 裂变、增强型地热、长时储能、燃气 CCS、带稳定化的可再生能源 | 未配套稳定化的纯间歇性可再生能源 | 超大规模云厂商、工业企业、公用事业 | 可在聚变到来前满足需求的替代集合 |
| 欧洲能源安全需求 | 本土低碳基荷、工业竞争力、主权技术政策 | 进口化石燃料供应和非稳定抵消 | 欧盟 / 德国政府、公用事业、大型工业 | 解释德国 / 巴伐利亚政策支持和 RWE 兴趣 |
| 工业热 / 过程能源 | 未来高温热或电力到工业应用 | 低品位热市场和无关效率服务 | 重工业和基础设施运营商 | 可能的细分,但 Proxima 公开证据支持较弱 |
边界把未来聚变收入,与更广泛的清洁能源支出、以及争夺同一清洁稳定电力预算的替代技术区分开。
[CM001, CM002, CM003, CM004, CM005, CM006]Proxima 的机会从广义清洁稳定电力需求,收窄到欧洲 / 德国政策支持,最后落到尚未数值披露的首座电站 SOM。
[CM008, CM015, CM025, CM026, CM036, CM037]2.2 规模测算视角:预测数很大,但近期 SAM/SOM 证据窄
最宽口径的已发布核聚变市场规模很醒目,但应视为情景,而不是当前 TAM。Precedence Research 预测核聚变市场 2030 年为 $471.99B,2040 年达 $843.46B;FIA 2025 年报告称行业过去 12 个月融资约 $2.6B,累计投资接近 $9.8B。 这些数字显示的是上升的期权价值,不是商业电力收入。更适合尽调的视角,应从可能为清洁稳定电力付费的需求池出发: McKinsey 预计 2024 至 2030 年美国数据中心用电需求将增加约 400 TWh,BloombergNEF 追踪到截至 2025 年末全球在建数据中心 IT 容量为 23.1 GW。对 Proxima 来说,SAM 还要更窄:欧洲和德国,能源安全、产业竞争力和核聚变政策在这里同向。 SOM 目前无法量化,因为 Proxima 披露了里程碑,但没有披露电站容量、合同价格或购电量。[CM008, CM009, CM010, CM011, CM012, CM013]
| 视角 | 发布方 / 来源 | 年份或期间 | 地域 | 价值 / 信号 | 置信度 | 局限 |
|---|---|---|---|---|---|---|
| 聚变 TAM 场景 | Precedence Research | 2030 | 全球 | $471.99B 核聚变市场 | 低 | 尽管今天还没有商用聚变,预测仍假设商业市场会形成 |
| 核聚变 TAM 情景 | Precedence Research | 2040 | 全球 | $843.46B 核聚变市场 | 低 | 高度推演的市场模型,不是已签约收入 |
| 核聚变融资代理指标 | FIA 2025 报告覆盖范围 | Jul 2024-Jul 2025 | 全球 | $2.6B+ 新增聚变投资;累计约 $9.8B | 中 | 融到的资本不等于终端市场需求 |
| 核聚变供应链代理指标 | FIA 供应链报告 | 2025-2026 | 全球 | 2025 年供应链支出披露为 $538M;2026 年预计 $681M | 中 | 调查覆盖约一半私营公司 |
| AI 数据中心电力需求 | McKinsey | 2024-2030 | 美国 | 约 400 TWh 增量数据中心用电需求 | 中 | 仅限美国;并非所有需求都会优先选择聚变 |
| 数据中心建设信号 | BloombergNEF | 2025-2026 | 全球 | 23.1 GW IT 容量在建;2026 年资本开支接近 $750B | 中 | IT 容量和资本开支是需求代理指标,不是购电承诺 |
| 欧洲数据中心 SAM 代理指标 | BloombergNEF | 2030 | 英国 / 德国 / 爱尔兰 / 挪威 / 荷兰 | 5.4 GW 中性情景、>7 GW 激进情景实时 IT 用电需求 | 中 | 2021 年研究较早,但对地域和灵活性逻辑仍有用 |
| Proxima SOM 证据 | Proxima / CNBC | 2030 年代初至末期 | 德国 | Alpha 示范装置,随后瞄准首座商用仿星器电站 | 低 | 未披露 MW 容量、PPA 价格或签约电量 |
测算口径刻意保留彼此不兼容的单位,而不是把它们平均;只有 Precedence 两行是市场规模预测, 其他行只是需求或准备度代理指标。
[CM008, CM009, CM010, CM011, CM012, CM013]公开资料里唯一口径一致的数字化市场规模区间,是带有推测性的 2030-2040 年核聚变预测;应把它作为情景展示,而不是当前收入。
所有数值都来自同一 Precedence Research 预测期,口径为十亿美元;市场尚未商业化,可信度仍低。
[CM008, CM009, CM036, CM037]2.3 需求细分与买方路径:公用事业先行,超大规模云厂商作为战略加速器
可能的买方路径由电力系统角色牵头,而不是单个工业场址。电气化和天气依赖型可再生能源提高系统平衡要求后,公用事业和电网运营商需要可靠、 低碳容量;RWE 的投资和 Gundremmingen 合作,是 Proxima 最清晰的公用事业信号。超大规模云厂商是第二条需求向量。 Google 自身可持续发展材料称,其 AI 基础设施建设快过电网脱碳,2025 年签署了超过 12 GW 清洁能源协议,并在扩大核能和增强型地热的同时, 对核聚变做长期押注。数据中心买家看重 24/7 无碳能源、接电速度和可信的长周期供应,但在核聚变到来前,也可以选择裂变、地热、 储能支撑的可再生能源或电网合同。工业热是可信的第三个细分,因为核聚变电站产生热能,供应链已聚焦热管理约束;但 Proxima 的公开证据对电网电力的支撑远强于工业过程热。[CM016, CM017, CM018, CM019, CM020, CM021]
| 需求细分 | 经济买方 | 日常用户 / 利益相关方 | 预算所有者 | 采用触发点 | 可能路径 |
|---|---|---|---|---|---|
| 电网基荷 / 公用事业容量 | 公用事业高管团队和发电战略负责人 | 电网运营、厂站工程、交易团队 | 公用事业资产负债表、项目融资、公共支持 | 需要稳定低碳容量和能源安全 | 示范验证 → 场址 / 承购合作 → 项目融资 → 并网电站 |
| AI 与超大规模数据中心 | 能源采购和可持续发展负责人 | 数据中心基础设施和并网团队 | 企业清洁能源采购和资本开支 | 全天候无碳能源需求和通电时间压力 | 战略投资 → 长期承购选择权 → 组合采购 |
| 工业热能和电力 | 工业能源 / 运营负责人 | 工艺工程和设施团队 | 能源采购和脱碳预算 | 高温热源和可靠性要求 | 只有电站性能得到验证后,才会试点热集成 |
| 政府 / 主权技术 | 科研、经济和能源部委 | 公共实验室、资助机构、区域发展机构 | 公共补助、IPCEI/Euratom、国家资金 | 能源安全、产业竞争力、战略自主 | 研发补助 → 枢纽 → 示范装置 → 监管框架 |
| 聚变供应链 | 聚变开发商和一级供应商 | 磁体、材料、热管理、真空、燃料循环团队 | 开发商资本开支和供应商扩张预算 | 商用机器到来前就需要先备好产能 | 供应商认证 → 长期可见度 → 规模化生产 |
买方地图反映公开证据;真实采购权限、电价结构和 PPA 经济性仍需在非公开尽调中核实。
[CM016, CM017, CM018, CM019, CM020, CM021]公用事业和超大规模云厂商是证据最充分的需求板块;工业热能需求可能成立,但 Proxima 公开材料验证较少。
[CM016, CM017, CM018, CM019, CM020, CM021]市场转化先要过技术验证和项目可融资性两关,广泛的清洁电力需求才会变成 Proxima 收入。
[CM006, CM014, CM015, CM025, CM026, CM027]2.4 时点、地理、政策顺风与约束
市场时点有利,但容错很低。德国 Fusion Action Plan 承诺到 2029 年投入超过 €2B,并把德国第一座核聚变电站列为国家目标; Bavaria 另行宣布最高 €400M,投向与 Proxima、RWE 和 IPP 相关的项目;EU/Euratom 政策也在走向 2026-2027 年核聚变商业化和公私合作框架。 这给 Proxima 在 Munich/Garching 和 Bavaria 周边一个异常强的本地政策楔子,而公司自身也把欧洲描述成在与美国和中国赛跑。 反向观点同样重要:Clean Energy Wire 报道核聚变仍处实验阶段,商业使用至少还要数十年;CNBC 引用 Google 称商业化极其艰难,成功没有保证。 正确的市场结论因此应分阶段。2030 年代初的 Alpha 可以扩大 SAM 可信度;2030 年代末的并网电站可以创造首个 SOM 证据; 但估值不应像商业收入已经去风险一样,把宽口径 2040 年市场预测资本化。[CM025, CM026, CM027, CM028, CM029, CM030]
| 驱动因素 / 约束 | 方向 | 时间 | 含义 | 尽调问题 |
|---|---|---|---|---|
| AI 数据中心用电负荷快速增长 | 驱动因素 | 现在至 2030 年 | 为可靠清洁电力创造支付意愿高的买方 | 询问 Google 的兴趣是否包括技术协作、承购权,还是只有财务敞口 |
| 欧洲能源安全和产业政策 | 驱动因素 | 2026-2030 政策周期 | 支持德国首座电站和供应商生态在本地落地 | 梳理补助、里程碑、追回条款、IP 限制和许可要求 |
| RWE / Gundremmingen 场址路径 | 驱动因素 | Alpha 至 2030 年代末 | 公用事业伙伴和原核电场址可能缩短市场形成路径 | 索取 MoU 条款、并网责任、场址约束和承购框架 |
| 聚变供应链支出上升 | 驱动因素 | 2025-2026 | 说明供应商已经在收入到来前开始建设产能 | 核实 Proxima 在 HTS、真空、热管理和燃料循环系统上的供应商承诺 |
| 目前没有商用聚变电站 | 约束 | 当前 | 市场预测体现的是期权价值,不是已验证销售 | 在净能量和电站经济性得到证明前,不要把 TAM 获取写入投资假设 |
| 竞争性稳定清洁技术 | 约束 | 当前至 2030 年代 | SMR、地热、储能和配套稳定化的可再生能源可能先拿下需求 | 按买方细分比较平准化成本、时间、许可和可融资性 |
| 工程和供应链瓶颈 | 约束 | 当前 | 电力系统、热管理、真空室、泵、燃料循环和材料都可能拖慢电站 | 对照 Alpha 和 Stellaris 里程碑跟踪关键路径部件 |
| 监管和 IP 商业化复杂度 | 约束 | 2026 年起 | 资金支持可能附带合作、开发利用和出口管制义务 | 审查公共资助条款,以及聚变是否不受传统核能法律约束 |
驱动因素真实存在,但会分阶段释放;约束会直接决定广义稳定清洁电力需求能否转成 Proxima 的收入。
[CM010, CM011, CM012, CM013, CM017, CM018]2.5 图表
03竞争对手
3.1 竞争格局:获融资的核聚变对手按约束路线分化
Proxima 竞争的不是单一产品品类;它参与的是一场资本密集竞赛,目标是成为首批可信的私人核聚变电站供应商之一。赛场首先按物理架构分开。 CFS 和 Tokamak Energy 走高场托卡马克;Type One、Thea、Gauss、Renaissance 和 Proxima 是最相关的私人仿星器队列; Helion、TAE、General Fusion 和 Zap 使用场反位形、磁化靶或 Z 箍缩变体;First Light、Marvel、Focused Energy、 Xcimer 和 Pacific Fusion 更接近惯性或脉冲功率路径。这一点重要,因为投资人、公用事业、超大规模云厂商和政府买的不是标准化反应堆, 而是时间线、技术降险、产业供应链接入和可信的电网项目。Proxima 的 QI 仿星器叙事在仿星器赛道内有差异化,但融资表显示, 全球资本最充足的对手仍主要是美国托卡马克、FRC 和惯性路线玩家。[CP006, CP007, CP009, CP011, CP012, CP013]
| 公司 | 路线 / 类别 | 披露融资或估值信号 | 目标细分 / 里程碑 | 差异化 | 局限 |
|---|---|---|---|---|---|
| Proxima Fusion | QI 仿星器 / 直接同业 | 含补助在内合计 €650M+;2026 年 7 月投后估值 €2.4B | 2030 年代初,慕尼黑附近的 Alpha 净能量仿星器 | QI-HTS 设计、W7-X 传承、AI 线圈优化、Stellaris 电站概念 | 比美国头部玩家更年轻、资金更少;模型线圈和制造验证仍待完成 |
| Commonwealth Fusion Systems | HTS 托卡马克 / 既有领先者 | 按 TechCrunch 列表,已融资接近 $3B | SPARC 净能量装置,2030 年代初 ARC 并网电站 | 资金最充足的私营聚变玩家,路线图是紧凑型 HTS 托卡马克 | 托卡马克破裂、材料、氚和建设风险仍在 |
| TAE Technologies | 场反位形 / 长期在场的既有玩家 | 按 TechCrunch/PitchBook,计划合并前已融资约 $1.79B | 将先进束流 FRC 平台商业化 | 运营历史深,另有替代燃料循环叙事 | 时间线漫长,复杂公司交易也削弱可比性 |
| Helion Energy | FRC / 直接发电 / 超大规模客户支持的对手 | 2026 年一轮 $465M;估值 $15.5B;累计融资 $1.5B | Orion 电站和 Microsoft 客户时间表 | 直接电能转换,以及 Sam Altman/OpenAI 邻近性 | 激进的 2028 目标抬高交付风险 |
| Tokamak Energy | 球形托卡马克 / HTS 磁体同业 | 按 TechCrunch/PitchBook,累计融资 $336M | ST40 和 HTS 磁体商业化 | 紧凑型托卡马克叠加磁体技术收入选项 | 资金基础小于 CFS,与仿星器的直接相关性也更弱 |
| General Fusion | 磁化靶聚变 | 按 TechCrunch,已融资超过 $600M | LM26 示范装置和首创电站路线图 | 液态金属压缩架构和长期运营历史 | 混合机械系统带来独特工程复杂度 |
| Zap Energy | 剪切流 Z 箍缩 / 相邻路线 | 按 TechCrunch/PitchBook,已融资 $327M | 聚变试点电站预概念里程碑 | 用紧凑 Z 箍缩路线绕开 HTS 磁体和激光 | 仍必须证明电站级等离子体性能稳定且可重复 |
| First Light Fusion | 惯性 / 弹丸与 FLARE 架构 | 按 TechCrunch/PitchBook,已融资 $108M | 可扩展惯性架构,以及相邻极端条件能力 | 峰值功率较低的惯性概念 | 商业路径已重新聚焦,且无法与磁约束路线直接比较 |
| Type One Energy | 仿星器 / 直接同业 | 按 TechCrunch,含 Series B 前股权在内已融资 $269M | 2030 年代中期,与 TVA 相关的 350 MW 电站概念 | 美国仿星器路线,采用公用事业主导建设 / 持有 / 运营模式 | 2026 年 7 月后资本少于 Proxima,且仍处于商业化前 |
| Thea Energy | 仿星器 / 直接同业 | 按 TechCrunch,2026 年完成 $100M Series B 轮 | 软件控制的模块化磁体仿星器 | 用可量产磁体阵列化解线圈复杂度 | 必须证明软件控制磁场质量和反应堆集成 |
| Marvel Fusion | 激光惯性 / 德国同业 | > $400M 公司披露的公共和私人资金;按 TechCrunch,私人融资 $162M | 激光靶电站技术和 Fort Collins 基础设施 | 硅靶和快点火激光策略 | 资金口径不一,惯性路线经济性仍未证明 |
| Focused Energy | 激光惯性 / 德美同业 | 按 TechCrunch,Series A 轮 $240M,另有补助 | Biblis 首台激光,2030 年代试点电站路径 | NIF 点火传承和 RWE 场址准入 | 需要高重复率靶制造和激光经济性跑通 |
| Gauss Fusion | 欧洲磁约束联盟 / 仿星器相邻 | 公司提到 BMBF 支持的 €10M 氚增殖资金 | 面向欧洲聚变电站的产业联盟 | 联盟可触达德国、法国、意大利、西班牙的产业伙伴 | 私人融资和具体装置里程碑披露较少 |
| Renaissance Fusion | 仿星器 / 欧洲直接同业 | 保留来源对融资披露不一致 | HTS 线圈沉积、液态金属屏蔽、面向电网的仿星器 | 用直接沉积 HTS 线圈简化仿星器工程 | 资本规模和试点时间的公开证据比 Proxima 更薄 |
| Xcimer | 激光惯性 / 美国同业 | 按 TechCrunch 列表语境为 >$100M 级别;官方路线图强调 Phoenix | 10-MJ 级激光聚变路线图 | 大型准分子激光架构,连接已验证的 NIF 物理 | 资本和重复率经济性仍不确定 |
| Pacific Fusion | 脉冲磁惯性 / 美国同业 | 按 TechCrunch,Series A 轮超过 $1B,按里程碑支付 | 使用同步 Marx 发生器的脉冲功率惯性系统 | 大规模分阶段资本承诺和 Eric Lander 领衔团队 | 按里程碑分档的融资,以及极端脉冲功率时序风险 |
融资数值是截至 2026 年运行日期的公开来源快照;披露口径不同,有的给累计融资,有的给单轮规模, 有的给估值。
[CP001, CP006, CP008, CP009, CP011, CP012]| 路线 | 代表公司 | 核心承诺 | 主要风险 | 为何影响 Proxima |
|---|---|---|---|---|
| QI / 优化仿星器 | Proxima | 稳态运行、降低湍流,并有 W7-X 带来的可信度 | 复杂 3D 工程、线圈精度和制造爬坡 | 这是 Proxima 的核心差异化,也是主要技术尽调负担 |
| 其他仿星器 | Type One、Thea、Renaissance、Gauss 等仿星器同行 | 用不同简化策略跑稳态磁约束 | 同业可能更快或更便宜地解决线圈复杂度 | 直接挑战 Proxima「胜出仿星器路线」的说法 |
| HTS 托卡马克 / 球形托卡马克 | CFS、Tokamak Energy | 更成熟的托卡马克物理,叠加紧凑 HTS 磁体 | 破裂、脉冲 / 稳态运行、材料和氚系统 | 可能先打到并网里程碑,并占住投资人想象力 |
| FRC / Z 箍缩 / 磁化靶 | Helion、TAE、Zap、General Fusion 等路线对手 | 几何形态可能更简单,或可直接能量转换 | 等离子体稳定性、压缩机制和重复运行 | 不需要仿星器线圈,也能争夺同一批清洁稳定电力客户 |
| 激光 / 惯性 / 脉冲磁惯性 | First Light、Marvel、Focused、Xcimer、Pacific 等惯性路线对手 | 从 NIF 式点火或脉冲功率物理出发 | 驱动器效率、靶材量产、重复频率、腔室存活性 | 物理路线不同,仍可吸引大额融资和产业伙伴 |
本路线表把复杂反应堆项目压缩成尽调类别;它是决策地图,不是物理证明。
[CP003, CP006, CP009, CP011, CP012, CP013]Proxima 在仿星器专属性上处于高位,资本规模居中偏高;CFS 和 Helion 在全球资本 / 客户可见度上领先。
X 轴是路线与 Proxima 的接近度,1=其他 / 惯性、10=直接仿星器;Y 轴是公开资本 / 客户可见度,基于本报告保留来源做序数判断,不是经审计会计口径。
[CP006, CP009, CP015, CP016, CP020, CP022]3.2 Proxima 对比仿星器同行:QI、W7-X 传承、HTS 与 AI 线圈设计
直接的仿星器同行范围比完整核聚变地图窄。Type One 是最接近的美国类比,因为它也在商业化仿星器,并有一个与 TVA 相关的项目概念; Thea 从相反方向处理同一个仿星器历史弱点,用大量模块化磁体和软件控制替代复杂 3D 线圈;Renaissance 通过 HTS 沉积和液态金属屏蔽简化线圈制造; Gauss 则是一个面向磁约束电站交付的欧洲产业联盟。Proxima 的回答是另一组组合:准等动力等离子体物理、HTS 磁体、AI 加速的线圈和工程优化, 以及来自 Max Planck IPP 的 Wendelstein 7-X 机构传承。Stellaris 论文强化了这个故事,因为它把卖点转成可检查的电站概念, 而不只是公司单方面宣称。风险在于,每个仿星器同行也都在试图消除同一个复杂性反对意见,所以 Proxima 的护城河取决于执行速度和制造证明, 而不是“仿星器”这个词本身。[CP003, CP004, CP005, CP015, CP016, CP019]
| 购买标准 | Proxima | CFS | Helion | Type One | Thea / Renaissance | 惯性 / 脉冲同业 |
|---|---|---|---|---|---|---|
| 核心约束路线 | QI 仿星器 | HTS 托卡马克 | FRC / 脉冲直接转换 | 仿星器 | 仿星器变体 | 激光或脉冲惯性变体 |
| 稳态叙事 | 强 | 中 | 中 / 脉冲 | 强 | 强 | 低 / 脉冲重复挑战 |
| HTS 磁体依赖 | 高 | 高 | 低 | 可能较高 | Thea/Renaissance 为高 | 激光同业为低;Pacific 不定 |
| 复杂线圈风险 | 高,但经 QI/AI 优化 | 3D 线圈风险较低 | 不同的 FRC 稳定性风险 | 高 | 明确的简化策略 | 不同的靶 / 激光 / 脉冲功率风险 |
| 同行评审 / 已发布电站概念 | Stellaris 概念自称首个经同行评审的商用仿星器电站概念 | SPARC/ARC 公开技术路线图 | 公开技术页面,但传统电站验证较少 | 公开项目路线图和 IEEE 审视 | 公开简化主张 | 激光同业有 NIF 相邻物理基础 |
| 战略客户 / 场址准入 | RWE 场址路径和 Google 投资 | Google ARC 承购 | Microsoft 客户和 OpenAI 兴趣 | TVA 相关模式 | 在保留来源中可见度较低 | Focused 有 RWE/Biblis;其他情况不一 |
| 资金深度 | 欧洲较高 | 全球很高 | 估值很高 | 中高 | 阶段更早 | 范围很宽:Focused/Marvel/Pacific/Xcimer 比许多玩家资金更充足 |
| 公开定价证据 | 无产品定价;收入前 | 无反应堆定价 | 无反应堆定价 | 无反应堆定价 | 无反应堆定价 | 无反应堆定价 |
矩阵评级是基于公开来源的序数化解读;没有公开支撑的商业定价单元格明确标为不可得,而不是推断填入。
[CP003, CP004, CP005, CP006, CP007, CP009]直接仿星器竞赛的关键,在于复杂磁场究竟靠 QI/AI、模块化磁体、沉积 HTS 线圈,还是工业联合体执行来解决。
[CP003, CP004, CP005, CP006, CP007, CP009]3.3 融资、估值与客户入口不对称
Proxima 2026 年 7 月融资让它进入全球牌桌,并似乎使其成为欧洲融资最充足的核聚变公司,但它仍比美国最大竞争者更小、更年轻。 CFS 融资接近 $3B,正在推进 SPARC 到 ARC 的托卡马克路线图;Helion 披露估值 $15.5B、累计融资 $1.5B,并把 Microsoft 指为首个客户; Pacific Fusion 披露了规模很大的分阶段 A 轮融资;TAE 的运营历史也多出近 20 年。RWE 和 Google 参与帮助 Proxima 缩小商业入口差距, 因为这把公司连到德国电网站址路径和战略电力需求买家;但对手的购电和项目公告抬高了尽调门槛。承保问题不是 Proxima 是否融到一大轮, 而是 Alpha、模型线圈、HTS 生产和制造系统能否在资金更充足的替代方案锁住伙伴、人才和供应链产能之前,把资本转成里程碑。[CP001, CP002, CP006, CP007, CP008, CP009]
| 公司 / 同类组 | 用作代理指标的资本信号 | 商业化包装信号 | 渠道优势 | Proxima 启示 |
|---|---|---|---|---|
| Proxima Fusion | 2026 年 €411M 融资;投后估值 €2.4B;含补助总额 €650M+ | 未来电站 / 开发商模式;暂无产品定价 | RWE、Google、巴伐利亚、IPP 生态 | 欧洲背书强,但仍需有约束力的项目经济性 |
| CFS | 融资接近 $3B | 自持 / 运营 ARC 电站;Google 购电 | 美国项目、Google、深厚投资人联合体 | 为资本厚度和项目可信度设定标杆 |
| Helion | 2026 年 $465M 融资,估值 $15.5B,累计融资 $1.5B | 向 Microsoft 售电;电力销售叙事直接 | Microsoft、Altman / OpenAI 关联 | 更高估值和客户叙事可能吸走人才与资本 |
| TAE | TechCrunch / PitchBook 称,合并前累计 $1.79B | 技术平台及最终电站平台 | 长期投资人和企业关系 | 运营成熟度凸显 Proxima 仍年轻 |
| Pacific Fusion | Series A 轮超过 $1B,按里程碑支付 | 按里程碑融资的脉冲功率电站开发 | 大规模分阶段资金池 | 可激烈争夺惯性聚变人才和供应商 |
| General Fusion | 融资超过 $600M | 磁化靶电站路线图 | 加拿大基地和长期投资人历史 | 显示韧性,也暴露开发周期漫长 |
| Type One Energy | 已融资 $269M,并在募集更大规模 Series B 轮 | 围绕仿星器技术的公用事业建造 / 拥有 / 运营模式 | TVA 与英国财团信号 | 最直接的美国仿星器商业化同业 |
| Thea Energy | $100M Series B 轮 | 模块化磁体仿星器开发商 | Princeton / PPPL 渊源和美国投资人基础 | 直攻 Proxima 的线圈复杂度痛点 |
| Focused Energy / Marvel Fusion | Focused 完成 $240M Series A;Marvel 获得公私资金 >$400M | 激光聚变基础设施和靶材生态 | Focused 有 RWE / Biblis;Marvel 有 Siemens Energy | 德国激光聚变同业可争夺公共资金和产业伙伴 |
| Gauss / Renaissance | 公私资金透明度较低 | 欧洲仿星器或磁约束技术平台 | 产业财团或 Grenoble 生态 | 欧洲同业具参考性,但可见资本规模较低 |
保留来源均未提供可比反应堆定价;本表用公开融资、场址 / 客户和商业包装信号,作为竞争强度的代理指标。
[CP001, CP002, CP006, CP008, CP009, CP010]公开融资信号显示,Proxima 在欧洲领先,但仍低于美国最大核聚变公司的资本堆叠。
私营公司披露口径不一致,图中混合使用融资总额和估值;EUR 按报告共用换算指引折算。
[CP001, CP002, CP006, CP008, CP009, CP011]3.4 护城河耐久性:差异化真实存在,但尚未在电站尺度证明
Proxima 最强的竞争论点,是它把稳态仿星器架构与 QI 优化、HTS 磁体、AI 设计工具、W7-X 学习曲线,以及异常强的欧洲公私支持组合在一起。 反向论点同样直接:公司年轻、收入前、融资少于 CFS 和 Helion,并身处拥挤赛场;其他架构可能先达成可见并网里程碑。仿星器复杂性仍是核心风险。 Thea 和 Renaissance 明确营销简化策略,而 CFS 和 Helion 通过不同约束选择绕开同一个 3D 线圈问题。供应链证据也指向先进组件、真空容器、 电力电子、第一壁材料和燃料循环基础设施的瓶颈;这些瓶颈会拖慢所有路线,尤其拖慢需要定制磁体和精密制造的硬件设计。因此,尽调应聚焦模型线圈性能、 制造良率、供应商集中度、时间线现实性,以及 RWE/Google 支持是否会转化为具约束力的项目经济性,而不只是战略信号。[CP003, CP005, CP026, CP027, CP028, CP030]
| 护城河主张 | 威胁 | 严重性 | 重要性 | 缓释 / 尽调索取 |
|---|---|---|---|---|
| QI 仿星器物理叠加 W7-X 传承 | Type One、Thea、Renaissance 和 Gauss 也主打仿星器简化 | 高 | 若同业解决线圈成本和可制造性,架构级差异化会收窄 | 审查样机线圈测试结果、优化基准和制造良率数据 |
| 欧洲融资最多的聚变公司 | CFS、Helion、TAE 和 Pacific 的全球资本信号更强或相当 | 高 | 资本厚度决定人才、供应商、项目可信度和时间表韧性 | 用竞争对手融资能力对标从 Alpha 到首座电站的剩余现金需求 |
| AI 与计算线圈设计 | 软件驱动的线圈优化并非独家;Thea 和 Renaissance 也把复杂度转移到控制或制造 | 中 | 没有硬件证明,AI 主张会变成入场标配 | 审计从设计到制造的周期、仿真验证和公差叠加 |
| RWE / Google 战略入口 | CFS 有 Google 购电;Helion 有 Microsoft 及 OpenAI 关联需求叙事 | 中 | 战略 logo 不等于可融资的电站合同 | 拆分投资、MOU、场址准入、购电和项目融资义务 |
| Stellaris 电站概念可信度 | 纸面概念可能扛不住部件、供应链、燃料循环和第一壁约束 | 高 | 概念可信只是必要条件,不足以交付可执行电站 | 把每个 Stellaris 子系统映射到供应商、成本、进度和测试证据 |
风险严重性反映尽调优先级,不代表发生概率;承销前,每一行都需要私有技术和商业证据。
[CP003, CP004, CP005, CP006, CP009, CP015]3.5 图表
04财务
4.1 没有产品收入:售电前的模式是里程碑融资
Proxima 应按收入前深科技承保,而不是按 SaaS、市场平台或经常性基础设施业务承保。公司公开材料描述的是一条从工程工作走向 Alpha 的路径:Alpha 是计划在 2030 年代初落在慕尼黑附近的净能量仿星器演示装置,之后再走向 Stellaris,也就是十年后期的首座商业仿星器电站; 这些材料没有描述当前产品收入、ARR、已实现售电、毛利率或客户合同。从公开证据看,近期唯一可见的经济“单元”是里程碑完成: Stellarator Model Coil、HTS 线缆和磁体生产、Alpha 场址/建设,以及产业伙伴就绪度。RWE 和 Google 在战略上重要, 但 RWE 2026 年 7 月披露的是投资和场址/产业化关系,而不是收入合同。因此,今天无法给收入质量打分。尽调路径应在给予商业收入信用前, 拆分公共补助、股权融资、伙伴实物支持,以及任何未来购电或项目融资承诺。 [CI014, CI015, CI016, CI017, CI041, CI042]
| 潜在来源 | 机制 | 单位 | 当前价值 / 状态 | 质量信号 | 尽调索取 |
|---|---|---|---|---|---|
| 商业售电 | 未来销售 Stellaris 规模电站电力 | MWh / PPA | 当前无产品收入;电站目标在 2030 年代后期 | 索取已签署的承购、电价、PPA 或售电条款清单 | |
| Alpha 里程碑资金 | 股权和公共补助为示范机里程碑供资 | 里程碑 / 分期 | 资金正在使用 | 里程碑融资,不是收入 | 索取补助协议、提款条件和里程碑预算 |
| 战略能源伙伴关系 | RWE 场址 / 工业化关系和 Google 战略意向 | 伙伴承诺 | 仅投资 / 伙伴关系 | 支撑资本获取,但不是 ARR | 索取商业权利、排他性和未来承购经济性 |
| 技术许可或供应 | 未来可能把 HTS、仿星器或工程 IP 变现 | 许可 / 设备 | 无公开定价或合同 | 索取许可、部件销售和利润率假设路线图 | |
| 公共补助 | 巴伐利亚、德国、欧盟和公共基金支持战略技术 | 补助 / 类股权公共资本 | 公司披露 €95M 补助 | 若拨付可降低稀释,但增加政策依赖 | 索取已承诺与附条件补助明细,以及合规义务 |
null 值表示,已审阅公开来源未发现产品收入、ARR、定价或已确认收入;补助和伙伴投资是资金来源,不是客户收入。
[CI014, CI015, CI016, CI017, CI040, CI041]| 产品 / 资产 | 价格 / 单位 / 合同 | 标价与实际价格 | 公开证据 | 含义 |
|---|---|---|---|---|
| Alpha 示范机 | 不是商业产品 | 无实际价格 | 公司称 Alpha 是净能量示范机 | 按资本开支里程碑评估,不按收入评估 |
| Stellaris 电站 | 未来电站经济性未披露 | 无实际价格 | 商业电站计划在 2030 年代后期 | 未来 PPA / LCOE 经济性仍属推测 |
| RWE Gundremmingen 路径 | 无公开 PPA 价格 | 投资和场址合作,不是电价 | RWE 披露 €25M 投资和场址协作 | 有战略验证,但当前无收入 |
| Google 战略意向 | Proxima 无公开购电价格 | 仅投资意向 | 公司与 CNBC 提到 Google 参与 | AI 电力需求支撑投资论点,但不等于当前销售 |
| 公共补助 | 补助分期 / 公共共同出资 | 不是收入 | 公司披露 €95M 公共补助;巴伐利亚承诺另有报道 | 融资风险包含政策执行 |
Proxima 未披露标价、实际价格或当前客户收入来源;所有变现都处于未来状态,或与融资相关。
[CI001, CI004, CI006, CI014, CI041, CI042]当前融资先转化为技术里程碑,未来电站收入才可能出现。
由于没有公开产品定价、ARR 或 PPA 经济性,该桥接图为定性判断。
[CI014, CI015, CI016, CI017, CI025, CI041]4.2 资本形成在欧洲异常突出,但资本栈复杂
最强的财务证据,是 Proxima 能够快速组织一个泛欧洲且带战略属性的融资辛迪加。公开来源相互印证了 2025 年 6 月 €130M A 轮、2025 年 9 月 €15M 延伸轮使披露融资达到 €200M,以及 2026 年 7 月以 €2.4B 估值完成 €411M 融资。 2026 年 7 月轮由 XTX Ventures 和 East X Ventures 领投,RWE 和 Google 为战略投资方;公共和准公共支持者包括 KfW Capital、SPRIND、DTCF、Bayern Kapital、HTGF、CDP Venture Capital 和 EIC Fund。这是优势,因为核聚变公司在多个阶段都需要口袋够深的支持者; 但这也带来股权结构和治理问题。Northdata 和在线注册页将 Proxima Fusion GmbH 标识为 Munich HRB 283423,Northdata 列出 37 名已知活跃股东。公司可以可信地声称欧洲融资领先,但投资人尽调仍需要清算优先权、按比例跟投权、公共资金约束,以及战略投资者围绕 Alpha 或未来电站的任何权利。 [CI001, CI002, CI003, CI006, CI007, CI011]
| 日期 | 轮次 / 来源 | 金额 | 领投 / 重要投资人 | 包含的公共资本 | 财务解读 |
|---|---|---|---|---|---|
| 2023-05 | Pre-seed 轮 | 约 €7M-€7.5M | Plural、UVC Partners、HTGF、Wilbe、TOMORROW 等投资方 | Max Planck 生态支持 | 公司成立,以及从实验室转向创业公司 |
| 2024-04 | 种子轮 | €20M | redalpine;Bayern Kapital、DTCF、Max Planck Foundation 等投资方 | Bayern Kapital、DTCF、HTGF 体系公共出资方 | 从设计概念走向团队和伙伴关系 |
| 2025-06 | Series A 轮 | €130M | Cherry Ventures 和 Balderton | DTCF、Bayern Kapital、HTGF 参与 | 为 SMC 2027 和 Alpha 选址供资 |
| 2025-09 | Series A 延伸轮 | €15M;总融资 €200M | CDP Venture Capital、EIC Fund、Brevan Howard Macro 等投资方 | EICF;此前 €2.5M EIC 补助 | 增加欧洲主权公共基金支持 |
| 2026-07 | 大额私人融资 | €411M / $468M | XTX Ventures、East X Ventures、RWE、Google 等投资方 | KfW Capital、SPRIND、DTCF、Bayern Kapital、EIC Fund 等公共支持方 | 为 Alpha 提供支撑,但单独不足以覆盖整个示范机 |
| 2026-07 | 已披露公共补助基础 | €650M+ 总额中包含 €95M 补助 | 巴伐利亚和欧洲公共来源 | 公共补助 | 路线图中实质性补贴组成部分 |
本枚举并不完整,只覆盖对投资测算重要的公开股权 / 补助里程碑,不涵盖每一次股权发行或补助提款。
[CI001, CI003, CI004, CI007, CI008, CI009]| 资本来源 | 证据 | 资金用途 / 里程碑 | 若延迟的风险 | 尽调索取 |
|---|---|---|---|---|
| 2026 年 7 月私人融资 | €411M 融资,估值 €2.4B | Alpha、SMC、HTS 电缆 / 磁体、招聘 | 里程碑滑坡会迫使下一轮提前 | 索取交割后现金和董事会批准预算 |
| 公共补助 | 已披露 €95M 公共补助 | 抵消深科技开发成本 | 附条件补助可能滞后于支出 | 索取补助合同和拨付时间表 |
| 巴伐利亚承诺 | 报道称其为 €2B Alpha 计划贡献 €400M | Alpha 场址 / 测试设施 | 联邦共同出资缺口可能卡住设施 | 核验已签署的州级承诺和里程碑 |
| 德国联邦 / Fusion Action Plan | 报道称预期联邦份额为 €1.2B | Alpha 剩余公共融资组合 | 最大未落实公共分期 | 确认招标状态、预算科目和授予时间 |
| RWE 战略资本 | RWE 投资 €25M | 场址、工业化和审批经验 | 场址收益可能无法转化为收入 | 索取场址权利和电站经济性 |
| 未来项目融资 | 未披露 | FOAK 商业电站及厂区配套系统 | 若项目债务融资不可得,股权会被稀释 | 索取 2031 年之后的融资计划 |
资金用途包含公司披露用途和独立报道;巴伐利亚 / 联邦资金拆分需用一手赠款或招标文件确认。
[CI001, CI004, CI005, CI006, CI020, CI021]公开融资里程碑显示,公司从种子前轮快速跃升到 2026 年 €411M 融资。
Pre-seed 按 2024 年种子轮新闻稿中的 €7.5M 展示;总融资额是公司披露的四舍五入口径,不是仅把列出的股权轮次相加。
[CI001, CI003, CI007, CI009, CI010, CI011]4.3 资金跑道取决于 Alpha 烧钱,而不是普通运营开支
2026 年 7 月资产负债表应能买来时间,但不能消除融资依赖。Proxima 称新资本将支持 Alpha、模型线圈完成、HTS 线缆和磁体生产、招聘, 以及制造系统开发。围绕 Bavaria 相关 Alpha 计划的独立报道,指向一个约 €2B 的仿星器测试设施项目;Bavaria 和 Proxima 各承担 20%, 仍需要预期 €1.2B 联邦资金。即便这些报道只是方向性而非已签融资文件,它也框定了规模问题:€411M 是大额风险资本,但 Alpha 是项目融资级别的工程。 公开来源没有披露账面现金、月度现金消耗、已承诺资本开支、营运资本、采购定金、债务或公共补助的确切拨付计划。一个简单敏感性说明了为何这重要: 如果月度现金消耗为 €10M,€411M 约等于 41 个月资金跑道;如果为 €20M,约 21 个月;如果为 €30M,约 14 个月。 这些情形都不是公司指引,但它们说明资金跑道必须拿里程碑预算和公共资金时点来尽调验证。 [CI004, CI005, CI018, CI019, CI020, CI021]
| 指标 | 数值 / 状态 | 置信度 | 重要性 | 公开代理指标 | 尽调索取 |
|---|---|---|---|---|---|
| 产品收入 | 中 | 任何 P&L 模型的起点 | 无公开产品销售或上网电力收入 | 索取经审计收入台账,以及补助与收入分类 | |
| ARR | 中 | 检验经常性收入质量 | 无 ARR 披露,也无商业产品 | 索取任何已签约的经常性工程、许可或服务收入 | |
| 毛利率 | 中 | 检验长期经济性 | 无产品收入或 COGS 基础 | 索取建模 LCOE、电站 capex、O&M 和部件毛利率假设 | |
| 月度烧钱 | 未披露 | 低 | 决定现金续航和稀释时点 | 招聘、HTS、磁体和 Alpha capex 指向支出上升 | 按 R&D、capex、采购和 G&A 索取月度现金消耗 |
| 现金余额 | 未披露 | 低 | 决定融资后实际现金续航 | 披露的是 €411M 融资额,不是交割后现金 | 索取交割后资产负债表和受限现金明细 |
| 补助依赖 | 已披露 €95M 公共补助;Alpha 另有更大公共承诺报道 | 中 | 影响稀释和政策风险 | 公司及 HTGF / DTCF 披露 | 索取已承诺补助、条件和报销节奏 |
本表对产品收入、ARR 和毛利率刻意使用 null 而非 0,因为公司仍处收入前阶段,并非成熟指标为零。
[CI014, CI015, CI017, CI018, CI019, CI023]| 资本问题 | 公开答案 | 信号 | 重要性 | 尽调要求 |
|---|---|---|---|---|
| 在手现金 | 未披露 | unknown | 融资轮规模不等于可动用的非受限现金 | 索取交割后资产负债表和受限现金明细 |
| 月度烧钱额 | 未披露 | unknown | 在示例情形下,决定 €411M 能支撑 14、21 还是 41 个月 | 按工作流索取月度烧钱额和里程碑预算 |
| 跑道月数 | 未披露;若月烧 €30M–€10M,示例跑道为 14–41 个月 | 敏感 | 跑道很可能由里程碑驱动,而非稳态消耗 | 将现金与 SMC、HTS、Alpha 采购和招聘计划对齐核验 |
