Radiant Nuclear
面向国防与 AI 基础设施的便携式微反应堆
Radiant Nuclear 已凭公开里程碑证据跻身最可信的私有微堆开发商之一;但在零公开收入、燃料和许可关卡层层叠加、客户验证集中之下,当前 >$1.8B 估值信号仍显过早。
封面要素
公司概况
Radiant Nuclear 是一家位于加州 El Segundo 的先进核能初创公司,由 CEO Doug Bernauer 于 2020 年创立,目标是打造 Kaleidos:一款集装箱化、约 1 MWe 的微反应堆,使用 TRISO 燃料、氦冷却剂和干式冷却,替代国防基地、数据中心和偏远地点的柴油发电。公开证据显示,Radiant 在 2025-2026 年商业化势头异常强:先完成 $165M Series C,再完成 >$300M Series D,公开累计融资超过 $525M,Equinix 为 20 台反应堆支付定金,取得 DOE DOME 测试资格和 DARK 批准,入选 Buckley Space Force Base,并且 Oak Ridge R-50 工厂装料申请已获 NRC 受理审查。即便进展明显,Radiant 仍处于商业化前阶段,且私营披露有限:没有公开收入、毛利率、烧钱速度、现金跑道、员工数或完整治理信息。
- 成立时间
- 2020-01-01
- 创始人
- Doug Bernauer
- 创立地点
- El Segundo, California, USA
- 总部
- El Segundo, California, USA
- 产品
- Kaleidos 是一款可运输、约 1.0-1.2 MWe 的高温气冷微反应堆,采用 TRISO 燃料、氦冷却剂、棱柱石墨块和被动空气冷却,并封装在集装箱化产品中;Radiant 计划在工厂完成装料、服务和发运,为现场韧性供电。
- 客户
- 主要目标客户是国防设施、数据中心运营商,以及需要韧性柴油替代电力、并能通过长周期政府项目或企业能源决策采购的偏远或离网工业与关键基础设施地点。
- 商业模式
- Radiant 计划通过直接销售机组和购电协议变现 Kaleidos;目前早期牵引力来自有定金支持的预购,而不是已披露的经营收入。
- 阶段
- Series C/D
- 融资情况
- 2025 年 5 月 Series C:$165M,已披露累计融资升至 $225M;2025 年 12 月 Series D:由 Draper Associates 和 Boost VC 领投,金额 >$300M,意味着公开融资下限超过 $525M。DCVC 随后称该轮对 Radiant 的估值超过 $1.8B,但该估值来自投资方披露,而非监管文件背书。
执行摘要
主要优势
- 2025-2026 里程碑密度异常高:公共资金底线超过 $525M,拿到 DOME 准入、DARK 批准、Buckley 入选,以及 Part 70 审查受理
- Kaleidos 围绕可运输的约 1 MWe 干冷部署做差异化,瞄准国防、数据中心和偏远地区韧性电力场景
- Equinix 为 20 台机组支付订金,是公开记录里最清楚的商业付费意愿信号
- Rita Baranwal 加入,加上 2025 年更广泛的领导层补强,显著提高了监管和执行可信度
主要风险
- HALEU 和更广泛 TRISO 供应仍是桥接市场,而不是已验证的舰队级投入
- Radiant 仍要跨过多个独立关口:带燃料 DOME 表现、场址 / NEPA 工作,以及 Part 70 之外的独立反应堆路径
- 客户验证集中在一个具名商业买家和一条国防路径上,公开材料没有披露定价或合同价值
- >$1.8B 标记来自投资人 / 媒体报道,没有公开收入、利润率、烧钱速度或治理披露支撑
未决问题
- 当前收入、ARR、毛利率、月度烧钱速度、现金跑道和已确认现金转化仍未披露
- Equinix 和 Buckley 的订金条款、取消权、定价和部署时间表未公开
- 公开材料没有披露完整董事会构成、投资人控制权、所有权或清算优先权
- 成功带燃料测试、稳定 HALEU 供应和单堆单位经济的证据仍然缺失
目录
01公司概况
1.1 身份、产品与商业定位
Radiant Nuclear 是一家位于加州 El Segundo 的私营初创公司,成立于 2020 年,目标是批量生产便携式核微反应堆,在燃料物流困难或电网可靠性差的地点替代柴油发电。旗舰产品 Kaleidos 在公司材料中一贯被描述为 1 MWe 微反应堆,采用 TRISO 燃料、氦气冷却剂、被动空气冷却和可用卡车或飞机运输的集装箱形态。产品卖点明确不是大电网发电,而是为军事设施、数据中心、偏远工业、医院、应急响应和偏远社区提供有韧性的现场电力。Radiant 称 Kaleidos 可在补充燃料前运行五年或更久,总计可补充燃料四次,寿命 20 年,并可通过直接销售机组或购电协议销售。路线图仍然激进:2026 年完成首次带燃料 DOME 测试,2028 年开始首批客户部署。这些属性拼出差异化的便携性和不依赖用水叙事,但公开证据仍未到运行中反应堆或商业收入这一步。 [CO001, CO002, CO003, CO004, CO005, CO006]