| 计划资金用途 | SMC、HTS 电缆 / 磁体、工程 / 制造系统、招聘 | 强 | 本轮融资绑定了具体硬件里程碑 | 索取里程碑预算和供应商承诺 |
| 下一轮触发条件 | SMC 2027、Alpha 2031、公共共同资助奖项 | 重大 | 失败或延期可能重置估值和稀释幅度 | 定义融资 go/no-go 里程碑和备用情形 |
| 债务 / 项目融资义务 | 已审阅公开来源未披露债务 | unknown | FOAK 电站可能需要非风投资本 | 索取债务、租赁、担保和项目融资条款书 |
跑道情形只是示例敏感性,不是公司指引;公开来源未披露实际现金消耗或现金余额。
[CI018, CI019, CI023, CI024, CI025, CI026]2026 年融资支撑近期硬件里程碑,但 Alpha 设施资金仍是一套更大的公私资金组合。
瀑布图结合公司披露和独立报道;巴伐利亚 / 联邦 Alpha 资金堆叠仍需一手赠款确认。
[CI004, CI005, CI020, CI021, CI022, CI028]€411M 这一轮能买来多长时间,取决于 Alpha 相关支出爬坡多快,差异很大。
烧钱情景仅用于尽调敏感性分析;Proxima 未披露现金余额或月度烧钱速度。
[CI018, CI019, CI023, CI024, CI025]4.4 同行融资验证了品类,同时凸显稀释和补贴风险
Proxima 现在是欧洲融资最充足的核聚变公司之一,但同行背景提醒,不应把这一轮融资过度解读为财务去风险证明。Commonwealth Fusion Systems 已融资接近 $3B,Helion 称已融资约 $1.5B,TechCrunch 2026 年调查列出多家融资数亿美元或更多的私人核聚变公司。这个比较有双向含义。 一方面,它支持严肃核聚变项目需要传统风险轮次可能无法单独覆盖的资本规模;另一方面,它也说明 Proxima 超过 €650M 的总资本,相对于美国领跑者仍属早期。 反向财务证据很重要:MIT Technology Review 警告核聚变可能不会很快变便宜,且尚无核聚变电站存在;Bulletin of the Atomic Scientists 强调商业核聚变一再比预期更慢,经济性核聚变没有保证。公开结论因此是平衡的。Proxima 融资动能强,公私战略支持异常有力;但十年以上收入路径、 没有产品收入、重资本开支、可能稀释和补贴依赖,仍是核心财务风险。 [CI026, CI029, CI030, CI031, CI032, CI033]
| 公司 | 已披露融资 / 资本 | 最新大额轮次 | 技术 / 状态 | 对 Proxima 的启示 |
|---|---|---|---|---|
| Commonwealth Fusion Systems | 接近 $3B | 2025 年 $863M Series B2 轮 | SPARC/ARC 托卡马克路径 | Proxima 规模仍小于全球融资领跑者 |
| Helion Energy | 约 $1.5B | 2026 年 $465M Series G 轮 | 激进的 2028 年发电目标 | 头部聚变竞争者出现超大轮次并不异常 |
| TAE Technologies | 宣布合并前为 $1.79B | 2025 年 $150M,另有合并相关资本 | 场反位形 | 长开发周期会吃掉很大的资本栈 |
| General Fusion | 约 $612M | 现金承压后 $22M pay-to-play,外加 SAFE 票据 | 磁化靶聚变 | 资本稀缺会迫使公司接受降价轮或续命融资 |
| Zap Energy | $327M | 近期转向裂变 / 混合收入选项 | Z 箍缩路径 | 部分同业在寻找更近端的收入替代方案 |
| Proxima Fusion | 含赠款 €650M+ | 2026 年 7 月 €411M | QI-HTS 仿星器 / Alpha | 欧洲融资最多,但仍低于美国头部公司 |
同业数字来自公开报道和公司公告;除头部融资语境外,未统一币种和估值口径。
[CI026, CI033, CI034, CI035, CI036, CI037]| 缺失的私有指标 | 影响 | 为何卡住投资判断 | 具体尽调路径 |
|---|---|---|---|
| 已确认收入和赠款会计政策 | 高 | 无法区分商业收入和报销型 R&D 支持 | 审阅经审计账目、赠款台账和收入确认备忘录 |
| ARR / 已签约订单储备 | 高 | 公开信息看不到经常性收入基础 | 索取已签合同、购电 / 承购意向书和订单储备明细 |
| 毛利率 / LCOE 模型 | 高 | COGS 和电站经济性均未公开 | 审阅 LCOE 模型、每 MW 资本开支、O&M、替换件和氚 / 燃料假设 |
| 现金、烧钱额和跑道 | 高 | 决定稀释幅度和下一轮时点 | 审阅轮后资产负债表、月度现金计划和里程碑驱动的烧钱情形 |
| 公共赠款条件 | 高 | 政策时点可能控制 Alpha 建设进度 | 获取巴伐利亚、联邦、EIC 和 DTCF 文件及拨款契约 |
| 股权结构表和清算优先权 | 中 | 大型多方财团可能含有影响未来融资的权利 | 审阅股东名册、优先权栈、按比例认购权和战略否决权 |
这些缺口是财务尽调阻断点;它们不否定 Proxima 的融资动能,但会卡住精确估值测算。
[CI013, CI014, CI015, CI017, CI018, CI019]公开尽调必须先从融资证明走到电站经济性,才谈得上为收入质量背书。
经济性风险分支依赖反向独立来源,而不是公司指引。
[CI020, CI029, CI030, CI031, CI032, CI043]4.5 图表
05产品与技术
5.1 QI-HTS 架构是技术论点,今天还不是可交付产品
Proxima 的技术栈最适合被框定为一套一体化反应堆设计架构。公司正在开发准等动力仿星器,用外部、复杂的三维磁场而不是大等离子体电流来约束等离子体。 这给了该架构相对托卡马克的核心优势:设计难得多,但运行概念瞄准稳态、无破裂,并降低等离子体控制负担。Proxima 加入高温超导磁体, 以相对更老的仿星器概念缩小装置,并用仿真驱动优化搜索一个过去难以处理的巨大设计空间。最强的产品技术证据,是公司官方技术页、Stellaris 论文公告、 DOE 解释材料和 W7-X 技术来源都指向同一个取舍:仿星器可以更容易连续运行,但线圈几何、工程集成和制造精度会变成硬问题。因此, 这是一个有吸引力但仍处商业化前的架构故事,不是经过验证的能源产品。 [CE001, CE002, CE003, CE004, CE005, CE006]
| 模块 / 资产 | 主要用户 | 状态 / 成熟度 | 差异化 | 尽调缺口 |
|---|---|---|---|---|
| QI-HTS 仿星器架构 | 反应堆物理和电站设计团队 | 技术论文验证了概念;Proxima 尚无运行中装置 | 以 HTS 支撑紧凑化,瞄准稳态仿星器运行 | 需要 Alpha 等离子体运行和净能量证明 |
| Stellaris 电站概念 | 工程团队、投资人、公用事业伙伴 | 经同行评议的概念发表于 Fusion Engineering and Design | 整合电磁、结构、热和中子学仿真 | 需要完整工程设计、许可路径、成本模型和可维护性证明 |
| Alpha 演示装置 | Proxima、IPP、巴伐利亚、RWE | 计划建在 Garching 附近的演示装置 | 目标是在仿星器中证明净能量,并验证关键技术 | 尚未建成;进度、联邦资金和监管路径仍未确定 |
| HTS 磁体和 Stellarator Model Coil | 磁体工程和制造团队 | SMC 计划早于 Alpha 完成 | 相比传统设计,HTS 磁场可能缩小仿星器尺度 | 需要仿星器专用线圈演示和长时程辐照数据 |
| StarFinder / 仿真工作流 | 设计优化团队 | 公司声称的云端框架 | 可在大型 QI 设计空间内快速迭代 | 需要内部基准、验证,以及从设计到制造的可追溯性 |
| ConStellaration 开放基准 | ML 和等离子体优化社区 | 公开代码、数据集、挑战赛和预印本 | 为 QI 优化带来外部开发者信号 | 开放等离子体边界基准不能证明反应堆可制造性 |
成熟度标签区分公开概念证明、开放优化资产和尚未运行的硬件。
[CE001, CE002, CE008, CE009, CE010, CE019]| 维度 | QI 仿星器 / Proxima 论点 | 托卡马克基线 | 尽调含义 |
|---|---|---|---|
| 磁场来源 | 外部 3D 线圈生成约束场 | 等离子体电流加线圈生成扭曲磁场 | 仿星器把复杂度从运行转移到设计和制造 |
| 运行模式 | 瞄准稳态连续运行 | 脉冲运行或非感应电流驱动更难 | 只有连续高性能运行可以放大时,Proxima 的差异化才最关键 |
| 破裂风险 | 没有大环向等离子体电流;可消除电流驱动破裂 | 托卡马克等离子体电流可能发生内部破裂 | 安全性和可用性论点可信,但仍需要 Alpha 证明 |
| 设计复杂度 | 自由度多,非平面线圈复杂 | 轴对称几何更简单 | AI/ML 和数值优化是核心,不是可选项 |
| 功率密度路径 | HTS 高磁场旨在缩小仿星器尺寸 | HTS 托卡马克也在追求紧凑高场设计 | HTS 是共同使能项,不是 Proxima 独有护城河 |
该对比综合 DOE、IPP、Proxima 和 W7-X 技术证据,并不假定某一概念已获商业验证。
[CE001, CE002, CE004, CE005, CE006, CE029]架构分层覆盖优化、QI 等离子体物理、HTS 磁体、3D 线圈、排热、包层和分阶段电站。
[CE001, CE003, CE008, CE009, CE019, CE032]公开信息中,物理积累和开放优化得分好于反应堆硬件、燃料循环和许可就绪度。
定性准备度评级仅基于公开证据,不基于 Proxima 私下设计评审。
[CE008, CE013, CE015, CE019, CE025, CE032]5.2 Stellaris 和 Alpha 把 W7-X 传承转成分阶段证明计划
Stellaris 给 Proxima 提供了一个比多数早期核聚变创业公司更具体的技术产物:一个经同行评审的一体化电站概念,把 QI 等离子体设计、 HTS 磁体、支撑结构、热管理和中子学放进同一个连贯的设计主张。该概念仍是纸面设计,但其覆盖面重要,因为它迫使等离子体物理和工程约束进入同一优化回路。 Alpha 是下一步验证,计划建在 Garching 的 IPP 附近,被描述为净能量演示装置,而不是商业电站。路线图随后指向 Alpha 之后在前 Gundremmingen 裂变核电站址建设 Stellaris,RWE 和 Bavaria 参与场址、融资和产业化工作。Wendelstein 7-X 是科学传承: IPP 的装置已经证明优化仿星器的相关性,创下长脉冲纪录,并在 2026 年发表经同行评审的托卡马克级三重积性能。尽调问题在于, W7-X 验证的是物理案例的一部分,不是 Proxima 完整反应堆系统;Alpha 也尚未运行。 [CE008, CE009, CE010, CE011, CE012, CE013]
| 日期 / 阶段 | 功能 / 里程碑 | 状态 | 含义 | 来源 |
|---|---|---|---|---|
| 2015-2026 | W7-X 运行和创纪录实验活动 | 外部传承活跃 | 支撑优化仿星器物理案例 | IPP 与 APS |
| 2025 | Stellaris 论文和概念公告 | 已发表概念 | 把物理和工程约束压进同一套设计 | Proxima、ScienceDirect / DOI、BusinessWire |
| 2025-2026 | ConStellaration 数据集、仓库、挑战赛和预印本 | 公开开发者信号 | 向 ML 社区开放 QI 优化 | Proxima、GitHub、Hugging Face、arXiv 资料 |
| 2027 | Stellarator Model Coil 演示磁体 | 已规划 | 第一个重要 HTS 磁体降险里程碑 | Proxima |
| 2031 / 2030 年代初 | Alpha 净能量演示装置 | 已规划 | 测试净能量仿星器物理和关键子系统 | Proxima、DTCF、NEI |
| 2030 年代后期 | Stellaris 并网电站 | Alpha 之后规划 | 若实现,将从演示装置走向商业电站 | Proxima、RWE / 巴伐利亚 MoU |
路线图日期是公开目标,应视为有风险的里程碑,而非承诺交付日期。
[CE010, CE011, CE012, CE014, CE016, CE017]公开路线图从 W7-X 和 Stellaris 论文证据推进到 SMC、Alpha,再到后续 Stellaris 电站。
未来日期是公司和合作伙伴目标,不是已完成里程碑。
[CE010, CE011, CE012, CE014, CE016, CE017]公开证据显示概念准备度强,但一体化反应堆准备度仍低。
这个 KPI 使用分类准备度标签,因为未找到 Proxima 公开 TRL 审计。
[CE013, CE015, CE019, CE021, CE025, CE032]5.3 开放优化资产围绕艰难设计瓶颈,释放出有意义的开发者信号
Proxima 的公开开发者信号,与其核心技术风险异常相关。其 GitHub 组织列出 VMEC++、射线追踪、开放仿星器模型和 ConStellaration 相关代码库;ConStellaration 代码库、Hugging Face 数据集、PyPI 包和 arXiv 预印本则暴露了一个用于 QI 类仿星器等离子体边界的优化基准。 该数据集按不同来源呈现被描述为超过 160,000 到 182,000 行,并配有 ideal-MHD 平衡态和指标;相关挑战要求机器学习社区优化几何上更简单、 更易建造的 QI,以及 MHD 稳定的 QI 仿星器。这不能证明 Proxima 能制造线圈或实现净能量,但验证了设计工作流周围确有真实从业者界面。 因此,今天的客户工作流是内部且面向生态的:定义目标等离子体边界,评估物理和工程指标,简化与线圈相关的几何,并把有前景的设计送入更高保真工程。 公开开源活动降低了黑箱风险,同时也确认设计问题仍然高维且计算昂贵。 [CE019, CE020, CE021, CE022, CE023, CE024]
| 用户任务 | 当前工作流 | Proxima 方案 | 可衡量收益 | 局限 |
|---|---|---|---|---|
| 搜索 QI 设计空间 | 专家主导数值优化 | StarFinder 加开放基准工具 | 更快迭代高维仿星器候选方案 | 没有公开的端到端基准来证明电站级设计速度 |
| 评估等离子体边界 | VMEC 和理想 MHD 仿真工作流 | ConStellaration 数据和 VMEC++ 评估工具 | 为 ML 优化提供共同指标和基线 | 边界质量位于线圈和包层可行性的上游 |
| 简化仿星器可建造性 | 物理优化后手工权衡工程取舍 | 面向更易建造 QI 形状的基准问题 | 明确权衡紧凑性和线圈简洁度 | 制造公差仍未证明 |
| 验证净能量概念 | 实验室实验和仿真外推 | IPP Garching 附近的 Alpha 演示装置 | 对关键聚变技术做真实环境测试 | Alpha 尚未运行,也不是电站 |
| 走向并网电站 | 公用事业级电站开发 | RWE 提供场址支持,在 Gundremmingen 建 Stellaris | 借助既有工业电站经验和基础设施 | 需要许可、燃料循环、资本和 Alpha 成功 |
该表描述 Proxima 的内部和生态工作流;目前还没有商业客户工作流。
[CE018, CE019, CE020, CE021, CE022, CE023]当前流程是从设计到验证的闭环,而不是商业电力交付。
[CE019, CE020, CE021, CE022, CE023, CE024]5.4 子系统风险集中在磁体、包层、材料、燃料循环和制造
偏反向的技术观点是,Proxima 有可信的设计论点,但在达到可投资的商业就绪度前,仍有数个未解决证明点。HTS 磁体必须在仿星器几何中被验证, 先通过 2027 年 Stellarator Model Coil,再走到反应堆尺度。复杂非平面线圈和支撑结构必须满足毫米级磁场公差,同时仍可制造、可检查、可维护。 中子包层和内部热管理表面必须适配仿星器的 3D 几何,氘-氚燃料路径最终还需要氚增殖和抗中子材料,而这仍是全行业瓶颈。MIT 2025 年 REBCO 辐照结果排除了一个瞬时临界电流担忧,但明确把多年暴露下的长期退化留作后续工作。因此,风险登记表把 Proxima 视为技术上差异化但远未去风险: 净能量尚未演示,Alpha 融资和监管仍未完成,Stellaris 还排在成功演示装置之后多年。 [CE032, CE033, CE034, CE035, CE036, CE037]
| 层级 / 组件 | 作用 | 依赖 | 风险 |
|---|---|---|---|
| QI 等离子体构型 | 核心约束几何 | 数值优化,并用仿星器物理验证 | 性能好的等离子体形状仍可能很难用线圈实现 |
| 3D 外部线圈系统 | 不靠等离子体电流,生成扭曲磁场 | 毫米级线圈精度和结构支撑 | 制造复杂度和磁场误差容忍度 |
| HTS 磁体 | 支撑更高磁场和更小装置 | REBCO 带材供应、低温工程、接头、辐照数据 | 长期退化和仿星器集成尚未证明 |
| 岛式偏滤器 / 排热 | 管理等离子体边缘和热负荷 | W7-X 传承和 Proxima 工程适配 | 电站热通量仍是艰难的集成测试 |
| 中子包层 | 捕获中子、屏蔽结构,并可能支持增殖 | 材料、几何、氚化学 | 复杂仿星器几何会让包层设计更难 |
| 仿真和优化栈 | 协同优化物理和工程约束 | 类 VMEC 代码、算力、ML、验证数据 | 仿真置信度必须经得住硬件和等离子体现实检验 |
架构行基于公开证据,避免推断未披露的专有内部细节。
[CE001, CE003, CE006, CE009, CE015, CE019]| 控制 / 认证 | 状态 | 范围 | 缺口 |
|---|---|---|---|
| 经同行评议的 Stellaris 论文 | 发表在 Fusion Engineering and Design | 集成式 QI-HTS 电站概念 | 同行评议不等于运行证明或建设验证 |
| W7-X 实验传承 | 有公开记录的活跃 IPP 项目 | 优化仿星器物理,与长脉冲相关 | W7-X 不是 Proxima 发电演示装置 |
| 开源开发者界面 | 可在 GitHub、Hugging Face、PyPI、arXiv 看到 | 优化代码、数据、基准和 VMEC++ | 开放基准不能证明完整反应堆工程 |
| 与巴伐利亚、RWE 和 IPP 的 MoU | 2026 年签署 | 场址、科学、工业、融资和许可协作 | 不等同于已获资金支持的建设批准 |
| 聚变监管和燃料控制 | 商业化前 | 许可、氚处理、中子材料、牌照 | 欧洲聚变监管路径和燃料循环证据仍不完整 |
质量控制是证据触点,不是电站安全或商业就绪度认证。
[CE008, CE011, CE014, CE015, CE016, CE019]| 风险 | 当前证据 | 严重性 | 触发条件 / 里程碑 | 尽调路径 |
|---|---|---|---|---|
| 净能量尚未证明 | Alpha 计划证明净能量,但尚未运行 | 高 | Alpha 等离子体实验活动 | 审阅 Alpha 设计依据、Q 目标、功率核算和独立评审 |
| HTS 磁体规模化 | SMC 计划于 2027 年完成;MIT 结果缓解了一项 REBCO 辐照担忧 | 高 | 仿星器模型线圈与 Alpha 磁体采购 | 索取线圈测试计划、失超保护数据、接头、供应协议和辐照裕度 |
| 复杂 3D 线圈制造 | DOE 指出,仿星器线圈需要毫米级精度 | 高 | 首件生产线圈与计量验收 | 审查制造流程、公差预算、检测和返工经济性 |
| 中子包层和材料 | Stellaris 包含复杂几何包层方案 | 高 | 一体化包层和材料资质验证 | 审查中子学、结构寿命、远程维护和测试设施使用权 |
| 氚增殖和燃料循环 | 全行业氚短缺和增殖就绪度仍未解决 | 高 | 闭合燃料循环设计评审 | 验证增殖比、锂-6 供应、氚处理和监管路径 |
| 许可与资金 | MoU 写明合作方和融资意向;联邦资金与聚变许可仍未落定 | 中 | Alpha 建设决策 | 跟踪德国联邦承诺、许可节点和 RWE/Bavaria 资金文件 |
风险登记表刻意偏反向,因为反应堆层面的里程碑尚未消除净能量或电站集成风险。
[CE011, CE012, CE014, CE032, CE033, CE034]Alpha 和 Stellaris 依赖物理验证、HTS 线圈、材料、燃料循环、监管和工业伙伴。
[CE011, CE012, CE014, CE032, CE033, CE034]5.5 图表
06客户
6.1 “客户”基础是未来购电方和使能伙伴,而不是当前买家
Proxima 仍处收入前阶段,所以正确的尽调框架不是装机客户质量或账户扩张,而是可信未来买方和使能方是否正在围绕首座电站组织起来。 最强公开证明是 RWE:它是战略投资方,是具备电站建设和运营能力的公用事业公司,也是 Proxima 计划建设商业 Stellaris 电站的前 Gundremmingen 核电站址所有者。与巴伐利亚自由州、RWE 和 Max Planck IPP 的 MoU 定义了一条路径: 先在 Garching 附近做 Alpha,如果技术跑通,再走向 Gundremmingen 并网仿星器。这比标志墙更接近场址/购电市场验证, 但仍不等于经常性收入、具约束力的 PPA 或已交付的兆瓦时。因此,买方/用户/付款方角色被拆开:RWE 是场址和公用事业伙伴, 公共部门资金方是基础设施赞助者,IPP 是科学伙伴,最终电网或工业电力买家仍是尚未签约的需求池。[CU001, CU002, CU003, CU004, CU005, CU006]
| 客群 | 买方 / 用户 / 付款方 | 用例 | 规模信号 | 战略价值 | 缺口 |
|---|---|---|---|---|---|
| 公用事业 / 场址合作方 | RWE 管理层、项目开发团队、未来电网运营商对接方 | 复用 Gundremmingen,推动聚变工业化 | RWE 投资 €25M,并拥有退役场址 | 最能为未来购电和场址可信度背书 | 未披露有约束力的 PPA 或电价 |
| 公共部门基础设施出资方 | Bavaria 自由州及潜在德国 / EU 项目 | 共同出资 Alpha、选址、许可、地区就业 | Bavaria 表示可能出资 20%;德国枢纽竞标待定 | 降低非稀释资金和政策风险 | 资金取决于联邦拨款和项目审批 |
| 科研 / 技术合作方 | Max Planck IPP 领导层和 Proxima 工程团队 | Alpha 等离子体物理和 W7-X 知识转移 | IPP 担任 Alpha 科学负责人 | 增强技术路线图可信度 | 研究合作不等于客户收入 |
| AI / 数据中心清洁电力需求 | Google 能源和基础设施团队、未来超大规模买方 | 为负载增长提供长时、稳定清洁电力 | Google 参与投资,并另行签署 CFS 聚变 PPA | 显示买方愿意采购聚变电力 | 未披露 Proxima-Google 购电协议 |
| 工业和电网买方 | 德国 / EU 公用事业公司、工业企业、数据中心 | 未来来自 Stellaris 的稳定清洁电力 | 商业电站目标在 2030 年代末并网 | 若聚变能以目标成本跑通,TAM 很大 | 没有已签约容量或价格 |
| 聚变 PPA 可比案例 | Microsoft、Google、CFS、Helion 交易对手方 | 首创型购电承诺模板 | 200 MW Google-CFS 和 50 MW Helion-Microsoft 先例 | 显示市场设计路径 | 可比交付仍有条件 |
分层指向未来需求和赋能合作方,因为截至运行日期,Proxima 没有付费客户或产品收入。
[CU001, CU002, CU003, CU005, CU009, CU014]| 对手方 | 客群 | 部署 / 用例 | 生产与试点 | 结果 | 限制 |
|---|---|---|---|---|---|
| RWE | 公用事业 / 场址和未来购电合作方 | Gundremmingen 场址、电站专业经验、审批经验 | 商业化前协议 | €25M 投资,并愿意提供核基础设施 | 未披露购电合同 |
| Bavaria 自由州 | 公共部门赞助方和监管对接合作方 | 选址、州共同融资、区域聚变枢纽 | 商业化前政策合作 | 潜在 20% 州级出资和高层政府背书 | 资金取决于联邦支持和项目成功 |
| Max Planck IPP | 科学和研究合作方 | Alpha 等离子体物理主导权和 W7-X 积累 | 研究合作 | MoU 将其列为 Alpha 科学负责人 | 不是买方或收入来源 |
| 战略投资方 / 未来清洁电力买方信号 | AI 数据中心稳定清洁电力需求 | 投资信号,不是 Proxima 购电 | 参与 Proxima 轮次,并另行签署 CFS 聚变 PPA | 没有 Proxima 专属电力合同 | |
| 德国联邦 / EU 项目 | 潜在公共资金和枢纽支持 | 磁约束聚变枢纽和商业化支持 | 招标 / 政策支持 | 德国和 EU 政策将聚变列为战略优先事项 | 拨款规模和条件仍未落定 |
| 未来电网 / 工业买方 | 潜在公用事业、工业和数据中心电力客户 | 购买 Stellaris 商业聚变电力 | 尚未签约 | 可比 PPA 表明,只要聚变跑通,买方有需求 | 没有价格、容量或 COD 承诺 |
表中只是公开具名的未来客户和赋能合作方证据样本,不是收入客户名单。
[CU001, CU002, CU003, CU004, CU005, CU006]Proxima 的 GTM 路径先走机构合作,再做技术验证;电力买家要到风险下降后才可能贡献收入。
[CU001, CU002, CU003, CU004, CU010, CU011]6.2 采用轨迹是一条从 MoU 到演示装置再到 2030 年代末并网电力的里程碑漏斗
采用轨迹更像基础设施开发,而不是企业销售。公开来源识别出一条分阶段路径:Garching 的 IPP 附近的 Alpha 旨在 2030 年代初演示净能量; Stellaris 将在 2030 年代末跟进于 Gundremmingen;Proxima 和伙伴也在争取德国联邦磁约束核聚变枢纽支持。RWE 2026 年 7 月投资强化了 2 月协议,因为它在早期场址和合作信号上加入了资金。Bavaria 新闻简报和 Proxima/IPP 发布稿把公共部门需求案例说得很直白: 能源安全、工业就业、AI 和数据中心负荷,以及欧洲技术领导力。Google 的参与不是 Proxima 购电合同,但它是可信的战略信号,因为 Google 已另行与 Commonwealth Fusion Systems 签署 200 MW 核聚变 PPA,并面临为数据中心采购清洁稳定电力的压力。因此,这条轨迹令人鼓舞但二元性强: 只有 Alpha 验证成功,且可融资电站能够获批、融资、并网并售电,采用才会变成商业。[CU010, CU011, CU012, CU013, CU014, CU015]
| 指标 | 值 | 日期 | 来源 | 置信度 | 含义 | 缺失分母 |
|---|---|---|---|---|---|---|
| 付费客户 | 披露为 0 / 未见证据 | 2026-07-11 | 公开来源审查 | 中 | 客户牵引力只能视为未来期权价值 | 没有客户台账或收入披露 |
| RWE 战略投资 | €25M | 2026-07-07 | RWE 新闻稿 | 高 | 公用事业合作方从 MoU 信号走到资本承诺 | 未披露项目股权条款或购电价格 |
| Alpha 示范堆时间表 | 2030 年代初 / 目标约 2031 年 | 2026 年公开发布 | Proxima、IPP、WNN | 中 | 技术验证推进前,商业采用无法启动 | 尚无已验证净能量的仿星器 |
| Stellaris 商业电站时间表 | 2030 年代末 / 外部分析称不晚于 2040 年 | 2026 年公开发布 | Proxima、RWE、Neutron Bytes | 中 | 收入窗口还在十年尺度之外 | 未披露 COD、并网或 PPA |
| Bavaria 潜在项目出资 | 项目成本的 20% / Bavaria 相关报道称最高约 €400M | 2026-02-26 | Proxima 与 Bavaria | 中 | 公共部门需求信号支撑场址和枢纽申标 | 联邦 / 州最终拨款尚未公开 |
| Google 战略需求信号 | Proxima 投资方;另有 200 MW CFS 聚变 PPA | 2025-2026 | CNBC、PRNewswire、DCD | 中 | 超大规模买方需要稳定清洁电力 | 没有 Proxima 专属的 Google PPA |
表内数值混合一手来源里程碑和独立报道;客户收入为 null,表示审阅到的公开披露没有客户收入,不是公司文件口径。
[CU006, CU010, CU011, CU012, CU013, CU014]| 里程碑 | 运行日期状态 | 客户相关性 | 下一验证点 | 延迟风险 |
|---|---|---|---|---|
| 2026 年 2 月 MoU | 已与 Bavaria、RWE 和 IPP 签署 | 形成首条具名场址 / 合作方路径 | 披露最终项目协议 | 信号仍无约束力或带条件 |
| RWE 2026 年 7 月投资 | 宣布 €25M 战略投资 | 强化公用事业合作方承诺 | 项目股权、购电或基础设施合同条款 | RWE 仍只是战略投资方,不是买方 |
| 联邦磁约束聚变枢纽申标 | 公开材料显示已提交 / 待定 | 可能为场址提供资金和合法性背书 | 拨款决定和条件 | 公共资金缺口会推迟 Alpha/Stellaris |
| Alpha 示范堆 | 计划在 2030 年代初建于 Garching 附近 | 任何客户信任 Stellaris 电力前,都需要先看到该验证 | 净能量仿星器运行 | 商业购电缺乏可融资性 |
| Stellaris 电站 | 计划在 2030 年代末建于 Gundremmingen | 潜在首个收入产品 | 许可、并网、PPA、项目融资 | 客户需求仍属推测 |
状态行按里程碑组织,因为 Proxima 没有可跟踪的产品部署或付费账户。
[CU001, CU002, CU003, CU004, CU010, CU011]商业客户路径长达十年,并受公共资金、Alpha 和 Stellaris 里程碑约束。
未来日期是公开来源披露的目标窗口,不是合同交付日期。
[CU010, CU011, CU012, CU013, CU020, CU022]RWE 是最强直接客户验证信号;超大规模云厂商和公共部门也有意义,但不够直接。
基于公开承诺直接性的 1-5 序数评分,不代表市场规模或收入。
[CU003, CU004, CU005, CU014, CU015, CU022]6.3 具名证明在交易对手上可信,但购电尚未商业化
以具名交易对手看,Proxima 对一家年轻核聚变公司而言质量异常高:RWE、巴伐利亚政府、Max Planck IPP、Google、KfW Capital 和 EIC Fund 都是有分量的机构。不过,尽调标准应把“伙伴意愿证明”和“按价格与数量证明客户需求”分开。RWE 材料称,如果该场址赢得联邦枢纽, 它将提供核基础设施和监管专长;Proxima 发布稿称,各方将就许可、项目结构、融资和建设角色合作。这是相当实质的未来客户证明,因为它把首座电站锚定在真实公用事业场址。 相比之下,Google 目前是投资者和市场需求信号,不是已披露的 Proxima 电力买家。可比案例显示未来模式可能是什么样:Google-CFS 宣布了 ARC 的 200 MW PPA,Helion-Microsoft 宣布了首个核聚变 PPA,目标至少 50 MW。这些交易验证了买方对核聚变电力的需求意愿, 但也凸显 Proxima 当前缺口:Stellaris 尚未披露同等具约束力的商业购电协议。[CU020, CU021, CU022, CU023, CU024, CU025]
| 交易 / 先例 | 买方信号 | 容量 / 目标 | 与 Proxima 的相关性 | 限制 |
|---|---|---|---|---|
| Google 与 Commonwealth Fusion Systems | 超大规模厂商承诺采购未来稳定清洁电力 | 来自规划 ARC 电站的 200 MW | 表明 Google 不只是投资方,也可能成为聚变电力买方 | 对象是 CFS,不是 Proxima;取决于 SPARC/ARC 里程碑 |
| Helion 与 Microsoft | 云买方签下首个聚变 PPA 式协议 | 最早 2028 年至少 50 MW | 表明数据中心客户愿意提前签聚变协议 | 市场认为交付日期高度激进 |
| RWE 与 Proxima MoU / 投资 | 公用事业场址和项目开发合作方 | 未披露 MW | 最接近 Proxima 首座电站购电 / 场址证明的专属路径 | 没有公开电价、容量或购买义务 |
| Google 对 Proxima 投资 | 战略能源需求信号 | 投资金额未披露 | AI / 数据中心需求支撑其对先进清洁电力的兴趣 | 不是购电协议 |
| 公共部门枢纽资金 | 政府共同出资示范基础设施 | 披露的是资金占比,不是电力容量 | 可降低通往可融资项目的风险 | 政治和招标结果仍不确定 |
可比交易用于勾勒未来客户模型;不能证明 Proxima 当前收入。
[CU020, CU021, CU022, CU023, CU024, CU025]合作伙伴 / 场址验证质量最高,留存和有约束力的承购最弱。
[CU002, CU003, CU004, CU005, CU020, CU021]6.4 反向客户观点是:承诺有条件且相隔十年
主要客户风险不是客户流失;它没有可流失的收入基础。风险在于,今天的“客户”是有条件的利益相关方,其经济性只有在重大科学、工程、许可、 融资和电网集成里程碑之后才会兑现。公开来源没有披露 Proxima PPA、费率、合同容量、商业价格、客户数指标、NRR、GRR、合同期限或头部客户集中度。 Neutron Bytes 明确质疑 Proxima 的公关说法止于何处、硬事实始于何处;Energy Solutions 2026 年核聚变综述强调,氚、材料、监管和成本障碍, 使 2030 年代中期前出现有意义的并网供应不太可能。这种怀疑对客户尽调很重要:RWE 和 Bavaria 降低了选址和机构风险,但不能消除技术交付风险, 也不能证明公用事业或数据中心买家会在 2030 年代末为核聚变电力支付有吸引力的价格。在 Proxima 发布具约束力的购电、项目融资条款和并网路径前, 客户结论是“有前景的期权价值,而非商业牵引力”。[CU031, CU032, CU033, CU034, CU035, CU036]
| 指标 | 值 / null | 客群 | 置信度 | 重要性 | 尽调事项 |
|---|---|---|---|---|---|
| NRR / GRR | null | 所有未来客户 | 高 | 没有收入基数,就没有队列留存 | 签约后索取项目开发里程碑转化率 |
| 续约 / 流失 | null | RWE 和未来购电方 | 高 | 合作关系不能按 SaaS 续约打分 | 索取 MoU 终止权和先决条件 |
| 合同期限 | 购电协议为 null;已存在 MoU | 公用事业 / 公共部门合作方 | 中 | 期限决定可融资性和客户质量 | 在 NDA 下审阅 RWE/Bavaria/IPP 协议 |
| 客户满意度 | null | 未来电力买方 | 中 | 尚未交付产品,客户无从满意或失望 | 与 RWE、IPP、Bavaria 和可比购电买方做背调访谈 |
| 重复使用 / 扩张 | 当前不适用 | 电网和工业买方 | 高 | 扩张需要首座电站成功,并继续复制机组 | 索取从 Stellaris 到第 n 座同类电站的路线图 |
传统留存指标刻意保留为 null,因为 Proxima 没有披露客户、收入队列或已交付电力产品。
[CU006, CU031, CU032, CU033, CU034, CU035]| 扩张驱动因素 | 集中度 / 采购风险 | 影响 | 尽调路径 |
|---|---|---|---|
| RWE 作为首个场址 / 公用事业锚点 | 对 Gundremmingen 单一场址和联邦枢纽拨款存在依赖 | 高 | 审阅备选场址管线,以及 RWE 排他性或终止条款 |
| Bavaria 与联邦公共资金 | Alpha 建成前,公共预算和政治优先级可能变化 | 高 | 索取拨款状态、条件和追回触发条款 |
| Google 与超大规模厂商清洁电力需求 | 战略投资兴趣未必转化为 Proxima 购电 | 中 | 询问是否有超大规模厂商签署 LOI、期权或容量预留 |
| 可比聚变 PPA | 可比合同取决于首创技术交付 | 中 | 对标 CFS/Helion 式交易的可融资性、违约惩罚和先决条件 |
| 工业供应链生态 | 首座电站可能依赖专业供应商和许可节奏 | 中 | 审阅采购计划、关键路径和供应商集中度 |
| 2030 年代末商业化时间 | COD 前,需求、市场规则和电价都可能变化 | 高 | 在多种德国电力市场情景下压力测试购电经济性 |
风险项反映未来客户模型;不应解读为当前客户流失或现有收入集中度。
[CU012, CU013, CU014, CU022, CU024, CU025]6.5 图表
07风险
7.1 顶层风险集中在技术证明、资本强度和进度现实性
Proxima Fusion 的风险更适合看成层层叠加的里程碑问题,而不是单一科学风险。作为一家年轻深科技公司,它有少见的优势:2026 年 7 月完成 €411 million 融资,投后估值 €2.4 billion,继承 Max Planck IPP 技术,RWE 参与选址和运营,德国政策环境也希望落地首座聚变电站。这些事实让它相对多数欧洲聚变同行, 融资和可信度风险更低,但核心问题还没解除:Alpha 仍要在 2030 年代初证明仿星器的净能量,同时公司还要扩 HTS 磁体制造、氚与材料方案、许可审批、公私资金和未来购电经济性。 独立负面来源仍把剩余风险压在高位:GAO 称商业聚变仍面临燃烧等离子体、材料、系统工程和监管挑战;MIT Technology Review 质疑聚变成本能否快速下降;Bulletin 则认为商业化炒作一次次跑在有用功率证明前面。[CR001, CR002, CR003, CR011, CR021, CR022]
| 类别 | 风险 | 可能性 | 严重性 | 缓释成熟度 | 剩余敞口 | 投资含义 |
|---|---|---|---|---|---|---|
| 技术 / 科学 | Alpha 未能按目标时间表证明仿星器净能量 | 中 | 严重 | 中低 | 核心物理和一体化电站验证仍在前面 | 估值应卡在 Alpha 物理与功率平衡里程碑之后 |
| 技术 / 工程 | HTS 线圈与 REBCO 退化在电站级辐照、载荷或运行时长下显现 | 中 | 高 | 中 | 瞬时效应风险有所降低,但长期退化仍未验证 | 要求线圈测试数据、中子暴露计划和良率指标 |
| 材料 / 燃料 | 第一壁材料和氚增殖 / 处理未能按时成熟 | 中高 | 高 | 中低 | 燃料基础设施和第一壁材料是全行业瓶颈 | 下调时间表,并要求合作方燃料循环路线图 |
| 执行 / 时间表 | 2030 年代初 Alpha 或 2030 年代末并网显著推迟 | 高 | 高 | 中 | 公私合力很深,但多条工作线都是首台套 | 按里程碑拨付资金,不要现在就把估值打满 |
| 财务 | 数十亿级资本开支会倒逼多轮巨额融资、补助或项目融资 | 高 | 高 | 中 | 新一轮 €411M 规模不小,但仅 Alpha 就被引述约需 €2B | 测算稀释和降价融资敏感性 |
| 竞争 | CFS、Helion 或其他资金更足的同行定义商业化时间表或客户标准 | 中 | 中高 | 中 | 其在欧洲领先,但 CFS 资本基础接近 $3B | 里程碑要对标 CFS / Helion,不只看欧洲同行 |
| 监管 / 法律 | 聚变专项框架推进慢于预期,或限制更严 | 中 | 中高 | 中 | 框架方向友好,但各地尚未全部定稿 | 跟踪德国、欧盟、英国、美国及具体场址反馈 |
| 市场 / 商业 | 首批电站投运时,承购经济性撑不起发电成本 | 中 | 高 | 低 | AI / 数据中心需求存在,但成本和可靠性未知 | 要求可信的 LCOE、PPA 和电网研究 |
| 人才 / 生态 | Max Planck IPP、创始人和专业制造资源依赖会卡住进度 | 中 | 中 | 中 | 生态很强,但稀缺技能和 IP 依赖仍在 | 审查留才、接班和 IP / 控制安排 |
严重性按投资影响排序;可能性是基于公开证据的尽调判断,不是工程概率模型。
[CR001, CR003, CR021, CR022, CR025, CR027]净能量证明、资本需求和时间线执行重叠处,剩余风险最高。
[CR021, CR022, CR025, CR027, CR033, CR046]公开缓释因素之后,技术、财务和执行风险的剩余严重度最高。
1–5 序数严重度评分来自本章风险登记表,不是定量概率模型。
[CR003, CR005, CR011, CR027, CR034, CR047]7.2 技术风险不只是等离子体物理;它是一套硬件、燃料和材料的集成工程
最重要的技术风险在于,Proxima 必须把优化后的 QI-HTS 仿星器概念,推进成一台可运行、可维护、能实现净能量的机器。Wendelstein 7-X 和 Stellaris 提供了科学与设计可信度, 但商业运行的约束比研究记录能验证的更严苛:14 MeV 中子通量下的长寿命结构材料、氚增殖和处理、HTS 磁体耐久性、磁场容差、第一壁维护,以及复杂部件供应。部分风险边际上已经降低: MIT 研究人员发现,先前怀疑的 REBCO 瞬时束流开启效应在其测试中并不重要。但同一来源也说,REBCO 多年或数十年的长期退化仍在研究中;FIA 供应链报告则把燃料基础设施和第一壁材料列为重大未来隐忧。 尽调重点应放在 Alpha 的模型线圈证据、中子 / 材料验证计划、燃料循环设计和制造良率上,而不只是看等离子体头条里程碑。[CR007, CR008, CR013, CR024, CR025, CR026]
| 失效模式 | 可能性 | 严重性 | 缓释成熟度 | 剩余敞口 | 未解决缺口 |
|---|---|---|---|---|---|
| 净能量演示未达目标 | 中 | 严重 | 中低 | Alpha 仍是未来才会给出的验证,不是已经跑成的运行结果 | 需要物理里程碑计划和独立评审 |