| 指标 | 数值 / 状态 | 日期 | 置信度 | 缺口 / 注意事项 |
|---|---|---|---|---|
| 成立 | 2020 | 2020 | 高 | 稳定历史事实 |
| 总部 | El Segundo, California | 2026-05-21 | 高 | 仅公开总部;更广站点布局未充分披露 |
| 核心产品 | Kaleidos 便携式核微反应堆 | 2026-05-21 | 高 | 尚无运行中的客户机组 |
| 反应堆架构 | 1 MWe;TRISO 燃料;氦气冷却剂;被动空气冷却;现场无需用水 | 2026-05-21 | 高 | 基于加燃料测试前的公司和行业描述 |
| 燃料周期 / 寿命 | 换料前 5+ 年;20 年寿命 | 2026-05-21 | 高 | 现场表现尚未验证 |
| 首次加燃料测试 | Idaho National Laboratory 的 DOME,目标 2026 年 | 2026-02-09 | 高 | 目标日期取决于最终启动授权 |
| 初始部署 | 2028 年开始 | 2026-04-22 | 高 | 取决于测试、燃料、制造和许可进展 |
| 已披露公开融资 | 累计 >$525 million | 2025-12-17 | 高 | 仅来自公司公告融资轮 |
| 估值信号 | >$1.8 billion | 2025-12-17 | 中 | 由 DCVC 投资人标记;未获监管文件或中立市场来源确认 |
| 具名商业订单 | Equinix 预订单,20 台反应堆并有押金 | 2025-08-14 | 高 | 定价、站点清单和交付时间表未披露 |
| 具名防务路径 | Buckley Space Force Base 入选;首批反应堆目标 2028 年 | 2026-04-22 | 高 | 项目仍取决于选址、环境审查和执行 |
| 收入 / ARR | 未公开披露 | 2026-05-21 | 高 | 需要管理层财务资料包或数据室材料 |
| 员工数 | 未公开披露 | 2026-05-21 | 高 | 需要组织架构图、HR KPI 包或尽调访问权限 |
混合了稳定公司事实、融资披露和明确公开数据缺口;估值只是方向性,因为 >$1.8B 数字来自投资人帖子,而不是监管文件或中立市场报告。
[CO001, CO002, CO003, CO004, CO006, CO007]Radiant 在商业化路径中如何串联产品架构、资本、政府测试基础设施、已点名需求和燃料依赖。
[CO003, CO004, CO005, CO006, CO019, CO020]截至 2026-05-21 运行日,Radiant 的核心成熟度、牵引力和风险指标。
[CO006, CO016, CO019, CO020, CO022, CO026]1.2 领导层、创始人匹配度与治理可见性
Radiant 的对外叙事异常围绕创始人展开。Doug Bernauer 仍是公司最主要的公众面孔,投资方和合作伙伴页面也反复把他定位为 Kaleidos 项目的工程与执行锚点。2025-2026 年公开管理团队显著增强,部分抵消了这种集中度。Rita Baranwal 曾任 DOE 核能助理部长,之后在 Westinghouse 担任高管,2025 年 6 月加入 Radiant,成为公司首任 Chief Nuclear Officer,为监管和核运营增加了重量级可信度。到 2025 年末和 2026 年初,Mike Starrett 已公开担任 Chief Revenue Officer,Tori Shivanandan 已公开担任 Chief Operating Officer,Radiant 还称在工程、制造和供应链新增了近十名 VP 与总监级领导。即便如此,治理透明度仍然有限。已审阅的公开来源没有披露完整董事会名单、委员会结构、投资人席位分配或控制权安排;尽管领导层建设明显推进,关键人和治理尽调仍需保持较高权重。 [CO008, CO009, CO010, CO011, CO012, CO013]
| 人物 | 职务 | 背景 | 创始人 / 覆盖范围 | 关键人物依赖 |
|---|---|---|---|---|
| Doug Bernauer | CEO 兼创始人 | 前 SpaceX 工程师;2020 年创立 Radiant,用航空航天式执行力做便携式反应堆 | 创始人与市场匹配,体现在制造、便携性和技术叙事 | 关键 — 融资、工程可信度和外部叙事仍以创始人为中心 |
| Rita Baranwal | 首席核能官 | 前 DOE 核能助理部长;前 Westinghouse CTO 和 AP300 SVP | 补足核监管深度和运营可信度 | 高 — 安全、许可和启动准备的关键接口 |