| HTS 磁体 / 模型线圈放大不及预期 | 中 | 高 | 中 | 模型线圈和电缆生产已列入下一步 | 需要线圈测试结果、生产良率和淬灭计划 |
| REBCO 长期中子退化 | 中 | 高 | 中 | 一个瞬时效应看起来已被排除 | 需要多年剂量 / 退化验证 |
| 第一壁材料达不到商业耐久性 | 中高 | 高 | 中低 | 行业承认第一壁材料是未来隐患 | 需要材料测试设施访问权和更换经济性 |
| 氚增殖或处理短缺 | 中 | 高 | 中低 | ITER 和监管机构已有燃料循环概念定义 | 需要增殖比、库存、去氚和供应商计划 |
| 电力电子 / 真空室供应受限 | 中 | 中高 | 中 | 供应链风险意识在提升 | 需要供应商承诺和长周期采购图谱 |
| 复杂仿星器制造公差 | 中 | 高 | 中 | 仿真和产业委员会可缓释风险 | 需要可制造性评审和计量证据 |
各行强调一体化电站就绪风险,而不是单一学科的等离子体成果。
[CR003, CR007, CR008, CR021, CR024, CR025]每个里程碑都依赖前序技术、融资和监管证明时,风险会层层叠加。
[CR001, CR003, CR005, CR031, CR039, CR045]7.3 监管比裂变更友好,但选址、废物、氚和 IP 义务仍是现实风险
监管风险有两面。好的一面是,主要司法辖区正有意搭建相称的聚变框架,而不是把聚变照搬裂变监管。英国 EN-8 草案不绑定具体技术和输出,NRC 正沿着聚变机器副产品材料路径推进, 德国联邦政策也明确要为首座聚变电站创造有利创新的条件。这降低了陷入裂变式许可死胡同的概率。但许可并不因此轻松。Foley Hoag 指出,NRC 框架仍留下废物分类、氚报告和州联邦兼容性等实操问题; 英国流程仍要求规划和环境评估;Proxima 自己的 MoU 也把选址、许可、监管流程、项目结构和融资列为明确工作流。法律风险还不止安全法规:Noerr 强调,IP 策略、利用计划、拨款义务和跨境限制, 都是德国 / 欧盟公共资助聚变项目能否成功的关键。[CR012, CR014, CR015, CR016, CR017, CR018]
| 规则 / 案例 / 义务 | 法域 | 状态 | 可能性 | 严重性 | 缓释措施 | 剩余敞口 | 尽调路径 |
|---|---|---|---|---|---|---|---|
| 聚变安全 / 辐射框架 | 德国 | 政策行动计划和资助框架仍在演进 | 中 | 高 | 联邦行动计划和鼓励创新的政策信号 | 商业聚变先例仍然缺位 | 要求德国律师就 StrlSchG/AtomG 适用及巴伐利亚主管机关路径出具备忘录 |
| EN-8 国家政策声明草案 | 英国 | 2026 年咨询框架草案 | 中低 | 中 | 不限定技术、不限定产出的规划方式 | 仍需环境评估和规划解释 | 跟踪最终 EN-8,并与德国选址路径对比 |
| NRC Part 30 / 副产品材料框架 | 美国 | 拟议规则于 2026 年 2 月发布 | 中 | 中高 | 副产品材料路径避开裂变反应堆框架 | 废物、氚、州-联邦兼容性仍未厘清 | 跟踪最终规则及其对全球监管规范的参考意义 |
| 氚报告与废物处置 | 美国 / 国际类比 | 实施细节仍在征询 | 中 | 高 | NRC 和法律分析认可处置路径 | 废物分类缺口可能约束设计 | 要求氚库存、去氚、废物和退役计划 |
| Gundremmingen 选址与许可 | 德国 / 巴伐利亚 | MoU 将工作线分配给各合作方 | 中 | 高 | RWE 的场址和基础设施经验 | 原裂变电站场址仍可能招致审查和地方要求 | 要求许可路线图、主管机关图谱、公众参与计划和并网状态 |
| 公共补助和 IP 利用义务 | 德国 / 欧盟 | 资助规则越来越强调 IP 计划 | 中 | 中高 | Noerr 将 IP 策略和利用计划列为申请资助必须完成的工作 | 联盟权利,或 EEA / 瑞士利用限制,可能拖慢商业化 | 审查补助条件、联盟协议、背景 IP 和使用领域权利 |
仅列举最重要的公开监管和法律风险;不能替代具体法域法律尽调。
[CR012, CR014, CR015, CR016, CR017, CR018]Proxima 的剩余风险取决于合作伙伴、公共机构、稀缺供应商和未来买家。
[CR002, CR013, CR031, CR038, CR039, CR040]7.4 这轮融资买来现金续航,但也抬高了后续资本和商业证明的门槛
2026 年 7 月这轮融资既是优势,也是风险标记。它给 Proxima 带来真实资源和战略背书,但也在 Alpha 尚未证明净能量、任何产品收入尚不可见之前,就把一条商业化前的硬件路线图定价到 €2.4 billion。2026 年 2 月 MoU 称,仅 Alpha 就需要约 €2 billion;从 Alpha 走到 Gundremmingen 和并网 Stellaris,还需要更多公共、私人和项目融资资本。 这轮融资也意味着实质稀释和后续融资压力:€411 million 约等于投后价值的 17%,且还不含后续分批。竞争压力并非纸面假设。CFS 称已累计融资接近 $3 billion,并计划在 2030 年代初借 Dominion 和 Google 推进 ARC 并网发电;其他聚变同行也把投资者预期维持在激进水平。即便 AI 和数据中心电力需求形成支撑,购电价值仍取决于成本、可靠性、进度和并网能力。[CR001, CR005, CR027, CR033, CR034, CR035]
| 依赖项 | 对手方 | 角色 | 集中度 | 失效情景 | 严重性 | 缓释措施 | 剩余敞口 |
|---|---|---|---|---|---|---|---|
| 科学领导力 | Max Planck IPP | 等离子体物理和 W7-X 传承 | 高 | IPP 资源或知识转移瓶颈拖慢 Alpha | 高 | 正式 MoU 和学研拆分创业背景 | 仍依赖稀缺仿星器专业能力 |
| 场址和电站执行 | RWE / Gundremmingen | 原电站场址、运营经验、投资 | 高 | 许可、电网或退役约束拖延 Stellaris | 高 | RWE 投资和基础设施 | 没有商业聚变电站先例 |
| 公共资金 | 巴伐利亚 / 德国联邦项目 / 欧盟 | 补助和共同融资路径 | 高 | 政策优先级转向,或联邦资金慢于计划 | 高 | 行动计划和巴伐利亚承诺 | 路线图仍需要大规模公私资金组合 |
| 战略需求 | Google / AI 数据中心需求 | 长期需求信号和投资方 | 中 | 电价或可靠性让未来买家失望 | 中高 | 稳定清洁电力需求在增长 | 公开信息未披露 Proxima 有具约束力的 PPA |
| 供应链 | 先进部件、第一壁材料、燃料循环供应商 | 硬件和电站投入品 | 中高 | 长周期供应商缺乏可见度或产能 | 中高 | FIA 报告称供应商参与度在改善 | 燃料和第一壁瓶颈仍在 |
| 竞争基准 | CFS / Helion / 其他聚变公司 | 投资人和客户预期标杆 | 中 | 竞争者更早达成并网或承购里程碑 | 中高 | 欧洲领先地位和仿星器差异化 | CFS 资本基础明显更大 |
依赖风险按失效对 Alpha、Stellaris、后续融资或商业可信度的直接拖延程度排序。
[CR002, CR004, CR020, CR031, CR033, CR034]| 情景 | 假设 | 剩余严重性 | 概率信号 | 投资含义 |
|---|---|---|---|---|
| 乐观 | 模型线圈通过验证,Alpha 仍接近 2030 年代初目标,监管尺度保持相称,公共 / 私人资金按计划到位 | 中高 | 需要多个外部里程碑同时对齐 | 保留期权价值;只有价格克制时才考虑按比例跟投 |
| 基准 | Alpha 有进展但推迟,材料 / 氚问题仍未关闭,电站验证前还需要更多资本 | 高 | 与 GAO 和行业瓶颈证据一致 | 跟踪或继续研究;按里程碑融资 |
| 悲观 | 净能量验证大幅推迟,线圈 / 制造证据不及预期,融资依赖补助,竞争对手里程碑领先 | 极高 | 聚变历史和不利来源让这种情景仍然可信 | 回避,或按受压的研究期权重新定价 |
| 监管上行情景 | 德国 / 英国 / 美国框架保持定制化,选址流程配合顺畅 | 中 | 当前政策方向有利 | 监管成为缓释项,但不是核心证明 |
| 成本下行情景 | 聚变经验曲线像学习慢的复杂基础设施,首批电站成本高 | 高 | MIT Technology Review 的成本分析支持这个担忧 | 估值中不要假设低价基荷电力 |
情景按公开证据汇总严重性区间;私人预算和技术测试数据不可得,所以概率只能定性。
[CR011, CR014, CR017, CR022, CR027, CR030]技术延期会传导到融资、伙伴信心、客户经济性和估值。
[CR022, CR027, CR033, CR035, CR038, CR043]7.5 缓释因素可信,但承销必须用里程碑设闸,而且不能心软
正确的投资控制,不是因为聚变很难就排除 Proxima;而是要求每个主要缓释因素都转化为可量化的降险。Max Planck IPP 领导力应产出透明的 Alpha 物理里程碑。RWE 在 Gundremmingen 的角色应拿出具体的选址、并网、许可和项目交付证据。新融资应转化为模型线圈进展、HTS 产能、供应商承诺,以及不依赖模糊未来巨额融资的分阶段预算。 监管上,应尽早拿到主管机构对氚、废物、环境评估和公众沟通的反馈。如果这些里程碑延误,下行情形会非线性放大:Alpha 迟到或超预算,会削弱净能量论点、加重稀释、暴露政策依赖, 并让资金更足的竞争对手定义市场。因此,风险登记支持高风险、继续研究的立场;除非资料室证据能证明进度、预算、技术和监管推进已经超出公开公告。[CR009, CR010, CR040, CR041, CR044, CR045]
| 角色 / 职能 | 依赖或缺口 | 可能性 | 严重性 | 缓释措施 | 尽调路径 |
|---|---|---|---|---|---|
| 创始人 / CEO | 需要把科学愿景转成工业执行和融资纪律 | 中 | 高 | 2026 年 7 月大额融资和公私联盟 | 审查董事会治理、接班安排、里程碑问责 |
| Max Planck IPP 科学领导力 | Alpha 物理依赖 IPP 领导力和知识转移 | 中 | 高 | IPP 根据 MoU 牵头等离子体物理 | 审查 IPP 资源配置、发表路径和冲突管理规则 |
| 工程和制造领导力 | HTS 电缆、磁体、线圈和仿星器制造放大 | 中高 | 高 | 已宣布招聘和产业委员会 | 要求组织架构图、自建还是外购计划、供应商 QA 指标 |
| 财务 / 项目控制 | 数十亿级硬件项目需要预算控制和后续融资 | 中 | 高 | CFO 于 2026 年 6 月任命 | 审查预算基线、采购控制和现金续航模型 |
| 监管 / 公共事务 | 需要本地选址、许可、环境和公众参与能力 | 中 | 中高 | RWE 和公共部门合作方提供经验 | 审查许可负责人、本地利益相关方计划和主管机关矩阵 |
| 商业 / 承购领导力 | 需要把 AI / 数据中心和电力公司兴趣转化为可融资条款 | 中 | 中高 | Google 和 RWE 是战略投资方 | 要求 PPA 策略、目标价格区间、电网研究和买家管线 |
执行风险高:Proxima 要同时搭公司、建制造基地、跑监管路径,还要做首台套硬件。
[CR002, CR005, CR006, CR009, CR010, CR040]| 风险 | 可监测触发项 | 阈值 / 事件 | 行动含义 |
|---|---|---|---|
| Alpha 净能量风险 | 物理和一体化功率里程碑评审 | Alpha 推迟超过 24 个月,或功率平衡证据仍未经审计 | 暂停给予溢价估值,改按研究期权重新定价 |
| HTS / 线圈制造风险 | 模型线圈测试和生产良率 | 模型线圈未达到目标场强 / 公差,或良率仍不具经济性 | 要求技术缓冲、财团支持或估值折扣 |
| 氚 / 材料风险 | 燃料循环和第一壁路线图 | 到下一轮大额融资前,仍没有可信的氚增殖、库存、废物或更换计划 | 将商业电站时间表视为推测 |
| 资本强度 / 稀释 | 预算、现金续航和融资计划 | 技术里程碑前就需要下一笔资金,或只能接受惩罚性条款 | 测算严重稀释和下行情景下的优先权堆栈 |
| 监管 / 选址风险 | 主管机关反馈和公众流程 | 具体场址预申请工作后,德国 / 巴伐利亚许可路径仍未解决 | 延后计入 Stellaris 价值,并要求法律先决条件 |
| 竞争时点风险 | 同行里程碑比较 | Alpha 仍未达到净能量前,CFS / Helion 已拿到并网 / 承购证明 | 下调战略稀缺性溢价 |
| 市场 / 承购风险 | PPA 和 LCOE 证据 | 首座电站成本水平下,买方经济性还没有可信证据 | 经济性获得独立验证前,不要把公用事业级收入写进投资假设 |
否决标准刻意按里程碑设置;公司尚未产生收入,估值主要由技术进展和融资期权价值支撑。
[CR021, CR025, CR027, CR033, CR035, CR036]7.6 图表
08估值
8.1 建议:跟踪资产,不要把当前估值当作基本面支撑
Proxima 是可信的战略资产,但公开估值结论必须对价格敏感。2026 年 7 月披露的融资给市场一个清晰头条:融资 €411 million、投后估值 €2.4 billion, Google 和 RWE 等战略方参与。对一家成立两到三年的 Max Planck 分拆公司来说,这是有分量的背书,也支撑 Alpha、垂直整合磁体和欧洲能源安全议程的真实期权价值。 但它没有给出基本面支撑的价格。Proxima 没有公开产品收入、ARR、EBITDA 或电力销售历史,因此常规软件或工业倍数都用错了类别。因此,基于基本面应视为偏高 / 昂贵; 但如果投资者能拿到可接受的优先权、里程碑融资安排,以及足以扛过后续稀释的持股,这个长期期权并非非理性。IC 应跟踪或继续研究,而不是按头条估值盲买。[CV001, CV002, CV003, CV004, CV005, CV007]
| 建议 | 置信度 | 风险评级 | 估值立场 | 决策含义 |
|---|---|---|---|---|
| 跟踪 / 继续研究 | 中 | 高 | 按基本面看偏高到昂贵;只有配合条款,期权价值才说得通 | 股权结构、优先权、里程碑预算和稀释路径看清前,不要写入买入假设 |
| 只有低于头条估值或结构足够强时才有条件投资 | 低-中 | 高 | 只有 Alpha 概率和后续融资风险明显下降,估值才算合理 | 要求下行保护、按比例跟投权、里程碑分期出资和战略信息权 |
建议把 Proxima 的高战略质量,与收入前聚变公司仍未证明的普通股经济性拆开看。
[CV002, CV007, CV010, CV038, CV039, CV042]| 论点 | 方向 | 什么会改变判断 |
|---|---|---|
| 2026 年 7 月 €411M 融资后,成为欧洲资金最充足的聚变公司 | 投资论点 | 轮次结构弱或优先权过重,会降低头条融资的可投性 |
| Google 和 RWE 参与验证战略相关性,也验证未来电力需求拉动 | 投资论点 | Alpha 预算获批后,战略方不再加深承诺,会削弱信心 |
| Alpha 以及 Bavaria/RWE/IPP 商业电站路径,带来普通创业公司之外的期权价值 | 投资论点 | Alpha 进度大幅滑坡,会把案例推向残值逻辑 |
| 没有收入、利润或商业电力销售,传统倍数无法使用 | 反论点 | 签下可融资的购电协议,或拿到付费工程 / 磁体收入,会改善投资测算 |
| 聚变同行显示,盈亏平衡前资本强度极高、周期很长 | 反论点 | 一条资金已落实、按里程碑推进且稀释有限的 Alpha 路径,会降低折价 |
| 成本和炒作质疑压住投资者应给乐观情景的概率 | 反论点 | 净能量进展获得独立验证,并拿出可信电力经济性,会改变判断 |
反论点不是否认市场,而是价格、时间、稀释和证据质量风险。
[CV001, CV003, CV005, CV008, CV009, CV010]公开证据支持围绕披露估值的宽期权价值区间,而不是精确基本面价值。
作者估算来自 Proxima 披露估值和公开同业估值 / 融资标记;不是管理层指引。
[CV028, CV029, CV030, CV031, CV033]8.2 当前估值:Proxima 已接近全球聚变头部公司组
€2.4 billion 估值让 Proxima 按披露估值和融资额看成为明确的欧洲龙头,但仍低于美国同行中最激进的估值。Helion 是离群点,2026 年投后估值 $15.5 billion, 并宣布对 Microsoft 的电力交付义务。CFS 的估值透明度较低,但它是融资额标杆:2025 年融资 $863 million,到当时累计接近 $3 billion,分析师来源还称其后续资本领先。 TAE 展示了长周期风险:多年累计融资超过 $1.3 billion,较早披露的估值标记远低于 Helion。Tokamak Energy 在欧洲聚变技术上相关,但按披露融资明显更小。 同行给出的教训不是 Proxima 便宜,而是即便收入证据稀疏,聚变市场仍会用风险期权估值奖励可信里程碑。[CV013, CV014, CV015, CV016, CV017, CV018]
| 可比对象 | 指标 | 倍数 / 估值 / 状态 | 相关性 | 局限 |
|---|---|---|---|---|
| Proxima Fusion | 投后估值 | 2026 年 7 月 €2.4B / ~$2.7B | 公司自身估值,也是当前进入锚点 | 尚未产生收入;优先权和持股细节未披露 |
| Helion Energy | 投后估值 | 2026 年 6 月 $465M Series G 轮后估值 $15.5B | 显示投资者能给被视作聚变龙头的公司多高上行估值 | 技术不同,且里程碑与 Microsoft 挂钩 |
| Commonwealth Fusion Systems | 融资额 | $863M Series B2 轮;截至 2025 年累计融资近 $3B | 领先聚变平台的最佳融资额基准 | 官方来源没有清晰公开当前估值 |
| TAE Technologies | 融资额 / 旧估值 | Tracxn 显示累计融资 $1.32B;2022 年轮次投后估值 $1.2B | 显示长周期资本需求和估值重定价风险 | 公司更老,技术路径不同 |
| Tokamak Energy | 融资额 | 最新披露轮次 $125M;按来源范围不同,累计 $226.5M-$335M | 欧洲 / 英国磁约束聚变同行,规模低于 Proxima | 估值未披露;数据提供商对累计融资口径不同 |
| 聚变行业 | 市场融资 | 按来源范围和日期不同,累计 $7.1B-$15B+ | 显示广泛融资热潮支撑期权估值 | FIA、TechCrunch 和市场评论者定义不同 |
可比样本来自公开可检索的估值和融资标记;私募轮条款大多未披露。
[CV002, CV013, CV015, CV016, CV017, CV019]| 日期 | 融资 / 里程碑 | 金额 | 估值含义 | 注意事项 |
|---|---|---|---|---|
| 2023 | 公司成立 / 早期拆分孵化阶段 | 此处不作为估值锚点 | 技术期权仍主要处在研究阶段 | 本章来源集中没有公开商业估值 |
| 2025 | 基于报告规范事实的 Series A 轮背景 | 此前 €130M 轮次 | 从种子期资本转向硬件执行资本 | 已留存来源未披露估值 |
| Feb 2026 | RWE/Bavaria/IPP 商业电站协议 | 战略里程碑,不是定价轮 | 把 Proxima 连接到场址和公用事业路径,提升期权价值 | 协议条款和可融资性未公开 |
| Jul 2026 | €411M 融资 | €411M / ~$468M | €2.4B / ~$2.7B 投后估值;约为累计融资的 3.6x | 优先权堆栈和一级 / 二级交易组合未披露 |
历史表结合了本次报告的规范事实、保留的 2026 年 7 月融资来源和合作来源;缺失估值仍明确列为缺口。
[CV001, CV002, CV004, CV005, CV008, CV009]按披露规模,Proxima 低于 Helion 披露估值,但高于大多数欧洲融资同业。
由于私营聚变公司的估值标记稀少,条形图有意混合披露估值和已筹资本基准。
[CV017, CV019, CV021, CV023, CV024, CV032]从研究机构孵化公司升至 €2.4B 期权价值,靠的是里程碑,不是收入倍数。
时间线使用公开和核心事实中的融资里程碑,并保留 2026 年来源日期。
[CV001, CV002, CV004, CV008, CV009, CV025]8.3 场景方法:期权价值主导,稀释决定普通股价值
估值模型应以场景为核心,而不是以倍数为核心。熊市价值主要来自 IP、团队、拨款关系,以及 Alpha 延误或磁体制造无法放量时剩余的战略可选性。基准价值接近当前估值: 承认战略投资者已经验证 Proxima,同时折价处理缺乏收入、运营指标、股权结构细节和已验证净能量硬件。牛市价值可能大得多,但前提是 Alpha 和公用事业承诺让 Proxima 看起来更像 Helion/CFS 同行组的头部。即便如此,普通股回报仍取决于稀释:若 7 月轮是一级发行,仅这一轮就约占投后 17%,而公司在任何商业电站前大概率还需要几轮大额融资。 按风险投资法在 €2.4 billion 进入的投资者,需要的是非常大的退出,而不只是成功完成下一轮。[CV025, CV026, CV027, CV028, CV029, CV030]
| 情景 | 假设 | 估值 / 回报逻辑 | 关键风险 | 概率信号 |
|---|---|---|---|---|
| 悲观 | Alpha 大幅推迟,磁体放大不及预期,后续资本成本高 | €0.4B-€1.0B 剩余期权价值;当前估值会受损 | 技术延误、稀释、战略方疲劳 | 25% 示意权重 |
| 基准 | Alpha 仍可信,但风险尚未完全排除;战略投资者保持参与 | €1.8B-€3.0B 区间围绕当前估值;回报取决于条款 | 里程碑滑坡和股权结构条款缺失 | 55% 示意权重 |
| 乐观 | Alpha 验证架构,RWE / 公用事业承诺加深,融资仍可获得 | €8B-€15B+ 期权情景,参照领先聚变公司估值 | 仍需要数十亿项目融资和电力成本证明 | 20% 示意权重 |
区间为作者基于公开证据的估算;不是管理层指引,也不是审计估值。
[CV028, CV029, CV030, CV031, CV033, CV034]| 驱动因素 | 悲观影响 | 基准假设 | 乐观影响 |
|---|---|---|---|
| Alpha 进度 | 推迟到 2030 年代初之后,期权价值会大幅下降 | 2030 年代初目标仍可信 | 独立里程碑验证提高龙头结果概率 |
| 磁体 / HTS 制造 | 原型到量产的成本曲线跑不通 | 放大仍有可能,但尚未证明 | 工业化磁体供应成为战略护城河 |
| 未来稀释 | 还要以持平 / 下行条款融入数十亿资本 | 轮次很大,但有按比例跟投,仍可管理 | 战略资本和非稀释性公共资金限制普通股稀释 |
| 电力成本可信度 | 聚变电力成本被证明很高 | 经济性仍未验证 | 可信 LCOE 路径打开公用事业 / 项目融资 |
| 战略承诺 | RWE / Google 兴趣没有加深 | 战略方持续跟踪并提供背书 | 可融资购电 / 场址 / 项目承诺落地 |
| 同行情绪 | 聚变融资热潮的裂缝扩大 | 资本继续流向龙头 | Helion / CFS 式估值把 Proxima 往上拉 |
敏感性为定性分析,因为 Proxima 没有披露收入、EBITDA 或单位经济基础。
[CV008, CV009, CV017, CV018, CV027, CV034]| 方法 | 适用性 | Proxima 的输出 | 为什么重要 |
|---|---|---|---|
| 收入 / EBITDA 倍数 | 不适用 | 没有公开收入、ARR、EBITDA 或产品销售 | 避免给收入前聚变公司制造虚假精度 |
| 融资额基准 | 有用但粗 | Proxima 估值约 ~$2.7B、累计融资约 ~$740M,接近一线同行但低于 Helion | 框定市场为聚变期权出资的意愿 |
| 同行估值比较 | 在估值披露处有用 | Helion 是高端参考;CFS / TAE / Tokamak 更多显示资本强度 | 显示当前估值既非独一无二,也未被基本面证明 |
| 里程碑加权期权价值 | 最适合公开信息的方法 | 给 Alpha、磁体放大、公用事业 / 场址承诺和未来融资通道定价 | 匹配真正创造或毁掉价值的位置 |
| 风险投资法 | 有助于约束回报纪律 | 不计后续稀释,5x-10x 毛回报大约需要 €12B-€24B 退出 | 显示入场价格要求龙头级结果 |
方法比较刻意排除传统倍数,依赖期权、里程碑和风险投资回报逻辑。
[CV025, CV033, CV042, CV043, CV044]概率加权视角由当前估值附近的基准情景存续驱动,同时受稀释风险封顶。
示例性加权贡献使用上述情景权重和中点估算;不含优先权和稀释折扣。
[CV028, CV029, CV030, CV031, CV034]8.4 尽调:在支付头条价格前,要求证明里程碑、条款结构和资本计划
负面情形不是聚变没有价值,而是投资者可能在决定性里程碑出现前就被要求为成功买单。TechCrunch 关于聚变融资热潮裂缝的报道、MIT Technology Review 对成本的提醒, 以及 Bulletin 对商业聚变炒作的怀疑,都指向同一个承销问题:时间线会滑、经济性会不及预期,公开市场也可能过早奖励叙事。Proxima 自己的备案和新闻稿证据确认了实体与头条融资, 但没有确认优先权堆栈、持股拆分、拨款条件、Alpha 资本开支、项目融资计划,或并网后的客户经济性。因此,最低尽调包应包括:到 Alpha 的里程碑预算、磁体生产成本曲线、 已签署的合作伙伴 / 购电条款、公共拨款附带条件,以及股权结构瀑布。没有这些,估值仍是昂贵期权,不是已降险的投资。最终 IC 模型还应同时展示企业价值, 以及至少两轮后续融资后的预期持股。如果所需资本堆栈迫使投资者不断续投、只为维持持股,那么入场时看似可接受的头条估值,即便技术进展真实,也会变得缺乏吸引力。 这也是为什么估值决策应在每个已融资技术里程碑之后重新审视,而不是被当作静态估值。[CV006, CV008, CV009, CV012, CV035, CV036]
| 主题 | 缺失证据 / 触发器 | 为什么重要 | 负责人 / 尽调路径 |
|---|---|---|---|
| 股权结构和优先权 | 完整投后持股、清算优先权堆栈、按比例跟投权、反稀释条款和一级 / 二级交易组合 | 头条投后估值可能夸大普通股吸引力 | 法律顾问和公司财务数据室 |
| Alpha 预算和进度 | 里程碑预算、技术成熟度等级、独立评审和应急计划 | Alpha 滑坡是悲观情景核心触发器 | 技术尽调,配合 IPP/RWE 里程碑复核 |
| 磁体制造 | HTS 线缆和磁体良率、成本曲线、供应商约束和垂直整合资本开支 | 制造失败会在收入出现前摧毁期权价值 | 工程尽调和供应商访谈 |
| 战略承诺 | RWE 场址 / 购电经济性、Google 需求逻辑、公共补助条件 | 战略标识不等于可融资的项目融资 | 合作伙伴尽调和补助文件审查 |
| 下一轮融资 | 下一笔数十亿欧元融资的金额、时间、投资团深度和下行情景计划 | 即使公司活下来,未来稀释也可能抹掉回报 | 董事会 / 投资者访谈和融资模型 |
| 电力经济性 | 预期 capex/MW、LCOE 区间、可用率、维护以及氚 / 燃料假设 | 成本质疑压住终值 | 独立电力市场和核工程审查 |
最终问题是把战略质量判断转成定价投资决定的最低要求。
[CV034, CV036, CV038, CV040, CV041, CV045]8.5 图表
免责声明
本报告是由 AI 辅助研究工作流生成的尽职调查研究成果。 所有财务估算和估值区间均基于公开信息,可能无法反映公司实际财务状况或交易条款。 文中列出来源,并受各章注明的访问日期约束。本报告不构成投资建议。 读者在作出任何投资决定前,应自行开展独立尽职调查。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Proxima Fusion GmbH is headquartered in Munich, Germany, with public company materials also naming Zurich and Oxford locations. | 高 | SO002, SO004, SO010 |
| CO002 | Proxima Fusion GmbH is registered at the District Court of Munich under HRB 283423. | 高 | SO010, SO023 |
| CO003 | Public sources date Proxima’s founding to April 2023, while IPP’s May 2023 release describes a launch at the beginning of 2023. | 高 | SO011, SO022 |
| CO004 | Proxima is the first spin-out company in the history of the Max Planck Institute for Plasma Physics. | 高 | SO004, SO011, SO015 |
| CO005 | The company is developing commercial fusion power plants based on a quasi-isodynamic high-temperature-superconducting stellarator concept. | 高 | SO003, SO004 |
| CO006 | Proxima explicitly builds on Wendelstein 7-X and IPP stellarator research as the technical heritage for its QI-HTS approach. | 高 | SO002, SO003, SO011, SO012 |
| CO007 | Proxima’s public roadmap names the Stellarator Model Coil in 2027, Alpha in the early 2030s, and Stellaris grid deployment in the late 2030s. | 高 | SO001, SO002, SO004 |
| CO008 | The July 2026 financing release says Proxima employs around 200 people across engineering, science, and operations. | 中 | SO004 |
| CO009 | The company remains pre-revenue in public evidence: no reviewed source discloses product revenue, ARR, paying customer count, or audited financials. | 中 | SO004, SO019, SO020, SO023 |
| CO010 | On July 7, 2026 Proxima announced a €411 million ($468 million) financing round at a €2.4 billion ($2.7 billion) valuation. | 高 | SO004, SO020 |
| CO011 | The July 2026 round was led by XTX Ventures and East X Ventures. | 高 | SO004, SO020 |
| CO012 | Google and RWE participated as strategic investors in the July 2026 financing round. | 高 | SO004, SO020, SO021 |
| CO013 | Proxima said the July 2026 round brought total secured capital to more than €650 million ($740 million), including €95 million in public grants. | 中 | SO004 |
| CO014 | RWE separately disclosed a €25 million investment in Proxima and linked it to a cooperation agreement covering the Gundremmingen site. | 中 | SO021 |
| CO015 | Proxima raised a €130 million Series A in June 2025, described by independent outlets as the largest European fusion-startup round at that time. | 高 | SO018, SO019 |
| CO016 | The June 2025 Series A was co-led by Cherry Ventures and Balderton Capital and brought public and private funding to roughly €185 million. | 高 | SO018, SO019 |
| CO017 | Max Planck Innovation and EU-Startups reported a roughly €7 million pre-seed round for Proxima in May 2023. | 高 | SO015, SO016 |
| CO018 | NucNet reported that Proxima raised €20 million for its QI stellarator reactor work in April 2024. | 中 | SO017 |
| CO019 | Public founder lists identify Francesco Sciortino, Lucio Milanese, Jorrit Lion, Jonathan Schilling, and Martin Kubie as Proxima founders. | 中 | SO022 |
| CO020 | Francesco Sciortino is Proxima’s co-founder and CEO and remains the central public strategy and financing spokesperson. | 高 | SO005, SO006, SO011 |
| CO021 | Lucio Milanese is listed in public sources as a co-founder and appears with Sciortino as a managing director in Proxima’s registry-imprint disclosure. | 高 | SO010, SO022 |
| CO022 | Jorrit Lion is a co-founder and Chief Scientist publicly quoted on the Stellaris design. | 中 | SO005, SO022 |
| CO023 | Jonathan Schilling and Martin Kubie are publicly listed as co-founders, but their current operating scope is less fully described in fetched public sources. | 中 | SO022 |
| CO024 | Proxima appointed Sergei Galperin as Chief Financial Officer on June 1, 2026 to lead financial strategy for the SMC, Alpha, and Stellaris roadmap. | 中 | SO006 |
| CO025 | Proxima announced an Industrial Development Board in May 2026 including Luc Rémont, Michael Bolle, Ann Mettler, and Erich Clementi. | 中 | SO007 |
| CO026 | The Alpha Alliance had grown to more than 50 industrial partners by May 2026, according to Proxima’s Industrial Development Board announcement. | 高 | SO004, SO007 |
| CO027 | Alpha is Proxima’s planned net-energy stellarator demonstrator near Munich, targeted for the early 2030s. | 高 | SO004, SO005 |
| CO028 | Proxima’s commercial plant roadmap points to Stellaris later in the 2030s, after Alpha demonstrates net fusion energy in steady state. | 高 | SO004, SO006, SO007 |
| CO029 | RWE says Proxima chose the Gundremmingen site for development of the first commercial magnetic fusion power plant and will initiate approvals with Bavaria’s environment ministry. | 中 | SO021 |
| CO030 | The July 2026 financing followed a cooperation agreement among Bavaria, Proxima, IPP, and RWE around the roadmap and Gundremmingen site. | 高 | SO004, SO021 |
| CO031 | Stellaris is Proxima’s peer-reviewed commercial stellarator power-plant concept, published in Fusion Engineering and Design in 2025. | 高 | SO003, SO005, SO029 |
| CO032 | Proxima says Stellaris integrates electromagnetic, structural, thermal, and neutronics simulations into one coherent design. | 高 | SO003, SO005 |
| CO033 | Proxima’s engineering model relies on simulation, rapid prototyping, computational optimization, and an internal stellarator-design framework called StarFinder. | 中 | SO003, SO008 |
| CO034 | High-temperature superconducting magnets are a central enabling technology in Proxima’s public plan and are the focus of SMC de-risking before Alpha. | 高 | SO003, SO004, SO005 |
| CO035 | Proxima signed a framework agreement with PSI to develop high-temperature superconducting magnet technology for its stellarators. | 中 | SO013 |
| CO036 | IPP’s 2024 BMBF project source shows continuing public-sector collaboration between Proxima and IPP after spin-out. | 中 | SO014 |
| CO037 | XTX Ventures’ own materials emphasize deep AI and machine-learning expertise, making it strategically relevant to Proxima’s AI-enabled engineering story beyond capital alone. | 中 | SO004, SO025 |
| CO038 | Redalpine frames Proxima’s Max Planck access, MIT/Google technical talent, and AI use as part of the investor thesis. | 中 | SO024 |
| CO039 | Fusion power has not yet been deployed commercially, and current nuclear power plants still use fission rather than fusion. | 高 | SO020, SO028 |
| CO040 | Sifted’s Series A coverage cautioned that no fusion machine had yet managed to produce more energy than it consumes, underscoring net-energy risk for startup timelines. | 中 | SO019 |
| CO041 | CNBC quoted Google’s broader fusion view that commercialization is immensely challenging and success is not guaranteed. | 中 | SO020 |
| CO042 | CNBC reported that Proxima is best-funded in Europe by distance, but U.S. fusion startups Commonwealth Fusion Systems and Helion had raised materially more total funding. | 中 | SO020 |
| CO043 | The July 2026 financing release says Proxima ranks among the world’s best-funded fusion companies and establishes it as the best-funded fusion company in Europe. | 高 | SO004, SO020 |
| CO044 | RWE argues the decommissioning Gundremmingen site could create time and cost advantages because of existing nuclear infrastructure and regulatory-approval expertise. | 中 | SO021 |
| CO045 | Public registry and media sources do not disclose Proxima’s full cap table economics, liquidation preferences, board rights, debt facilities, or primary-versus-secondary mix. | 中 | SO004, SO020, SO023 |
| CO046 | North Data lists 37 known active shareholders for Proxima, indicating a broad ownership base but not the economics or control terms needed for underwriting. | 中 | SO023 |
| CO047 | The relevant near-term customers are better described as partners, potential off-takers, and ecosystem sponsors rather than current paying product customers. | 中 | SO004, SO020, SO021 |
| CM001 | Proxima’s defensible market boundary is clean firm power rather than the entire clean-energy or climate-tech market. | 中 | SM001, SM002, SM020 |
| CM002 | Included spend should cover future fusion electricity, power-plant development, grid services, and enabling fusion components. | 中 | SM003, SM004, SM008 |
| CM003 | Generic renewable PPAs, conventional fission output, SMRs, geothermal, storage and gas CCS are substitutes or adjacencies rather than Proxima’s core market. | 中 | SM015, SM020, SM021 |
| CM004 | Proxima’s market is pre-commercial because no cited source shows current product revenue from fusion electricity. | 中 | SM002, SM005, SM012 |
| CM005 | Google’s investment is framed as interest in abundant carbon-free firm energy over the long term, not near-term delivered power. | 高 | SM002, SM005, SM020 |
| CM006 | RWE became a strategic investor after signing an agreement with Proxima for a first stellarator plant at the former Gundremmingen nuclear site. | 高 | SM002, SM005, SM022 |
| CM007 | The relevant status quo for buyers is continued reliance on existing grids, PPAs, fossil backup, fission, and other firming technologies while fusion matures. | 中 | SM015, SM020, SM021, SM022 |
| CM008 | Precedence Research forecasts a global nuclear fusion market of $471.99 billion in 2030 and $843.46 billion by 2040. | 中 | SM017 |
| CM009 | Precedence Research forecasts the U.S. nuclear fusion market at $164.91 billion in 2030 and $239.54 billion by 2040. | 中 | SM017 |
| CM010 | FIA’s 2025 industry coverage reports over $2.5 billion of new fusion investment in the prior 12 months. | 中 | SM007 |
| CM011 | FIA-cited coverage says global fusion investment reached roughly $9.77 billion cumulative by the 2025 report cycle. | 中 | SM007, SM017 |
| CM012 | FIA’s 2026 supply-chain report says surveyed private fusion companies reported $538 million of 2025 supply-chain spending and projected $681 million in 2026. | 中 | SM008 |
| CM013 | FIA’s 2026 supply-chain report says 69% of suppliers still report a lack of long-term visibility of fusion needs. | 中 | SM008 |
| CM014 | Proxima says Alpha will be a net-energy stellarator demonstrator near Munich and is targeted for the early 2030s. | 高 | SM001, SM002, SM003 |
| CM015 | Proxima and CNBC report that the commercial Stellaris power plant is targeted for the late 2030s. | 高 | SM001, SM002, SM005 |
| CM016 | McKinsey expects U.S. data-center power demand to grow from 25 GW in 2024 to more than 80 GW in 2030. | 中 | SM015 |
| CM017 | McKinsey expects U.S. data-center electricity demand to increase by about 400 TWh between 2024 and 2030. | 中 | SM015 |
| CM018 | BloombergNEF says 23.1 GW of data-center IT capacity was under construction globally at the end of September 2025. | 中 | SM018 |
| CM019 | BloombergNEF estimates capex by the 14 largest publicly owned data-center operators will be close to $750 billion in 2026. | 中 | SM018 |
| CM020 | BloombergNEF says large data centers in the UK, Germany, Ireland, Norway and the Netherlands could draw 5.4 GW live IT power demand in 2030, or more than 7 GW in an aggressive scenario. | 中 | SM019 |
| CM021 | Google says its AI infrastructure buildout is accelerating faster than the grid is decarbonizing. | 中 | SM020 |
| CM022 | Google says it signed agreements for more than 12 GW of net-new clean energy in 2025. | 中 | SM020 |
| CM023 | Google says it is advancing nuclear and enhanced geothermal while making long-term bets on breakthrough technologies like fusion. | 高 | SM020, SM021 |
| CM024 | RWE’s homepage frames surging demand from electrification and AI as a 2026 energy-sector trend and emphasizes clean, safe, affordable electricity. | 中 | SM022 |
| CM025 | Germany’s Fusion Action Plan commits more than €2 billion by 2029 to fusion research and pilot projects. | 高 | SM010, SM012 |
| CM026 | BMFTR says Germany’s Fusion Action Plan aims to speed the path to a fusion power plant and plans hubs for magnetic fusion, laser fusion, fuel cycle and materials. | 高 | SM011, SM010 |
| CM027 | Noerr reports that Bavaria announced up to €400 million for fusion projects tied to Proxima, RWE and IPP, including Alpha in Garching. | 高 | SM013, SM002 |
| CM028 | FIA reports that fusion is included as a priority in the European Commission’s 2026 Work Programme. | 高 | SM009, SM011 |
| CM029 | Noerr reports that the Euratom Work Programme 2026–2027 allocates €222 million to accelerate fusion from the lab to the grid. | 中 | SM013 |
| CM030 | BMFTR says Germany contributes to ITER, EUROfusion and the European Commission’s fusion strategy scheduled for 2026. | 中 | SM011 |
| CM031 | Clean Energy Wire reports that fusion remains experimental and commercial use is at least decades away. | 中 | SM012 |
| CM032 | CNBC quotes Google saying commercializing fusion is immensely challenging and success is not guaranteed. | 中 | SM005 |
| CM033 | Clean Energy Wire says Germany cannot count on fusion for climate-neutrality efforts by 2045 because commercial use is at least decades away. | 中 | SM012 |
| CM034 | Competing clean-firm technologies can satisfy buyer demand before fusion, because Google and McKinsey both discuss nuclear, geothermal, storage, gas CCS and clean fuels as alternatives. | 中 | SM015, SM020, SM021 |
| CM035 | Public market sources do not disclose Proxima-specific PPA price, plant MW capacity, or contracted offtake volume. | 中 | SM002, SM005, SM017 |
| CM036 | The fusion TAM forecast should be treated as low-confidence because it begins in 2030 despite current sources saying commercial deployment remains unproven. | 中 | SM005, SM012, SM017 |
| CM037 | Approximate 2030 regional forecast baselines for China and France can be inferred from Precedence’s 2040 values and CAGRs, but this is a derived scenario rather than a source-stated TAM. | 中 | SM017 |
| CM038 | Proxima’s realistic near-term SAM is Europe/Germany clean-firm-power and policy-backed demonstration demand, not the full global fusion forecast. | 中 | SM002, SM010, SM011, SM013 |
| CM039 | Industrial heat is a plausible demand segment for fusion but public Proxima evidence is stronger for grid electricity and continuous power-plant operation. | 中 | SM003, SM008 |
| CM040 | Regulatory, funding, IP, supply-chain and engineering constraints must be resolved before macro clean-firm-power demand converts into repeatable Proxima revenue. | 中 | SM008, SM013, SM012, SM005 |
| CP001 | Proxima's €411 million July 2026 financing ranks as the largest single fusion round in Europe and among the largest globally that year, positioning it ahead of European stellarator and laser-fusion peers on disclosed capital. | 高 | SP026, SP027, SP028 |
| CP002 | Proxima reports more than €650 million of total funding including public grants and is described as the best-funded fusion company in Europe. | 高 | SP026, SP027, SP028 |
| CP003 | Proxima is developing a quasi-isodynamic high-temperature-superconducting stellarator that builds on Wendelstein 7-X plasma-physics heritage. | 高 | SP026, SP029, SP030, SP033 |
| CP004 | Proxima and partners describe Stellaris as the first peer-reviewed commercial stellarator fusion power-plant concept. | 高 | SP032, SP030 |
| CP005 | Proxima publicly says AI and computational tooling changed its design workflow for complex stellarator engineering and coil design. | 中 | SP031, SP030 |
| CP006 | TechCrunch reports that CFS has raised near $3 billion, roughly a third of private capital invested in fusion companies to date. | 高 | SP004, SP005 |
| CP007 | CFS says SPARC is an HTS compact tokamak intended to demonstrate net fusion energy and that ARC is planned as a grid-scale plant in the early 2030s. | 高 | SP005, SP006, SP007 |
| CP008 | TechCrunch reports that TAE had raised about $1.79 billion before its planned merger transaction. | 中 | SP004, SP008 |
| CP009 | Helion announced a $465 million 2026 round at a $15.5 billion valuation, bringing total funds raised to about $1.5 billion. | 高 | SP010, SP009 |
| CP010 | Helion uses a field-reversed-configuration approach with direct electricity recovery and has positioned Microsoft as its first power customer. | 高 | SP009, SP010 |
| CP011 | Tokamak Energy positions itself around spherical tokamaks and high-temperature-superconducting magnet technology, and TechCrunch cites $336 million raised. | 中 | SP004, SP011 |
| CP012 | General Fusion is developing magnetized target fusion through LM26 and TechCrunch reports more than $600 million raised. | 中 | SP004, SP012 |
| CP013 | Zap Energy uses a Z-pinch approach that avoids high-temperature-superconducting magnets or powerful lasers, with TechCrunch citing $327 million raised. | 中 | SP004, SP013 |
| CP014 | First Light Fusion presents FLARE as an inertial fusion architecture and TechCrunch cites $108 million raised. | 中 | SP004, SP014 |
| CP015 | Type One Energy is a direct stellarator peer pursuing utility-linked projects, and TechCrunch reports $269 million raised. | 中 | SP004, SP015, SP024 |
| CP016 | Thea Energy is a direct stellarator peer using arrays of mass-manufacturable magnets and software controls to reduce complex 3D coil burden. | 中 | SP016, SP004 |
| CP017 | Marvel Fusion pursues laser inertial fusion and says it has secured more than $400 million in public and private funding. | 中 | SP017, SP004 |
| CP018 | Focused Energy is commercializing laser fusion from National Ignition Facility lineage and TechCrunch reports a $240 million Series A. | 中 | SP018, SP004 |
| CP019 | Gauss Fusion is a European magnetic-confinement consortium with industrial partners and public BMBF-backed tritium-breeding funding signals. | 中 | SP019 |
| CP020 | Renaissance Fusion is a European stellarator peer using high-temperature superconducting magnets, direct deposition, and liquid-metal shielding. | 中 | SP020, SP021 |
| CP021 | Xcimer is commercializing laser fusion around very large excimer lasers and an inertial-fusion roadmap. | 中 | SP022, SP004 |
| CP022 | Pacific Fusion uses pulsed magnetic inertial fusion and TechCrunch reports a Series A of more than $1 billion paid in milestones. | 中 | SP004, SP023 |
| CP023 | The most direct stellarator peers for Proxima are Type One Energy, Thea Energy, Gauss Fusion, and Renaissance Fusion. | 中 | SP015, SP016, SP019, SP020, SP021, SP024 |
| CP024 | Tokamak competitors CFS and Tokamak Energy are not direct stellarator peers but compete for capital, HTS supply chains, public credibility, and grid customers. | 中 | SP004, SP005, SP006, SP007, SP011 |
| CP025 | First Light, Marvel, Focused Energy, Xcimer, Pacific Fusion, General Fusion, Helion, TAE, and Zap represent alternative non-stellarator routes to the same clean firm-power buyer. | 中 | SP004, SP009, SP012, SP013, SP014, SP017, SP018, SP022, SP023 |
| CP026 | The FIA 2026 supply-chain report says 25 fusion companies completed its survey, with 2025 reported spend of $538 million and projected 2026 spend of $681 million. | 中 | SP003 |
| CP027 | The FIA 2026 report identifies power electronics, vacuum vessels, future fuel infrastructure, and first-wall materials as major supply-chain concerns. | 中 | SP003 |
| CP028 | Proxima is younger and less funded than CFS, Helion, TAE, and Pacific Fusion by public capital or valuation signals. | 中 | SP004, SP010, SP026, SP027, SP028 |
| CP029 | Proxima’s July 2026 round appears to put it ahead of visible European stellarator peers on disclosed funding scale. | 中 | SP026, SP027, SP019, SP020, SP021, SP024 |
| CP030 | Stellarator complexity remains a competitive risk because peers such as Thea and Renaissance explicitly position their designs around simplifying difficult coil engineering. | 中 | SP016, SP020, SP021, SP024, SP025 |
| CP031 | Wendelstein 7-X remains a major stellarator benchmark and supports Proxima’s claimed technical heritage. | 中 | SP030, SP033, SP024 |
| CP032 | The broader private fusion field is crowded, with TechCrunch enumerating many startups above $100 million in funding. | 中 | SP004 |
| CP033 | Most leading private fusion companies publicly target first major grid or pilot milestones in the late 2020s to mid-2030s, making timeline credibility a competitive variable. | 中 | SP006, SP007, SP010, SP015, SP018, SP024, SP026 |
| CP034 | Proxima’s RWE and Google participation gives it strategic site, utility, and potential power-demand relevance rather than only financial sponsorship. | 中 | SP026, SP027, SP028 |
| CP035 | CFS and Helion have stronger public offtake or customer signals through Google ARC output and Microsoft power plans. | 中 | SP007, SP010 |
| CP036 | No retained public source provides comparable reactor pricing, realized power purchase prices, or product revenue across Proxima and its private fusion competitors. | 中 | SP004, SP010, SP026, SP027 |
| CP037 | Fusion supply-chain constraints can slow Proxima even if its plasma physics works because custom magnets, vacuum systems, power components, first-wall materials, and fuel-cycle systems remain scarce. | 中 | SP003, SP026, SP030 |
| CP038 | Stellarators offer a steady-state operating narrative but historically require complex magnetic fields and precision coil systems. | 中 | SP016, SP021, SP024, SP029 |
| CP039 | Proxima says the 2026 financing will fund its Stellarator Model Coil, HTS cable and magnet production, and engineering and manufacturing systems. | 高 | SP026, SP028 |
| CP040 | Helion, Zap, General Fusion, Pacific, and laser-inertial peers avoid Proxima’s exact stellarator coil problem but introduce different stability, compression, repetition-rate, or target-manufacturing risks. | 中 | SP009, SP012, SP013, SP014, SP017, SP018, SP022, SP023, SP025 |
| CP041 | Type One is the most direct U.S. stellarator commercialization comparison because public sources tie it to TVA-linked plant concepts and IEEE compares it directly with Proxima. | 中 | SP015, SP024, SP004 |
| CP042 | Thea and Renaissance challenge Proxima’s coil-design moat by shifting complexity respectively into software-controlled modular magnets and directly deposited HTS coil surfaces. | 中 | SP016, SP020, SP021 |
| CI001 | Proxima announced a €411 million ($468 million) financing round on July 7, 2026, at a €2.4 billion ($2.7 billion) valuation. | 高 | SI001, SI002, SI003, SI005, SI007 |
| CI002 | The July 2026 financing was led by XTX Ventures and East X Ventures, with RWE and Google as strategic investors. | 高 | SI001, SI003, SI004, SI005, SI006, SI007 |
| CI003 | Proxima disclosed more than €650 million of total funding in less than three years, including €95 million in public grants. | 高 | SI001, SI007 |
| CI004 | Public and quasi-public backers in Proxima's funding stack include KfW Capital, SPRIND, DTCF, Bayern Kapital, HTGF, CDP Venture Capital, and the EIC Fund. | 中 | SI001, SI007, SI013, SI021, SI022 |
| CI005 | Proxima says the July 2026 financing provides backing to build Alpha, its net-energy stellarator demonstrator near Munich. | 高 | SI001, SI005, SI007 |
| CI006 | RWE disclosed a €25 million investment in Proxima's July 2026 funding round and described an intent to collaborate toward a commercial magnetic fusion power plant at Gundremmingen. | 高 | SI004, SI001 |
| CI007 | Proxima's June 2025 Series A was €130 million and was co-led by Cherry Ventures and Balderton Capital. | 高 | SI011, SI012, SI013 |
| CI008 | The June 2025 Series A brought Proxima's public and private funding to more than €185 million. | 高 | SI011, SI012, SI013 |
| CI009 | Proxima raised a €20 million seed round in April 2024 led by redalpine with participation from Bayern Kapital, DTCF, the Max Planck Foundation, Plural, UVC Partners, HTGF, Wilbe, and TOMORROW. | 高 | SI014, SI015, SI016 |
| CI010 | Proxima's first disclosed pre-seed financing was approximately €7 million to €7.5 million in 2023. | 中 | SI017, SI018, SI014 |
| CI011 | Proxima announced a €15 million Series A extension in September 2025 that brought total funding to €200 million. | 高 | SI021, SI022 |
| CI012 | The September 2025 extension included CDP Venture Capital, the EIC Fund, and Brevan Howard Macro Venture Fund, and Proxima said the EICF investment followed a prior €2.5 million EIC grant. | 高 | SI021, SI022 |
| CI013 | Public registry pages identify Proxima Fusion GmbH under Amtsgericht München HRB 283423, and Northdata lists 37 known active shareholders. | 高 | SI019, SI020 |
| CI014 | Reviewed public sources do not show current product revenue, grid-electricity revenue, ARR, or commercial plant revenue for Proxima. | 中 | SI001, SI004, SI013, SI021 |
| CI015 | Because Proxima has no public product revenue, gross margin and revenue-recognition metrics should be treated as null rather than estimated from funding or grants. | 中 | SI001, SI013, SI021 |
| CI016 | Proxima's current economic model is milestone financing for engineering, manufacturing, and demonstrator work before any future power sales. | 中 | SI001, SI005, SI007, SI013 |
| CI017 | RWE's disclosed relationship is an investment and site/industrial collaboration, not a disclosed revenue contract or PPA for current Proxima output. | 中 | SI004, SI001 |
| CI018 | Reviewed public sources do not disclose Proxima's cash on hand, monthly burn, committed capex schedule, or runway. | 中 | SI001, SI005, SI007, SI021 |
| CI019 | Reviewed public sources do not disclose Proxima's debt, leases, guarantees, or project-finance obligations. | 中 | SI001, SI004, SI005, SI007, SI021 |
| CI020 | Independent reporting described Alpha as part of a roughly €2 billion stellarator test-facility plan near Munich. | 中 | SI023, SI001, SI007 |
| CI021 | The same reporting said Bavaria and Proxima were each responsible for 20% of the Alpha facility plan, with a remaining €1.2 billion federal funding tranche expected. | 中 | SI023 |
| CI022 | TechFundingNews reported that without Berlin's expected €1.2 billion contribution, the Alpha test plant could not move forward. | 中 | SI023 |
| CI023 | The July 2026 financing is likely to increase spend because Proxima plans to complete the SMC, expand HTS cable and magnet production, and hire across engineering, manufacturing, and operations. | 中 | SI001, SI007, SI013, SI021 |
| CI024 | A €411 million cash pool would fund about 41 months at €10 million monthly burn, 21 months at €20 million, or 14 months at €30 million before considering restricted cash or grants. | 低 | SI001, SI007 |
| CI025 | Proxima's public milestone path includes SMC hardware in 2027 and Alpha operation around 2031 or the early 2030s. | 中 | SI013, SI021, SI001, SI007, SI023 |
| CI026 | Proxima expects venture capital to fund it to about 2031 and then expects other forms of capital to become necessary. | 中 | SI011 |
| CI027 | The public evidence does not yet show the project-finance package needed for a first-of-a-kind commercial plant after Alpha. | 中 | SI001, SI004, SI011, SI023 |
| CI028 | Public grants and public investors reduce near-term private dilution but create policy and disbursement risk for Alpha. | 中 | SI001, SI007, SI021, SI022, SI023 |
| CI029 | MIT Technology Review reported in April 2026 that fusion may not become cheap quickly because large, complex, customized plants may have slower experience-rate cost declines than solar or batteries. | 中 | SI024 |
| CI030 | MIT Technology Review wrote that fusion power plants do not yet exist and that no private fusion company has a working reactor producing electricity. | 中 | SI025 |
| CI031 | The Bulletin of the Atomic Scientists described commercial fusion as repeatedly harder and longer than expected and warned about hype around overly rosy projections. | 中 | SI026, SI027 |
| CI032 | The Bulletin quoted Dennis Whyte saying economic commercialized fusion is not a cakewalk and is not assured. | 中 | SI027 |
| CI033 | Commonwealth Fusion Systems disclosed that it had raised close to $3 billion after an $863 million Series B2 round. | 高 | SI029, SI028 |
| CI034 | TechCrunch reported in June 2026 that Helion had raised about $1.5 billion after a $465 million Series G. | 中 | SI028 |
| CI035 | TechCrunch's 2026 peer survey reported large funding stacks for TAE, General Fusion, Zap Energy, Type One Energy, and other private fusion companies. | 中 | SI028 |
| CI036 | Proxima is the best-funded fusion company in Europe but remains materially smaller than global funding leaders such as Commonwealth Fusion Systems. | 中 | SI001, SI007, SI028, SI029 |
| CI037 | Proxima's decade-plus commercialization path creates continued dilution risk because equity rounds and public funding are still needed before product revenue exists. | 中 | SI011, SI023, SI024, SI025 |
| CI038 | Northdata's shareholder count and the multi-investor July 2026 syndicate make cap-table rights, liquidation preferences, and strategic investor covenants material diligence asks. | 中 | SI019, SI001, SI007 |
| CI039 | Proxima employed around 200 people across engineering, science, and operations at the time of the July 2026 financing. | 中 | SI001, SI005 |
| CI040 | Proxima's funding base includes multiple state-backed or publicly backed European capital providers rather than only private venture funds. | 中 | SI001, SI007, SI013, SI014, SI021, SI022 |
| CI041 | Google's participation supports the strategic power-demand thesis but does not disclose any current Proxima customer revenue or power-purchase price. | 中 | SI001, SI003, SI008 |
| CI042 | RWE and future utility/site relationships should be treated as future customer or partner optionality until binding offtake economics are disclosed. | 中 | SI004, SI001 |
| CI043 | The public financial verdict is strong capital access but weak revenue-quality underwriting because revenue, ARR, margin, cash, burn, runway, and plant-level economics remain undisclosed. | 中 | SI001, SI007, SI024, SI025, SI027 |
| CE001 | Proxima Fusion is developing quasi-isodynamic stellarators using high-temperature superconducting magnets. | 高 | SE001, SE002 |
| CE002 | Proxima's official technology page says its cloud-based StarFinder framework rapidly iterates on QI stellarator designs. | 中 | SE001 |
| CE003 | Proxima says Stellaris integrates electromagnetic, structural, thermal, and neutronics simulations into a coherent QI-HTS plant concept. | 中 | SE001, SE002, SE019 |
| CE004 | QI stellarators avoid a large toroidal plasma current, which underpins Proxima's claim of eliminating current-driven instabilities and disruptions. | 中 | SE001, SE027, SE028 |
| CE005 | DOE describes stellarators as having greater design flexibility and simpler plasma-control aspects than tokamaks, at the cost of more complex magnetic-field coils. | 高 | SE027, SE028 |
| CE006 | Tokamaks use plasma current and magnetic coils to generate confinement fields, while stellarators rely on external coils for twisted magnetic fields. | 高 | SE026, SE027, SE028 |
| CE007 | High-temperature superconductors can operate at higher temperatures and magnetic field strengths than conventional superconductors, enabling a smaller stellarator design space. | 中 | SE001, SE002, SE024 |
| CE008 | Proxima and partners announced Stellaris as a peer-reviewed commercial stellarator power-plant concept published in Fusion Engineering and Design. | 高 | SE002, SE011, SE012, SE019 |
| CE009 | The Stellaris concept includes a magnetic-field design, full-power support structures, HTS integration, heat management, and a complex-geometry neutron blanket concept. | 中 | SE002, SE018, SE019 |
| CE010 | Proxima targets a Stellarator Model Coil demonstration in 2027 to de-risk HTS technology for stellarators. | 中 | SE002, SE018, SE019 |
| CE011 | Proxima, Bavaria, RWE, and IPP signed a 2026 MoU that places Alpha near IPP in Garching and Stellaris at Gundremmingen. | 高 | SE004, SE021, SE022, SE023 |
| CE012 | Alpha is described as a planned net-energy stellarator demonstrator rather than a commercial power plant. | 高 | SE004, SE021, SE022 |
| CE013 | IPP's Wendelstein 7-X is the world's largest stellarator and uses 50 non-planar superconducting magnet coils to test optimized magnetic confinement. | 中 | SE013 |
| CE014 | IPP reports that W7-X achieved a world record for triple product in long plasma discharges during its OP 2.3 campaign. | 高 | SE014, SE015 |
| CE015 | The 2026 APS W7-X paper reports a triple product of (1.10 ± 0.15) × 10^20 m^-3 keV s held stable for 1.9 seconds. | 高 | SE015, SE014 |
| CE016 | The W7-X performance paper says stellarators offer inherent steady-state capability and lack disruptions, but historically lagged tokamaks in triple product. | 中 | SE015 |
| CE017 | Fusion Future reports that Alpha will scale confinement toward heat loads relevant for a power plant, unlike W7-X's research instrumentation. | 中 | SE021 |
| CE018 | RWE's role in the MoU is to contribute large power-plant construction experience and industrial-network support. | 中 | SE004, SE022, SE023 |
| CE019 | Proxima's GitHub organization exposes public repositories for VMEC++, ray tracing, ConStellaration, coilstellaration, and open stellarator models. | 中 | SE005 |
| CE020 | The ConStellaration repository provides code for analyzing and evaluating stellarator plasma boundaries and points to the Hugging Face dataset. | 中 | SE006, SE007 |
| CE021 | The Hugging Face ConStellaration page listed 182k rows with boundary and metric fields during this run. | 中 | SE007 |
| CE022 | The Proxima-Hugging Face challenge defines geometric, simpler-to-build QI, and multi-objective MHD-stable QI stellarator optimization problems. | 中 | SE003, SE009, SE010 |
| CE023 | The ConStellaration arXiv paper says stellarator design is a high-dimensional constrained optimization problem requiring expensive physics simulations and domain expertise. | 中 | SE009, SE010 |
| CE024 | ConStellaration provides reference code, evaluation scripts, and baselines to lower the barrier for optimization and machine-learning researchers. | 中 | SE003, SE006, SE009 |
| CE025 | Proxima's open-source and dataset surfaces verify a real practitioner signal, but they validate design tooling rather than reactor operation. | 中 | SE005, SE006, SE007, SE009 |
| CE026 | VMEC++ and ConStellaration create public evidence that Proxima invests in numerical optimization infrastructure. | 中 | SE005, SE006, SE009 |
| CE027 | The ConStellaration benchmark explicitly includes coil-simplicity and compactness tradeoffs through its simple-to-build QI problem. | 中 | SE003, SE009 |
| CE028 | A reasonable Proxima design workflow runs from QI boundary search to MHD simulation, coil-simplicity optimization, subsystem prototyping, Alpha validation, and Stellaris translation. | 中 | SE001, SE002, SE003, SE006, SE009 |
| CE029 | Stellarators can be designed for continuous stable operation, while tokamaks have historically led fusion science because their geometry is simpler and performance is better studied. | 中 | SE001, SE026, SE027, SE028 |
| CE030 | Proxima's technology thesis is differentiated by combining QI stellarator physics with HTS magnets and computational optimization. | 中 | SE001, SE002, SE003, SE009 |
| CE031 | HTS magnets are an enabling technology shared with compact tokamak efforts, so they are not a Proxima-only moat. | 中 | SE001, SE024, SE026 |
| CE032 | Proxima has not publicly demonstrated net energy gain because Alpha is still a planned demonstrator. | 高 | SE002, SE004, SE021, SE022 |
| CE033 | MIT News reports that a specific instantaneous REBCO critical-current suppression concern under irradiation was experimentally retired, while long-term degradation remains under investigation. | 中 | SE024 |
| CE034 | DOE says stellarator coil manufacturing is challenging because large-bore wire coils require millimeter precision. | 高 | SE027, SE028 |
| CE035 | Clean Energy Platform's 2026 analysis says civilian tritium stockpiles remain only 20 to 30 kilograms and a 1 GW reactor could require about 55 kilograms per year. | 中 | SE025 |
| CE036 | Stellaris includes a neutron blanket concept adapted to complex stellarator geometry, but public evidence does not prove an operating closed tritium breeding cycle. | 中 | SE002, SE018, SE025 |
| CE037 | Fusion Future notes Alpha's schedule depends on unresolved federal funding decisions, fusion regulatory frameworks, and HTS engineering challenges. | 中 | SE021 |
| CE038 | The Bavaria-RWE-IPP MoU de-risks partner alignment and site sequencing but does not itself prove funded construction, permitting, or reactor performance. | 中 | SE004, SE021, SE022, SE023 |
| CE039 | Public Proxima materials do not disclose an independently audited TRL, so readiness must be inferred from concept, code, partner, and milestone evidence. | 中 | SE001, SE002, SE004, SE005, SE006 |
| CE040 | The largest open technical unknowns are net energy, HTS magnet scaling, 3D coil manufacturing, neutron damage, tritium breeding, materials lifetime, and regulation. | 中 | SE002, SE021, SE024, SE025, SE027, SE028 |
| CU001 | Proxima signed a February 2026 agreement with the Free State of Bavaria, RWE, and Max Planck IPP to pursue the world’s first commercial stellarator fusion power plant in Europe. | 高 | SU001, SU004, SU007 |
| CU002 | The agreement sets a two-step roadmap: Alpha near Garching first, followed by Stellaris at the former Gundremmingen nuclear plant site. | 高 | SU001, SU004, SU007 |
| CU003 | RWE invested €25 million in Proxima Fusion’s July 2026 funding round. | 高 | SU002, SU003, SU011, SU012 |
| CU004 | RWE owns the former Gundremmingen nuclear site and is positioned to provide infrastructure, operational expertise, regulatory approvals know-how, and industrial networks. | 中 | SU001, SU002, SU007 |
| CU005 | RWE said it would continue decommissioning at Gundremmingen to make existing nuclear infrastructure available as early as possible if the site receives the magnetic-fusion-hub award. | 中 | SU002 |
| CU006 | Reviewed public sources do not disclose any paying Proxima customers, product revenue, delivered power product, or commercial customer count. | 中 | SU001, SU002, SU003, SU008, SU011 |
| CU007 | Proxima’s current customer-proof file is best understood as future offtakers, site partners, technical partners, public funders, and strategic energy-demand signals rather than current buyers. | 中 | SU001, SU002, SU004, SU005, SU008 |
| CU008 | RWE is the closest Proxima-specific customer-proof signal because it combines utility credibility, site control, operating expertise, and capital commitment. | 中 | SU001, SU002, SU007, SU012 |
| CU009 | Future grid, industrial, utility, and data-center electricity buyers remain uncontracted in public sources. | 中 | SU001, SU002, SU003, SU008, SU011 |
| CU010 | Proxima says Alpha will be built near IPP in Garching and, when operational in the 2030s, is intended to demonstrate net energy gain in a stellarator. | 中 | SU001, SU004, SU007 |
| CU011 | Stellaris is planned for the Gundremmingen site after Alpha, with a target of putting a commercial fusion power plant on the grid in the late 2030s. | 中 | SU001, SU002, SU007, SU023 |
| CU012 | The partners are seeking federal funding under Germany’s High-Tech Agenda and magnetic-fusion-hub process. | 中 | SU001, SU002, SU005, SU006, SU022 |
| CU013 | Bavaria indicated a potential state co-financing contribution of 20% of project costs, subject to federal funding. | 中 | SU001, SU005 |
| CU014 | Bavaria’s government framed the fusion project as relevant to electricity demand from AI, data centers, electromobility, energy security, jobs, and high-tech leadership. | 中 | SU005 |
| CU015 | Max Planck IPP is named as scientific lead for Alpha’s plasma physics work under the MoU. | 中 | SU001, SU004 |
| CU016 | IPP and Proxima’s spin-out relationship give Proxima access to W7-X stellarator heritage and plasma-physics credibility, but IPP is not a paying customer. | 中 | SU004, SU013, SU014 |
| CU017 | Germany’s fusion-policy push and federal roadmap improve the public-sector demand signal but do not by themselves create customer revenue. | 中 | SU006, SU022, SU025 |
| CU018 | RWE’s willingness to participate financially was first signaled in the February MoU and then hardened by the July 2026 €25 million investment. | 中 | SU001, SU002 |
| CU019 | Proxima’s adoption trajectory is milestone-based, not account-based, because meaningful customer adoption depends on Alpha, permits, project finance, interconnection, and Stellaris construction. | 中 | SU001, SU002, SU007, SU023, SU024 |
| CU020 | Google participated as a strategic investor in Proxima’s July 2026 financing round. | 中 | SU003, SU008, SU009, SU010, SU011 |
| CU021 | Google’s clean-energy and data-center materials support the inference that hyperscale electricity demand is a relevant future buyer segment for fusion power. | 中 | SU020, SU021, SU008 |
| CU022 | Google and Commonwealth Fusion Systems announced a strategic partnership and 200 MW power purchase agreement for CFS’s planned ARC fusion power plant. | 中 | SU016, SU017, SU018 |
| CU023 | The Google-CFS PPA is a useful comparable for Proxima because it shows a hyperscaler can contract for future fusion electricity before commercial delivery. | 中 | SU016, SU017, SU018, SU021 |
| CU024 | Helion announced a first-of-a-kind fusion power purchase agreement with Microsoft targeting electricity from its first plant. | 中 | SU019 |
| CU025 | The Helion-Microsoft agreement is a relevant comparable because it shows cloud buyers are willing to make early fusion commitments despite delivery risk. | 中 | SU019, SU024 |
| CU026 | No reviewed source discloses a Proxima-specific power purchase agreement with Google, Microsoft, RWE, or any other named electricity buyer. | 中 | SU001, SU002, SU003, SU008, SU011, SU016, SU019 |
| CU027 | Fusion PPA comparables validate the likely future customer model but do not prove Proxima’s own offtake, price, capacity, or delivery date. | 中 | SU016, SU017, SU019, SU023, SU024 |
| CU028 | Google is currently a Proxima investor and demand signal rather than a disclosed Proxima electricity customer. | 中 | SU003, SU008, SU011, SU016, SU017 |
| CU029 | A bankable Proxima customer case would require definitive offtake, capacity, tariff, COD, penalties, and conditions-precedent disclosures that are not public today. | 中 | SU001, SU002, SU016, SU019, SU023 |
| CU030 | RWE, Bavaria, IPP, Google, and comparable PPA buyers together make future demand credible but do not convert Proxima into a commercial power supplier yet. | 中 | SU001, SU002, SU004, SU005, SU008, SU016, SU019 |
| CU031 | Public sources do not disclose NRR, GRR, renewal rates, churn, customer satisfaction, or contract length for Proxima because there is no operating customer base. | 中 | SU001, SU002, SU003, SU008, SU011 |
| CU032 | Classical retention analysis is not applicable to Proxima today; diligence should instead test MoU durability, project milestones, and future offtake conversion. | 中 | SU001, SU002, SU023, SU024 |
| CU033 | Customer concentration risk is currently project concentration risk because RWE and Gundremmingen anchor the first public commercial site pathway. | 中 | SU001, SU002, SU007 |
| CU034 | Procurement and customer-conversion friction may arise from public funding, federal hub selection, permitting, decommissioning reuse, project finance, and grid-connection dependencies. | 中 | SU001, SU002, SU005, SU006, SU022, SU025 |
| CU035 | Neutron Bytes questioned where Proxima’s claims to build and operate a German grid fusion plant by the mid-2030s end and hard facts begin. | 中 | SU023 |
| CU036 | Energy Solutions’ 2026 review identified tritium breeding, materials science, regulatory paths, and high costs as hurdles between fusion milestones and commercial power. | 中 | SU024 |
| CU037 | Outside analysis placed Proxima’s integrated power-plant effort around a TRL 3-6 transition and said Alpha must validate the basis for a commercial plant. | 中 | SU023 |
| CU038 | Neutron Bytes said Proxima must still prove Alpha and raise major funding to cross from demonstration to commercial plant construction. | 中 | SU023 |
| CU039 | The customer value proposition is speculative until Proxima proves it can deliver reliable grid-connected fusion power at acceptable cost and schedule. | 中 | SU023, SU024, SU025, SU026 |
| CU040 | The chapter verdict is that Proxima has strong institutional future-customer proof but no commercial customer traction as of 2026-07-11. | 中 | SU001, SU002, SU005, SU008, SU016, SU019, SU023, SU024 |
| CR001 | Proxima announced a €411 million financing round at a €2.4 billion post-money valuation, taking total secured funding to more than €650 million including public grants. | 高 | SR001, SR007 |
| CR002 | The July 2026 round included strategic participation from RWE and Google, linking the financing story to future power-plant sites and long-term firm clean-energy demand. | 高 | SR001, SR006 |
| CR003 | Alpha is Proxima’s planned net-energy stellarator demonstrator near Munich and is targeted for the early 2030s. | 高 | SR001, SR002, SR030 |
| CR004 | The planned commercial Stellaris plant is tied to the former Gundremmingen nuclear fission site in Bavaria. | 高 | SR002, SR006, SR030, SR031 |
| CR005 | The February 2026 MoU states that Alpha alone requires about €2 billion, while Proxima and Bavaria each indicated roughly 20% financing shares subject to further funding. | 中 | SR002 |
| CR006 | Proxima’s July 2026 release says the company employs around 200 people across Munich, Zurich, and Oxford while hiring across engineering, manufacturing, and operations. | 中 | SR001 |
| CR007 | Proxima’s technology thesis depends on quasi-isodynamic stellarators using HTS magnets and on scientific inheritance from Wendelstein 7-X. | 中 | SR003, SR007 |
| CR008 | Proxima says its Stellaris concept integrates electromagnetic, structural, thermal, and neutronics simulations into a coherent stellarator design. | 中 | SR003 |
| CR009 | The Industrial Development Board is a mitigation signal for industrialization, but it also implies that Proxima is moving from research into execution disciplines it has not yet proven publicly. | 中 | SR004, SR001 |
| CR010 | The June 2026 CFO appointment is a useful finance-control signal for a company entering a capital-intensive hardware build phase. | 中 | SR005, SR001 |
| CR011 | The German federal government explicitly says fusion is not a short-term solution and that significant technological challenges must still be overcome before a first fusion power plant. | 中 | SR008 |
| CR012 | Germany’s action plan aims to create innovation-friendly conditions for a first fusion power plant in Germany, which is a policy tailwind rather than a completed regulatory approval. | 高 | SR008, SR009 |
| CR013 | BMFTR’s planned fusion hubs include magnetic fusion, laser fusion, and fuel-cycle/materials work, implying that Germany treats fuel and materials as explicit unresolved development lanes. | 高 | SR009, SR025 |
| CR014 | The UK draft EN-8 framework is technology- and output-agnostic and says no local population-density restrictions will apply to fusion energy infrastructure because of low radiological risk. | 高 | SR013, SR026 |
| CR015 | The NRC says the ADVANCE Act amended the Atomic Energy Act definition of byproduct material to include radioactive material produced by fusion machines. | 高 | SR010, SR011 |
| CR016 | The NRC published a proposed fusion-machines rule on February 26, 2026 and opened a comment period through May 27, 2026. | 高 | SR011, SR012, SR027 |
| CR017 | Foley Hoag’s legal analysis says the NRC proposal would use a Part 30 byproduct-material pathway rather than fission-reactor construction and operating requirements. | 高 | SR027, SR010 |
| CR018 | The same legal analysis warns that fusion waste classification, tritium reporting, and state-federal compatibility remain practical implementation issues. | 中 | SR027 |
| CR019 | Noerr says IP arrangements, protection strategies, exploitation plans, and cross-border restrictions are now central to German fusion funding applications and commercialization structures. | 中 | SR025 |
| CR020 | Noerr says the Euratom Work Programme 2026–2027 allocates €222 million to accelerate fusion from the lab to the grid. | 中 | SR025 |
| CR021 | GAO identifies burning-plasma physics, materials that withstand fusion conditions for decades, and unresolved systems engineering as core obstacles to commercial fusion. | 高 | SR014, SR028 |
| CR022 | GAO reports that stakeholder projections for commercial fusion range from 10 years to several decades. | 高 | SR014, SR028 |
| CR023 | GAO says regulatory uncertainty and public-private misalignment could slow fusion development even as regulation must still protect safety. | 中 | SR014 |
| CR024 | World Nuclear Association describes the fundamental fusion challenge as producing more heat from plasma than the energy injected into it. | 中 | SR016 |
| CR025 | World Nuclear Association says present global tritium inventory is around 20 kilograms and that tritium breeding inside future fusion systems is important for large-scale power. | 高 | SR016, SR020 |
| CR026 | World Nuclear Association says D-T fusion creates 14 MeV neutrons and significant structural-materials challenges. | 中 | SR016 |
| CR027 | MIT Technology Review reports a 2026 study estimating fusion experience rates of only 2% to 8%, implying slow cost declines versus solar, batteries, and wind. | 中 | SR017 |
| CR028 | The Bulletin of the Atomic Scientists argues that controlled fusion has not yet demonstrated useful work and remains far from commercial application. | 中 | SR018 |
| CR029 | The Bulletin critique says NIF ignition publicity can overstate power relevance when total laser-system energy input is considered. | 中 | SR018, SR033 |
| CR030 | MIT News reports that one instantaneous REBCO neutron-irradiation concern was retired, but longer-term REBCO degradation over years or decades remains under investigation. | 中 | SR019 |
| CR031 | The FIA 2026 supply-chain report says immediate concerns include advanced components such as power electronics and vacuum vessels, while future concerns include fuel infrastructure and first-wall materials. | 中 | SR024 |
| CR032 | The FIA report says 48% of surveyed fusion companies view fuel supply as a major future concern and 54% plan to work with external suppliers on fuel-cycle technologies. | 中 | SR024 |
| CR033 | CFS says it has raised close to $3 billion and is pursuing ARC grid power in the early 2030s with Dominion Energy and Google, making it a better-funded benchmark for Proxima. | 中 | SR021, SR001 |
| CR034 | Proxima is Europe’s best-funded fusion company, but CFS remains materially better funded in absolute private capital. | 中 | SR001, SR021 |
| CR035 | The €411 million primary round represents roughly 17% of Proxima’s €2.4 billion post-money valuation, before considering future capital needs. | 中 | SR001 |
| CR036 | No public source reviewed discloses Proxima product revenue, ARR, customer revenue, gross margin, cash burn, or runway. | 中 | SR001, SR002, SR005, SR007 |
| CR037 | Proxima’s public roadmap is pre-commercial: current capital is directed to Alpha, HTS cable and magnet production, manufacturing systems, and later Stellaris rather than current product sales. | 中 | SR001, SR002, SR003 |
| CR038 | Public and semi-public funding dependence is material because the roadmap references Bavarian co-financing, federal funding, public grants, and European fusion programs. | 中 | SR001, SR002, SR008, SR009, SR025 |
| CR039 | Permitting risk remains site-specific because the MoU itself assigns work on site selection, permitting, regulatory processes, project structure, and financing to the partners. | 中 | SR002, SR030 |
| CR040 | IPP dependence is real because IPP leads plasma physics and scientific leadership for Alpha while Proxima leads engineering, procurement, and construction. | 中 | SR002, SR030, SR007 |
| CR041 | RWE’s Gundremmingen infrastructure and operating experience are meaningful mitigants, but they do not prove fusion plant licensing, construction, or grid connection on schedule. | 中 | SR002, SR006, SR031 |
| CR042 | Proxima’s near-term execution depends on completing the Stellarator Model Coil and scaling HTS cable and magnet production. | 中 | SR001, SR003 |
| CR043 | Strategic energy demand from AI and data centers supports the long-term buyer narrative, but actual offtake economics remain contingent on plant cost, schedule, reliability, and grid readiness. | 中 | SR001, SR002, SR017 |
| CR044 | The strongest public mitigants are deep public-private partnerships, Max Planck scientific credibility, RWE site/operator involvement, and a large fresh financing round. | 中 | SR001, SR002, SR006, SR007, SR030 |
| CR045 | The highest-severity thesis-break event would be Alpha missing net-energy operation by a wide margin, because Alpha is the bridge between research inheritance and commercial deployment. | 中 | SR001, SR002, SR014, SR016 |
| CR046 | Residual risk remains high because technical proof, plant economics, licensing implementation, and multi-billion-euro financing are all unresolved at the same time. | 中 | SR001, SR002, SR014, SR017, SR018, SR024 |
| CR047 | Proportionate fusion regulation reduces fission-style burden but does not eliminate tritium, waste, environmental assessment, local siting, or public-acceptance risk. | 中 | SR010, SR013, SR027, SR014 |
| CR048 | Legal/IP restrictions on publicly funded R&D can affect commercialization if exploitation rights, EEA/Suisse transfer limits, consortium access, or grant obligations are not cleanly controlled. | 中 | SR025 |
| CR049 | Follow-on financing should be gated by verified model-coil progress, Alpha schedule/budget control, regulator feedback, tritium/materials plans, and credible offtake economics. | 中 | SR001, SR002, SR014, SR017, SR024, SR027 |
| CR050 | The reviewed evidence supports a high-risk, milestone-gated underwriting posture rather than treating the July 2026 mega-round as proof that commercial fusion risk is solved. | 中 | SR001, SR014, SR017, SR018, SR024 |
| CV001 | Proxima announced a €411 million ($468 million) financing round in July 2026. | 高 | SV001, SV002, SV003 |
| CV002 | The July 2026 round was disclosed at a €2.4 billion post-money valuation, approximately $2.7 billion. | 高 | SV001, SV002, SV005, SV007 |
| CV003 | The round was led by XTX Ventures and East X Ventures, with strategic participation from Google and RWE. | 高 | SV001, SV002, SV003, SV004 |
| CV004 | Proxima says it has raised more than €650 million in total including public grants since its 2023 founding. | 高 | SV001, SV002, SV007 |
| CV005 | RWE disclosed a €25 million investment in Proxima in connection with the July 2026 financing. | 高 | SV004, SV001 |
| CV006 | Registry sources identify Proxima Fusion GmbH in Munich under Amtsgericht München HRB 283423. | 高 | SV010, SV011, SV012 |
| CV007 | Proxima is still pre-revenue in the investment sense because its public milestones concern demonstrators and future power plants rather than current electricity sales. | 中 | SV001, SV017, SV031 |
| CV008 | The central milestone behind the valuation is Alpha, a planned net-energy stellarator demonstrator near Munich in the early 2030s. | 中 | SV001, SV031, SV009 |
| CV009 | The commercial power-plant option is tied to a Bavaria/RWE/IPP agreement targeting a stellarator power plant on the grid in Europe later in the 2030s. | 中 | SV031, SV001, SV009 |
| CV010 | On fundamentals alone the €2.4 billion mark is stretched because the company has no product revenue, no earnings, and years of technical milestones ahead. | 中 | SV001, SV017, SV018, SV019 |
| CV011 | The strongest thesis is option value: a differentiated quasi-isodynamic stellarator path, strategic European energy-security relevance, and enough capital to reach Alpha-scale hardware execution. | 中 | SV001, SV006, SV013, SV031 |
| CV012 | The strongest anti-thesis is that fusion remains capital-intensive and pre-commercial, with no private company yet demonstrating a full commercial electricity business. | 中 | SV017, SV018, SV019 |
| CV013 | The Fusion Industry Association's 2025 global-industry presentation counted 53 verified private fusion companies and $9.76 billion of funding. | 高 | SV030, SV013 |
| CV014 | The FIA 2026 supply-chain report says fusion supply-chain spending increased 24% in 2025 to $538 million and was projected to rise 27% in 2026 to $681 million. | 中 | SV013 |
| CV015 | Clean Energy Platform frames fusion as having crossed a $15 billion cumulative investment milestone by late 2025, showing that market-data sources differ on scope and timing. | 中 | SV015, SV030 |
| CV016 | TechCrunch reported in June 2026 that fusion startups had raised $7.1 billion to date, concentrated among a handful of companies. | 中 | SV016 |
| CV017 | Helion's June 2026 Series G raised $465 million at a $15.5 billion post-money valuation. | 高 | SV023, SV024 |
| CV018 | Helion says the Series G brought total funding to more than $1.5 billion and supports its effort to supply at least 50 MW to Microsoft. | 中 | SV023, SV024 |
| CV019 | Commonwealth Fusion Systems raised $863 million in a 2025 Series B2 round and TechCrunch said it had raised nearly $3 billion by then. | 高 | SV020, SV021 |
| CV020 | Sacra's current CFS profile describes CFS as having raised more capital than any other private fusion company globally, but its 2026 funding totals are an analyst-data point rather than an audited filing. | 中 | SV022, SV020, SV021 |
| CV021 | TAE Technologies announced more than $150 million in latest funding in 2025, and Tracxn lists $1.32 billion of total TAE funding over 11 rounds. | 中 | SV025, SV026 |
| CV022 | Tracxn lists TAE's July 2022 Series G post-money valuation at $1.2 billion while not providing a current post-money valuation for the 2025 extension. | 中 | SV026 |
| CV023 | Tokamak Energy announced a $125 million 2024 financing to commercialize fusion and high-temperature superconducting magnet technologies. | 中 | SV027 |
| CV024 | Tokamak Energy's disclosed funding scale is below Proxima's, with CB Insights showing $226.5 million raised while Tokamak's own announcement cites $335 million including public and private investment. | 中 | SV027, SV029 |
| CV025 | Using the disclosed dollar equivalents, Proxima's $2.7 billion valuation is roughly 3.6 times its approximately $740 million total capital raised. | 中 | SV001, SV002 |
| CV026 | If the €411 million round were treated as all primary capital, it would represent about 17% of the €2.4 billion post-money valuation before any preference details. | 中 | SV001, SV002 |
| CV027 | The current mark implies investors are underwriting Alpha, magnet vertical integration, and strategic project credibility before revenue proves the business model. | 中 | SV001, SV006, SV031 |
| CV028 | A bear-case public-evidence valuation range of €0.4 billion to €1.0 billion is appropriate if Alpha slips materially and the asset is valued mainly as IP, team, and residual strategic optionality. | 中 | SV017, SV018, SV019 |
| CV029 | A base-case public-evidence valuation range of €1.8 billion to €3.0 billion is appropriate when the July 2026 mark is accepted but discounted for milestone and dilution risk. | 中 | SV001, SV002, SV013, SV017 |
| CV030 | A bull-case valuation range of €8 billion to €15 billion or more requires credible Alpha execution, strategic offtake, and peer-like validation closer to Helion's 2026 valuation benchmark. | 中 | SV023, SV024, SV031, SV001 |
| CV031 | An illustrative probability-weighted valuation using 25% bear, 55% base, and 20% bull cases lands around €3.5 billion, but dilution and preference risk pull the investable common-equity value below that headline. | 中 | SV001, SV017, SV023, SV024 |
| CV032 | The peer comparable set is necessarily a sample because most private fusion valuations and round terms remain undisclosed or paywalled. | 中 | SV016, SV022, SV026, SV029 |
| CV033 | A venture-method investor seeking a 5x to 10x gross outcome at a €2.4 billion entry would need an eventual exit value roughly in the €12 billion to €24 billion range before later dilution. | 中 | SV001, SV023, SV024 |
| CV034 | Future dilution is a material risk because fusion companies can require billions of upfront capital before electricity revenue begins. | 中 | SV017, SV018, SV022, SV030 |
| CV035 | TechCrunch's adverse reporting highlights concern that some fusion companies may go public before milestones such as scientific breakeven are achieved. | 中 | SV017 |
| CV036 | MIT Technology Review's cost skepticism argues against assuming fusion electricity will automatically be cheap even if the physics works. | 中 | SV018 |
| CV037 | The Bulletin frames commercial fusion as a hype-prone big-bet technology whose risks and benefits are not centrally assessed by a U.S. agency. | 中 | SV019 |
| CV038 | Entry at the €2.4 billion mark should require cap-table terms, preference stack, milestone financing plan, and grant conditionality before underwriting common-equity upside. | 中 | SV001, SV010, SV017, SV018 |
| CV039 | The best public recommendation is track or research-more rather than buy, because company quality is high but valuation support is milestone- and structure-dependent. | 中 | SV001, SV013, SV017, SV018, SV023 |
| CV040 | Kill triggers include Alpha slipping beyond the early 2030s, magnet manufacturing failing to scale, partner commitments weakening, or capital markets refusing the next multibillion-euro step-up. | 中 | SV006, SV017, SV018, SV031 |
| CV041 | Upside triggers include verified Alpha net-energy progress, HTS magnet production at industrial scale, strengthened RWE or utility commitments, and non-dilutive public support. | 中 | SV001, SV004, SV006, SV013, SV031 |
| CV042 | Revenue, EBITDA, and ARR multiples do not apply to Proxima because no public evidence shows current product revenue or earnings. | 中 | SV001, SV017, SV031 |
| CV043 | Capital-raised benchmarking is useful for fusion because peers disclose large funding rounds even when valuations and operating metrics are private. | 中 | SV016, SV020, SV021, SV023, SV026, SV027 |
| CV044 | A milestone-weighted option method is the best public valuation method because almost all value depends on staged technical, financing, regulatory, and project-development proof. | 中 | SV001, SV013, SV017, SV018, SV031 |
| CV045 | Public registry and press sources verify Proxima's existence and headline financing, but they do not disclose liquidation preferences, ownership, project-finance needs, or investor protections. | 中 | SV001, SV010, SV011, SV012 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | Proxima Fusion | Proxima Fusion | Building stellarators to power the future | We are Europe's fastest-growing fusion company, building the first generation of fusion power plants using quasi-isodynamic (QI) stellarators. |
| SO002 | Proxima Fusion | About | Proxima Fusion | Proxima is a fusion energy company headquartered in Munich. |
| SO003 | Proxima Fusion | Technology | Proxima Fusion | QI-HTS stellarators like Proxima's first-of-a-kind fusion power plant concept, Stellaris, also offer a proven heat exhaust concept. |
| SO004 | Proxima Fusion | Proxima Fusion Raises €411 Million to Build Europe’s Commercial Fusion Champion | Proxima Fusion today announced a €411 million ($468 million) financing round, bringing the company’s valuation to €2.4 billion ($2.7 billion). |
| SO005 | Proxima Fusion | Proxima Fusion and Partners Publish Stellaris Fusion Power Plant Concept | New peer-reviewed paper published in Fusion Engineering and Design presents Stellaris, the world’s first integrated concept for a commercial fusion power plant. |
| SO006 | Proxima Fusion | Proxima Fusion Welcomes Sergei Galperin as Chief Financial Officer | Sergei will lead the company's financial strategy as it advances toward commercial fusion. |
| SO007 | Proxima Fusion | Proxima Fusion convenes Industrial Development Board to scale Europe’s fusion future | The Industrial Development Board builds on the momentum of the Alpha Alliance, Proxima’s industrial partner network, which has already grown to more than 50 companies since February 2026. |
| SO008 | Proxima Fusion | How AI Changed the Way We Build at Proxima | How AI Changed the Way We Build at Proxima. |
| SO009 | Proxima Fusion | Careers | Proxima Fusion | Help us put fusion energy on the grid. |
| SO010 | Proxima Fusion | Proxima Fusion | Press & News | Registered at the District Court of Munich under HRB 283423, Managing Directors Dr. Francesco Sciortino, Lucio Milanese. |
| SO011 | Max Planck Institute for Plasma Physics | Max Planck Institute for Plasma Physics signs cooperation agreement with German fusion start-up Proxima Fusion | The Max Planck Institute for Plasma Physics will work with the start-up Proxima Fusion, the first spin-out company in the history of the IPP. |
| SO012 | Max Planck Institute for Plasma Physics | Wendelstein 7-X | Wendelstein 7-X is the world's most modern and powerful stellarator experiment. |
| SO013 | Proxima Fusion | Proxima Fusion and the Paul Scherrer Institute Sign Framework Agreement | Proxima Fusion and the Paul Scherrer Institute sign framework agreement for high-temperature superconducting magnet technology. |
| SO014 | Max Planck Institute for Plasma Physics | BMBF project by Proxima Fusion and IPP starts with joint workshop | BMBF project by Proxima Fusion and IPP starts with joint workshop. |
| SO015 | Max Planck Innovation | 7 million euros for MPG spin-out Proxima Fusion | 7 million euros for MPG spin-out Proxima Fusion. |
| SO016 | EU-Startups | Munich-based Proxima Fusion raises €7 million to bring in the next generation fusion power plants | Munich-based Proxima Fusion raises €7 million to bring in the next generation fusion power plants. |
| SO017 | NucNet | Germany’s Proxima Raises €20 Million For QI Stellarator Reactor | Germany’s Proxima Raises €20 Million For QI Stellarator Reactor. |
| SO018 | TechCrunch | Proxima Fusion joins the club of well-funded nuclear contenders with €130M Series A | Commercial nuclear fusion power isn’t a reality yet. |
| SO019 | Sifted | Germany’s Proxima Fusion secures €130m | No fusion machine has yet managed to produce more energy than it consumes, but investor appetite for the industry is on the up. |
| SO020 | CNBC | Google backs nuclear fusion startup targeting Europe’s first commercial power plant | Commercializing the tech is immensely challenging, and success is not guaranteed. |
| SO021 | RWE | RWE invests 25 million euros in fusion technology start-up Proxima Fusion | RWE has invested 25 million euros in the funding round of Munich-based magnetic fusion start-up Proxima Fusion. |
| SO022 | Wikipedia | Proxima Fusion | Proxima was founded in April 2023 by Francesco Sciortino, Lucio Milanese, Jorrit Lion, Jonathan Schilling, and Martin Kubie. |
| SO023 | North Data | Proxima Fusion GmbH, Munich, Germany, District Court of Munich HRB 283423 | Proxima Fusion GmbH; District Court of Munich HRB 283423; corporate purpose includes research and development in nuclear fusion. |
| SO024 | redalpine | redalpine portfolio assets | proxima fusion | Proxima Fusion's team of nuclear physicists, engineers, and computer scientists from Max Planck Institute, MIT, and Google combines research, technological breakthroughs, and AI. |
| SO025 | XTX Markets | Ventures | XTX Markets | XTX Ventures invests in technical founders who combine deep AI and machine learning expertise with a clear vision. |
| SO026 | arXiv | Quasi-isodynamic stellarators with low turbulence as fusion reactor candidates | Quasi-isodynamic stellarators with low turbulence as fusion reactor candidates. |
| SO027 | U.S. Department of Energy | DOE National Laboratory Makes History by Achieving Fusion Ignition | DOE National Laboratory Makes History by Achieving Fusion Ignition. |
| SO028 | ITER Organization | Advantages of fusion | Advantages of fusion. |
| SO029 | EE Times | Proxima Fusion Unveils Stellaris: A Breakthrough in Fusion Power | Proxima Fusion Unveils Stellaris: A Breakthrough in Fusion Power. |
| SM001 | Proxima Fusion | Proxima Fusion | Building stellarators to power the future | We are Europe's fastest-growing fusion company, building the first generation of fusion power plants using quasi-isodynamic (QI) stellarators. |
| SM002 | Proxima Fusion | Proxima Fusion Raises €411 Million to Build Europe’s Commercial Fusion Champion | Proxima announced a €411 million financing round, bringing the company’s valuation to €2.4 billion. |
| SM003 | Proxima Fusion | Proxima Fusion and Partners Publish Stellaris Fusion Power Plant Concept | Stellaris is the world’s first integrated concept for a commercial fusion power plant designed to operate reliably and continuously. |
| SM004 | Proxima Fusion | Technology | Proxima Fusion | Proxima is building fusion power plants using QI stellarators and high-temperature superconducting magnet technology. |
| SM005 | CNBC | Google backs nuclear fusion startup targeting Europe’s first commercial power plant | Google added that while fusion could change the world, commercializing the tech is immensely challenging, and success is not guaranteed. |
| SM006 | EU-Startups | Largest European fusion investment on record sees Proxima Fusion raise €411 million | Largest European fusion investment on record sees Proxima Fusion raise €411 million. |
| SM007 | Fusion Industry Association | In the News: The Global Fusion Industry in 2025 | The FIA’s Global Fusion Industry in 2025 report launched on July 22 and reported over $2.5 billion in new investment. |
| SM008 | Fusion Industry Association | FIA Launches 2026 Fusion Industry Supply Chain Report | Supply chain spending by the fusion industry increased by 24% in 2025. |
| SM009 | Fusion Industry Association | EU Commission Includes Fusion As A Priority In Its 2026 Work Programme | The 2026 Work Programme includes a plan for a European strategy to accelerate fusion commercialization. |
| SM010 | Fusion Industry Association | Germany Unveils "Fusion Action Plan" | The government will invest more than €2 billion by 2029 to support fusion research and pilot projects. |
| SM011 | BMFTR | Fusion energy: call for European collaboration | Fusion offers the prospect of a new energy age: clean, safe, low-carbon and not subject to the weather. |
| SM012 | Clean Energy Wire | Germany joins EU nuclear fusion push as Europe looks to energy security | Fusion technology is still in the experimental stage, and even many proponents say commercial use is at least decades away. |
| SM013 | Noerr | Fusion power in Germany and the EU: political momentum and IP strategies as a key success factor | EURATOM Work Programme 2026–2027 allocates €222 million to accelerate fusion energy from the lab to the grid. |
| SM014 | Fusion Industry Association | IAEA Director General Grossi Launches World Fusion Outlook 2025 | The World Fusion Outlook 2025 includes a country-by-country assessment of fusion development. |
| SM015 | McKinsey & Company | How data centers and the energy sector can sate AI’s hunger for power | Between 2024 and 2030, electricity demand for data centers in the United States is expected to increase by about 400 terawatt-hours. |
| SM016 | IEEE Spectrum | AI Data Centers Are Driving an Energy Crunch | AI data center energy consumption is creating new pressure on electricity systems. |
| SM017 | Precedence Research | Nuclear Fusion Market Size, Trends, Growth, Report 2040 | The global nuclear fusion market size is expected at USD 471.99 billion in 2030 and around USD 843.46 billion by 2040. |
| SM018 | BloombergNEF | AI Data Center Build Advances at Full Speed: Five Things to Know | Capex of largest data center firms nears $750 billion in 2026. |
| SM019 | BloombergNEF | Data Centers Set to Double Their Power Demand in Europe | Large data centers in the U.K., Germany, Ireland, Norway and the Netherlands are projected to draw 5.4GW in live IT power demand in 2030. |
| SM020 | Google Data Centers | Operating sustainably – Google Data Centers | In 2025 alone, we signed agreements for over 12 GW of net-new clean energy. |
| SM021 | BloombergNEF | Google Sees 2026 as Year of Scaling Up Advanced Energy Technologies | Google aims to scale up next generation technologies like geothermal, nuclear and long duration energy storage this year. |
| SM022 | RWE | This is RWE | RWE has reliably supplied people and companies with electricity for more than 125 years – night and day. |
| SM023 | Proxima Fusion | How AI Changed the Way We Build at Proxima | Proxima describes itself as physics simulation, engineering design, advanced manufacturing, procurement and industrial facility development company. |
| SM024 | Proxima Fusion | Proxima Fusion convenes Industrial Development Board to scale Europe’s fusion future | The Industrial Development Board brings together experienced industrial and energy leaders to accelerate large-scale industrial deployment. |
| SM025 | BloombergNEF | Nuclear Fusion Powers Up for Commercial Breakthrough | Fusion energy is entering a new phase, attracting billions in private and public capital. |
| SM026 | Statista | Data Center - Worldwide | Statista Market Forecast | The data center market worldwide is expected to see growth with increasing demand for data storage and processing. |
| SM027 | International Energy Agency | Energy and AI – Analysis - IEA | IEA publishes dedicated analysis on energy and AI, underscoring the relevance of AI electricity demand to energy planning. |
| SP001 | Fusion Industry Association | Members - Fusion Industry Association | The Fusion Industry Association is the leading voice of the fusion industry. |
| SP002 | Fusion Industry Association | Fusion Industry Reports | |
| SP003 | Fusion Industry Association | 2026 The Fusion Industry Supply Chain | 25 fusion companies completed the survey; total reported spend in 2025 was $538 million and projected spend for 2026 was $681 million. |
| SP004 | TechCrunch | Every fusion startup that has raised over $100M | Commonwealth Fusion Systems has raised about a third of all private capital invested in fusion companies to date. |
| SP005 | Commonwealth Fusion Systems | Technology | |
| SP006 | Commonwealth Fusion Systems | SPARC: Proving commercial fusion energy is possible | |
| SP007 | Commonwealth Fusion Systems | ARC: Putting fusion energy on the grid | |
| SP008 | TAE Technologies | Clean energy solutions for a bright future | |
| SP009 | Helion Energy | Technology | |
| SP010 | TechCrunch | Helion, the Sam Altman-backed fusion startup, raises $465M to build a power plant for Microsoft | Helion raised $465 million in a new funding round that values the company at $15.5 billion. |
| SP011 | Tokamak Energy | Delivering fusion energy and HTS magnet technology | |
| SP012 | General Fusion | Home | |
| SP013 | Zap Energy | Zap Energy: The atom, twice unlocked | |
| SP014 | First Light Fusion | First Light Fusion | Enabling Inertial Fusion Energy | |
| SP015 | Type One Energy | Home | |
| SP016 | Thea Energy | Thea Energy | Fusion Power | Reinventing the Stellarator | |
| SP017 | Marvel Fusion | Scalable and Globally Competitive Fusion Energy for a Sustainable Future on Earth | |
| SP018 | Focused Energy | Focused Energy — Laser Fusion | |
| SP019 | Gauss Fusion | Leading European Industries to build Fusion Power Plants | |
| SP020 | Renaissance Fusion | Renaissance Fusion | |
| SP021 | Renaissance Fusion | Stellarator Fusion Technology & HTS Magnets | |
| SP022 | Xcimer Energy | Laser Fusion Leader | |
| SP023 | Pacific Fusion | Pacific Fusion – Powering a prosperous planet | |
| SP024 | IEEE Spectrum | Stellarator Showdown: Proxima Fusion vs. Type One Energy | Both companies face a common challenge: funding. |
| SP025 | Undark | Startups Gamble on Fusion Energy | It remains to be seen whether these fusion startups can build reactors in the near future capable of generating hundreds of megawatts. |
| SP026 | Proxima Fusion | Proxima Fusion Raises €411 Million to Build Europe’s Commercial Fusion Champion | |
| SP027 | EU-Startups | Largest European fusion investment on record sees Proxima Fusion raise €411 million | |
| SP028 | Business Wire | Proxima Fusion Raises €411 Million at a €2.4B Valuation to Build Europe’s Commercial Fusion Champion | |
| SP029 | arXiv | Quasi-isodynamic stellarators with low turbulence as fusion reactor candidates | |
| SP030 | Proxima Fusion | Technology | |
| SP031 | Proxima Fusion | How AI Changed the Way We Build at Proxima | |
| SP032 | Proxima Fusion | Proxima Fusion and Partners Publish Stellaris Fusion Power Plant Concept | |
| SP033 | Max Planck Institute for Plasma Physics | Wendelstein 7-X | |
| SI001 | Proxima Fusion | Proxima Fusion Raises €411 Million to Build Europe's Commercial Fusion Champion | In less than three years, Proxima has secured more than €650 million ($740 million), including €95 million in public grants. |
| SI002 | Business Wire | Proxima Fusion Raises €411 Million at a €2.4B Valuation to Build Europe's Commercial Fusion Champion | |
| SI003 | CNBC | Google backs nuclear fusion startup targeting Europe's first commercial power plant | |
| SI004 | RWE | RWE invests 25 million euros in fusion technology start-up Proxima Fusion | RWE has invested 25 million euros in the funding round of Munich-based magnetic fusion start-up Proxima Fusion. |
| SI005 | Orrick | Proxima Fusion Raises €411 Million to Build Europe's Commercial Fusion Champion | |
| SI006 | Burda Principal Investments | Burda Principal Investments invests in Proxima Fusion | |
| SI007 | High-Tech Gründerfonds | Europe's Fusion Champion Comes of Age: Proxima Fusion Raises €411 Million for Alpha | |
| SI008 | Data Center Dynamics | European fusion company Proxima raises €411m with backing from Google | |
| SI009 | Global Banking & Finance Review | Proxima Fusion Secures €411M from Google, RWE in Major Financing Round | |
| SI010 | TechFundingNews | Proxima Fusion raises €411M at €2.4B valuation, marking Google's first fusion investment | |
| SI011 | TechCrunch | Proxima Fusion joins the club of well-funded nuclear contenders with €130M Series A | |
| SI012 | Sifted | Proxima Fusion raises €130m | |
| SI013 | DeepTech & Climate Fonds | Proxima Fusion raises €130M Series A to build world's first stellarator-based fusion power plant in the 2030s | |
| SI014 | Proxima Fusion | Proxima Fusion Raises €20M Seed to Accelerate Timeline to Fusion Power | |
| SI015 | High-Tech Gründerfonds | HTGF Investment Proxima Fusion | |
| SI016 | NucNet | Germany's Proxima Raises EUR20 Million For QI Stellarator Reactor | |
| SI017 | Max Planck Innovation | 7 million euros for MPG spin-off Proxima Fusion | |
| SI018 | EU-Startups | Munich-based Proxima Fusion raises €7 million to bring in the next generation fusion power plants | |
| SI019 | Northdata | Proxima Fusion GmbH, München, Amtsgericht München HRB 283423 | |
| SI020 | Online-Handelsregister | Handelsregisterauszug von Proxima Fusion GmbH aus München (HRB 283423) | |
| SI021 | Proxima Fusion | Proxima Fusion Extends Series A to €200M Total Funding As It Accelerates Into Hardware Execution | |
| SI022 | CDP Venture Capital | Proxima Fusion Fires Up Europe's Race for Clean Energy with €200M Milestone | |
| SI023 | TechFundingNews | Proxima Fusion secures €400M from Bavaria to build €2B fusion test facility | |
| SI024 | MIT Technology Review | Will fusion power get cheap? Don't count on it. | Fusion power could provide a steady, zero-emissions source of electricity in the future—if companies can get plants built and running. But a new study suggests that even if that future arrives, it might not come cheap. |
| SI025 | MIT Technology Review | Fusion power plants don't exist yet, but they're making money anyway | |
| SI026 | Bulletin of the Atomic Scientists | What's fueling the commercial fusion hype? | |
| SI027 | Bulletin of the Atomic Scientists | Introduction—Fusion, forever the energy of tomorrow? | Economic [commercialized] fusion is not a cakewalk; it is not an assured thing at all. |
| SI028 | TechCrunch | Every fusion startup that has raised over $100M | |
| SI029 | Commonwealth Fusion Systems | Commonwealth Fusion Systems Raises $863 Million Series B2 Round | |
| SE001 | Proxima Fusion | Technology | QI-HTS stellarators using high-temperature superconducting magnets offer the clearest path to putting fusion on the grid. |
| SE002 | Proxima Fusion | Proxima Fusion and Partners Publish Stellaris Fusion Power Plant Concept | Stellaris is the first QI stellarator-based power plant design that simultaneously meets all major physics and engineering constraints. |
| SE003 | Proxima Fusion | Proxima and Hugging Face Announce the ConStellaration Challenge | We propose three benchmark problems of increasing complexity, each with progressive relevance to fusion reactor design. |
| SE004 | Proxima Fusion | Proxima Fusion, RWE, Bavaria and IPP Sign Agreement | Alpha will become the first stellarator to demonstrate net energy gain. |
| SE005 | GitHub | proximafusion organization | vmecpp is a from-scratch C++ and Python reimplementation of the Variational Moments Equilibrium Code. |
| SE006 | GitHub | proximafusion/constellaration | ConStellaration is a dataset of diverse QI-like stellarator plasma boundary shapes and optimization benchmarks. |
| SE007 | Hugging Face | proxima-fusion/constellaration dataset | The fetched dataset page lists 182k rows and metrics including qi, max_elongation and vacuum_well. |
| SE008 | PyPI | constellaration | The fetched page showed a site-loading limitation, but the repository documents installation from PyPI. |
| SE009 | arXiv | ConStellaration: A dataset of QI-like stellarator plasma boundaries and optimization benchmarks | The preprint releases an open dataset of diverse QI-like stellarator plasma boundary shapes paired with ideal MHD equilibria and performance metrics. |
| SE010 | arXiv | ConStellaration PDF | The PDF contains the full ConStellaration methods paper and benchmark description. |
| SE011 | ScienceDirect | Stellaris: A high-field quasi-isodynamic stellarator for a prototypical fusion power plant | Fetch reached a robot challenge, so bibliographic use is corroborated with DOI and Proxima's peer-review announcement. |
| SE012 | DOI Foundation | DOI 10.1016/j.fusengdes.2025.114868 | The DOI identifies the Fusion Engineering and Design Stellaris article. |
| SE013 | Max Planck Institute for Plasma Physics | Wendelstein 7-X | W7-X is a large stellarator with modular superconducting coils which enable steady state plasma operation. |
| SE014 | Max Planck Institute for Plasma Physics | New performance records Wendelstein 7-X | W7-X achieved a world record for the triple product in long plasma discharges. |
| SE015 | American Physical Society | Tokamak level performance in the optimized stellarator Wendelstein 7-X with stable peaked density profiles | The paper reports a triple product of (1.10 ± 0.15) × 10^20 m^-3 keV s held stable for 1.9 s. |
| SE016 | arXiv | Enhanced performance in quasi-isodynamic max-J stellarators with a self-sustained density gradient | Recent stellarator reactor designs demonstrate mostly outward turbulent particle transport, which inhibits density gradients needed for confinement. |
| SE017 | EE Times | Proxima Fusion Unveils Stellaris: A Breakthrough in Fusion Power | |
| SE018 | Nuclear Engineering International | Proxima unveils Stellaris fusion plant design | This still leaves many technological and engineering challenges. |
| SE019 | Business Wire | Proxima Fusion and Partners Publish First-of-a-Kind Fusion Power Plant Concept | |
| SE020 | NS Energy | QI Stellarators Move Into the Spotlight for Fusion Commercialisation | |
| SE021 | Fusion Future | Proxima Fusion, Bavaria, RWE and IPP MoU for Alpha and Stellaris | Whether Alpha reaches operations in the 2030s at the stated budget depends on federal funding decisions not yet made, regulatory frameworks and engineering challenges. |
| SE022 | Nuclear Engineering International | Proxima signs stellarator roadmap | |
| SE023 | DeepTech & Climate Fonds | Press Release: Proxima Fusion Alpha | |
| SE024 | MIT News | Will neutrons compromise the operation of superconducting magnets in a fusion plant? | There remains the important issue of longer-term degradation of the REBCO that would occur over years or decades. |
| SE025 | Clean Energy Platform | Fueling Fusion Reactors: The Real Constraint in 2026 | The civilian stockpile remains at a critical threshold of just 20 to 30 kilograms. |
| SE026 | U.S. Department of Energy | DOE Explains...Tokamaks | |
| SE027 | U.S. Department of Energy | DOE Explains...Stellarators | Making stellarator coils is a challenge because it requires manufacturers to construct large bore wire coils with millimeter precision. |
| SE028 | U.S. Department of Energy | The Road to a Future Powered by Fusion May Be Twisty | Tokamaks can experience internal disruptions that can damage the devices' interior walls. |
| SU001 | Proxima Fusion | Proxima Fusion, RWE, the Free State of Bavaria and Max Planck Institute for Plasma Physics Sign Agreement to Build the World’s First Commercial Fusion Power Plant in Europe | Proxima Fusion signed an agreement with the Free State of Bavaria, RWE, and Max Planck IPP to put the world’s first commercial stellarator fusion power plant on the grid in Europe. |
| SU002 | RWE | RWE invests 25 million euros in fusion technology start-up Proxima Fusion | RWE has invested 25 million euros in the funding round of Munich-based magnetic fusion start-up Proxima Fusion. |
| SU003 | Proxima Fusion | Proxima Fusion Raises €411 Million to Build Europe’s Commercial Fusion Champion | Proxima Fusion raised €411 million to build Europe’s commercial fusion champion. |
| SU004 | Max Planck Institute for Plasma Physics | Memorandum of Understanding | IPP, the Bavarian State Government, Proxima Fusion and RWE signed a memorandum of understanding. |
| SU005 | Bayerische Staatsregierung | Newsletter der Bayerischen Staatsregierung – 27. Februar 2026 | Bavaria described the Alpha demonstrator and Gundremmingen fusion power plant project after the February 2026 agreement. |
| SU006 | BMFTR | Fusion energy: call for European collaboration | Germany called for European collaboration on fusion energy in 2026. |
| SU007 | World Nuclear News | Agreement to build German nuclear fusion power plant | The agreement covers Alpha near Garching and Stellaris at the former Gundremmingen nuclear site. |
| SU008 | CNBC | Google backs Proxima Fusion in $468 million round | Google backed Proxima Fusion in a $468 million funding round. |
| SU009 | Energy Connects | Google and RWE Back German Nuclear Startup Proxima Fusion at €2.4 Billion Valuation | Google and RWE backed Proxima Fusion at a €2.4 billion valuation. |
| SU010 | NucNet | Google And RWE Back Germany’s Proxima In Bid For Europe’s First Nuclear Fusion Plant | NucNet reported Google and RWE backing Proxima in its bid for Europe’s first nuclear fusion plant. |
| SU011 | Reuters via Yahoo Finance | Google, RWE back Proxima Fusion in €411 million financing round | Reuters reported Google and RWE backing Proxima Fusion in a €411 million financing round. |
| SU012 | Energy-pedia | Germany: RWE invests 25 million euros in fusion technology start-up Proxima Fusion | Energy-pedia covered RWE’s €25 million investment in Proxima Fusion. |
| SU013 | Max-Planck-Gesellschaft | Proxima Fusion raises 411 million euros | Max Planck reported that Proxima Fusion raised 411 million euros. |
| SU014 | Proxima Fusion | Technology | Proxima describes its stellarator fusion technology and its commercialization plan. |
| SU015 | Proxima Fusion | Proxima Fusion and Partners Publish Stellaris Fusion Power Plant Concept to Bring Limitless, Safe, Clean Energy to the Grid | Proxima and partners published the Stellaris fusion power plant concept for grid-connected fusion energy. |
| SU016 | PR Newswire | Google and Commonwealth Fusion Systems Sign Strategic Partnership Agreement to Commercialize Fusion Energy | Google and CFS announced a strategic partnership and power purchase agreement to commercialize fusion energy. |
| SU017 | Data Center Dynamics | Google signs 200MW fusion PPA with Commonwealth Fusion Systems | Google signed a 200 MW fusion PPA with Commonwealth Fusion Systems. |
| SU018 | World Nuclear News | Google signs up for power from future fusion plant | World Nuclear News reported Google signing up for power from a future fusion plant. |
| SU019 | Helion Energy | Helion announces world’s first fusion PPA with Microsoft | Helion announced an agreement to provide Microsoft electricity from its first fusion power plant. |
| SU020 | IEA | Energy and AI | The IEA analyzed the rising electricity demand implications of AI and data centers. |
| SU021 | Clean energy for Google data centers | Google describes its work to operate data centers with clean energy. | |
| SU022 | Fusion Industry Association | Germany Unveils Fusion Action Plan | Germany unveiled a Fusion Action Plan to accelerate commercial fusion energy. |
| SU023 | Neutron Bytes | Proxima Fusion Funding Fuels 2030s Date for Power Plant | Neutron Bytes questioned where Proxima’s public-relations claims end and hard facts begin. |
| SU024 | Energy Solutions | Fusion Energy Breakthroughs 2026: From Scientific Milestones to Commercial Reality | The 2026 review said engineering, regulatory, fuel-cycle, and cost challenges still stand between fusion milestones and commercial power. |
| SU025 | U.S. Department of Energy | Energy Department Releases Finalized Fusion Science and Technology Roadmap to Accelerate Commercial Fusion Energy | DOE released a roadmap intended to accelerate commercial fusion energy. |
| SU026 | Enlit World | Fusion energy expected in early 2030s, indicates survey | Enlit reported survey expectations for fusion energy in the early 2030s. |
| SR001 | Proxima Fusion | Proxima Fusion raises €411 million to build Europe’s commercial fusion champion | The financing provides the backing needed to build Alpha, Proxima’s net-energy stellarator demonstrator near Munich. |
| SR002 | Proxima Fusion | Proxima Fusion, RWE, the Free State of Bavaria and Max Planck IPP sign agreement to build the world’s first commercial fusion power plant in Europe | Under the MoU, the Free State of Bavaria, Proxima Fusion, RWE and IPP will work together on site selection, permitting and regulatory processes, project structure, and financing. |
| SR003 | Proxima Fusion | Technology | |
| SR004 | Proxima Fusion | Proxima Fusion convenes Industrial Development Board to scale Europe’s fusion future | |
| SR005 | Proxima Fusion | Proxima Fusion welcomes Sergei Galperin as Chief Financial Officer | |
| SR006 | RWE | RWE invests 25 million euros in fusion technology start-up Proxima Fusion | |
| SR007 | Max Planck Society | Proxima Fusion raises 411 million euros | |
| SR008 | Federal Government of Germany | Fusionskraftwerk in Deutschland | Fusion ist allerdings keine kurzfristige Lösung. Auf dem Weg zu einem ersten Fusionskraftwerk sind noch erhebliche technologische Herausforderungen zu überwinden. |
| SR009 | Federal Ministry of Research, Technology and Space | Europe needs more private investment in fusion energy | |
| SR010 | U.S. Nuclear Regulatory Commission | Fusion | |
| SR011 | U.S. Nuclear Regulatory Commission | Fusion Machine Rulemaking Status | |
| SR012 | Federal Register | Regulatory Framework for Fusion Machines | |
| SR013 | UK Department for Energy Security and Net Zero | Draft National Policy Statement for fusion energy generation (EN-8) | |
| SR014 | U.S. Government Accountability Office | Fusion Energy: Potentially Transformative Technology Still Faces Fundamental Challenges | Several challenges must be overcome to achieve commercial fusion, and stakeholders’ projections of this timeline range from 10 years to several decades. |
| SR015 | U.S. Department of Energy | Fusion Science and Technology Roadmap | |
| SR016 | World Nuclear Association | Fusion Power | |
| SR017 | MIT Technology Review | Will fusion power get cheap? Don’t count on it. | The final figure the researchers suggest for fusion’s experience rate is between 2% and 8%. |
| SR018 | Bulletin of the Atomic Scientists | What’s fueling the commercial fusion hype? | The hard truth is that scientists and engineers don’t even know yet whether controlled fusion can be achieved to make useful work. |
| SR019 | MIT News | Will neutrons compromise superconducting magnets in a fusion plant? | |
| SR020 | ITER Organization | Fusion fuels | |
| SR021 | Commonwealth Fusion Systems | Commonwealth Fusion Systems raises $863 million Series B2 round to accelerate the commercialization of fusion energy | |
| SR022 | Helion Energy | Articles and updates | |
| SR023 | Fusion Industry Association | Fusion Industry Reports | |
| SR024 | Fusion Industry Association | The Fusion Industry Supply Chain 2026 | Largest future concerns are fuel infrastructure and first wall materials. |
| SR025 | Noerr | Fusion power in Germany and the EU | IP and protection strategies must already be outlined at the application stage; they constitute a relevant evaluation criterion. |
| SR026 | CMS Law | The future for fusion: Government consults on National Policy Statement EN-8 for nuclear fusion energy | |
| SR027 | Foley Hoag | Fusion Update: NRC Publishes Proposed Regulatory Framework For Fusion Machines | Fusion waste would either need to fit within an existing waste-class description or go to a disposal site that has completed a site-specific intrusion assessment. |
| SR028 | ANS Nuclear Newswire | New GAO report focuses on fusion | |
| SR029 | Hogan Lovells | New GAO Report highlights advancements in fusion, challenges, and provides policy recommendations | |
| SR030 | Max Planck Institute for Plasma Physics | Framework agreement for Alpha and Stellaris | |
| SR031 | World Nuclear News | Agreement to build German fusion power plant | |
| SR032 | NucNet | Google And RWE Back Germany’s Proxima In Bid For Europe’s First Nuclear Fusion Plant | |
| SR033 | U.S. Department of Energy | DOE National Laboratory makes history by achieving fusion ignition | |
| SR034 | International Atomic Energy Agency | International Experience in the Regulation of Fusion Facilities | |
| SV001 | Proxima Fusion | Proxima Fusion Raises €411 Million to Build Europe’s Commercial Fusion Champion | Proxima Fusion today announced a €411 million ($468 million) financing round, bringing the company’s valuation to €2.4 billion ($2.7 billion). |
| SV002 | BusinessWire | Proxima Fusion Raises €411 Million at a €2.4B Valuation to Build Europe’s Commercial Fusion Champion | Proxima Fusion today announced a €411 million ($468 million) financing round, bringing the company’s valuation to €2.4 billion ($2.7 billion). |
| SV003 | CNBC | Google backs nuclear fusion startup targeting Europe’s first commercial power plant | Google backs nuclear fusion startup targeting Europe’s first commercial power plant. |
| SV004 | RWE | RWE invests 25 million euros in fusion technology start-up Proxima Fusion | RWE invests 25 million euros in fusion technology start-up Proxima Fusion. |
| SV005 | Yahoo Finance | Germany's Proxima Fusion secures $2.7B valuation as Google backs Europe's reactor race | Energy startup Proxima Fusion has raised €411 million from an investor group that includes Google and German energy giant RWE. |
| SV006 | Fusion Future | Proxima Fusion funding round hits €411 million | Proxima’s new capital is meant to industrialize exactly this kind of complexity, moving coil production from single prototypes to Alpha-scale volumes. |
| SV007 | EU-Startups | Largest European fusion investment on record sees Proxima Fusion raise €411 million | Munich-based Proxima Fusion today announced a €411 million financing round, bringing the company’s valuation to €2.4 billion. |
| SV008 | Invezz | Google, RWE back $469M funding round for Proxima Fusion | Google, RWE back $469M funding round for Proxima Fusion. |
| SV009 | Neutron Bytes | Proxima Funding Round Targets 2030s for Fusion Power | Proxima Fusion Funding Fuels 2030s Date for Power Plant. |
| SV010 | North Data | Proxima Fusion GmbH, Munich, Germany, District Court of Munich HRB 283423: Network, Financial information | Proxima Fusion GmbH, Munich, Germany, District Court of Munich HRB 283423. |
| SV011 | North Data | Proxima Fusion GmbH, München, Amtsgericht München HRB 283423: Netzwerk, Wirtschaftsinfos | Proxima Fusion GmbH, München, Amtsgericht München HRB 283423. |
| SV012 | Online Handelsregister | Handelsregisterauszug von Proxima Fusion GmbH aus München (HRB 283423) | Handelsregisterauszug von Proxima Fusion GmbH aus München (HRB 283423). |
| SV013 | Fusion Industry Association | FIA Launches 2026 Fusion Industry Supply Chain Report | Supply chain spending by the fusion industry increased by 24% in 2025, according to The Fusion Industry Supply Chain 2026. |
| SV014 | The Fusion Report | Fusion Funding | Explore the latest insights on fusion energy funding. |
| SV015 | Clean Energy Platform | Fusion Industry Investment Trends 2026: Where the Money Is Going | After crossing the $15 billion cumulative investment milestone in late 2025, the fusion industry entered 2026 with a fundamentally different capital structure. |
| SV016 | TechCrunch | Every fusion startup that has raised over $100M | Fusion startups have raised $7.1 billion to date, with the majority of it going to a handful of companies. |
| SV017 | TechCrunch | Cracks are starting to form on fusion energy’s funding boom | Most of those I spoke to were worried these companies were going public far too early and that they hadn’t achieved key milestones. |
| SV018 | MIT Technology Review | Will fusion power get cheap? Don’t count on it. | Fusion power could provide clean electricity, but only if the industry can get its economics to work. |
| SV019 | Bulletin of the Atomic Scientists | What’s fueling the commercial fusion hype? | Despite fusion power being a big-bet technology, there is no US government agency in charge of assessing its risks and benefits. |
| SV020 | Commonwealth Fusion Systems | Commonwealth Fusion Systems Raises $863 Million Series B2 Round to Accelerate the Commercialization of Fusion Energy | CFS raised $863 million in a Series B2 fundraising round. |
| SV021 | TechCrunch | Nvidia, Google, and Bill Gates help Commonwealth Fusion Systems raise $863M | Commonwealth Fusion Systems has raised nearly $3 billion to commercialize its fusion power plant. |
| SV022 | Sacra | Commonwealth Fusion Systems funding, news & analysis | With the latest funding, CFS has raised more capital than any other private fusion company globally. |
| SV023 | Helion Energy | Helion Raises $465 Million Series G Funding Round to Meet Surging Global Demand for Power | Led by Thrive Capital at $15.5 billion post-money valuation. |
| SV024 | GeekWire | Helion hits $15.5B valuation with $465M in new cash as it aims to commercialize fusion this decade | Fusion startup Helion raised $465M, boosting its valuation to $15.5B. |
| SV025 | TAE Technologies | TAE Technologies raises $150 million in latest funding round | TAE Technologies has raised more than $150 million in its latest funding round. |
| SV026 | Tracxn | TAE Technologies | TAE Technologies has raised a total funding of $1.32B over 11 rounds. |
| SV027 | Tokamak Energy | Tokamak Energy raises $125m to commercialise transformative fusion and magnet technologies | Tokamak Energy has raised $125 million to accelerate ambitious plans to commercialise fusion energy. |
| SV028 | Tracxn | Tokamak Energy | Tokamak Energy is a Series C company. |
| SV029 | CB Insights | Tokamak Energy Stock Price, Funding, Valuation, Revenue & Financial Statements | Tokamak Energy has raised $226.5M over 10 rounds. |
| SV030 | Fusion Industry Association | The Global Fusion Industry | 53 verified private fusion companies of diverse technologies; $9.76 billion in funding. |
| SV031 | Proxima Fusion | Proxima Fusion, RWE, the Free State of Bavaria and Max Planck Institute for Plasma Physics sign agreement to build the world’s first commercial fusion power plant in Europe | Proxima has signed an agreement with the Free State of Bavaria, RWE, and Max Planck Institute for Plasma Physics to put the world’s first commercial stellarator fusion power plant on the grid in Europe. |