| Mike Starrett | 首席营收官 | Equinix 和 Buckley 公告中的公开商业与防务发言人 | 负责把市场从试点转化为客户部署 | 中 — 商业证明仍集中在少数具名交易 |
| Tori Shivanandan | 首席运营官 | 在 Lockheed 战略投资和商业化进展中被引用发言 | 覆盖运营和规模化执行 | 中 — 运营角色重要,但公开披露仍较少 |
公开高管证据对 Bernauer 和 Baranwal 最强。已审阅来源未披露完整董事会名单、委员会或投资人席位分配,因此治理覆盖仍不完整。
[CO008, CO009, CO010, CO011, CO012, CO013]1.3 资本形成、利益相关方与具名需求信号
2025 年,Radiant 从一家有风险投资支持的概念公司,变成重融资的商业化故事。公司在 2025 年 5 月完成 $165M Series C,使公开累计融资达到 $225M;随后在 2025 年 12 月宣布由 Draper Associates 和 Boost VC 领投的新一轮融资,金额超过 $300M。这些披露意味着,到 2025 年底,公开风险投资资金超过 $525M。DCVC 还称 12 月这一轮对 Radiant 的估值超过 $1.8B,但该估值来自既有投资人,而不是监管文件或中立市场参照,因此最好把它当作方向性信号,而不是充分交叉验证的公允价值锚点。投资方如今横跨财务 VC 和战略方,包括 DCVC、Draper Associates、Boost VC、Andreessen Horowitz、Founders Fund、Union Square Ventures、Chevron Technology Ventures 和 Lockheed Martin Ventures。最强的公开需求证据同时带有商业和国防属性:Equinix 为 20 台反应堆支付定金,Department of the Air Force/DIU 选择 Radiant 用于 Buckley Space Force Base。公开视野里仍缺的是定价、积压订单转化、反应堆利润率,以及具名交易方之外的积压订单。 [CO014, CO015, CO016, CO017, CO018, CO019]
| 利益相关方 | 角色 | 控制 / 经济重要性 | 公开证明 | 尽调问题 |
|---|---|---|---|---|
| DCVC | Series C 轮领投方;Series D 轮参与方 | 主要成长资本支持方和公开估值标记来源 | 领投 $165M Series C;后称 2025 年 12 月融资轮给 Radiant 估值超过 $1.8B | 持股比例、董事会席位、按比例跟投权,以及 DCVC 的估值标记是否匹配定价轮 |
| Draper Associates | 2025 年 12 月领投方 | 共同领投 >$300M 融资轮,为商业化和工厂计划提供资金 | 在 Radiant 2025 年 12 月融资公告中被列为共同领投方 | 融资轮条款、优先股权利和预期后续支持 |
| Boost VC | 2025 年 12 月领投方 | 共同领投最新一轮,也出现在过往投资人名单中 | 在 2025 年 12 月公告中被列为共同领投方 | 董事会权利、持股比例和早期各轮投资历史 |
| Lockheed Martin Ventures | 战略投资人 | 补强防务验证,并可能带来战略渠道价值 | 2026 年 2 月宣布战略投资 | 任何商业合作、排他性或技术开发权利 |
| Equinix | 商业预订单客户 | 最强具名商业需求信号;据报道已为 20 台机组提交押金 | 2025 年 8 月预订单和押金公告 | 定价、里程碑、取消权、部署站点和先决条件 |
| Department of the Air Force / DIU 政府路径 | 政府部署发起方 | 创造首条公开的军事基地部署路径 | 2026 年 4 月宣布 Buckley Space Force Base 入选 | 合同金额、NEPA 路径、运营方责任划分和联邦退出条款 |
| DOE / NRIC / INL | 测试基础设施和燃料交易对手 | DOME 准入和 HALEU 支持是卡点式外部依赖 | DOME 入选加联邦 HALEU 分配流程 | 准确燃料合同条款、里程碑要求和进度滑期补救 |
| Andreessen Horowitz / Founders Fund / Union Square Ventures 投资人群体 | 现有财务支持方 | 显示来自防务和前沿科技导向风投的持续支持 | 出现在公司投资人名单和合伙人页面 | 当前持股、清算优先权、二级交易活动和任何董事会观察员权利 |
仅为部分公开利益相关方图谱。公开材料识别了许多投资人和交易对手,但没有披露持股比例、清算优先权或完整治理权利。
[CO014, CO015, CO016, CO017, CO018, CO020]1.4 执行里程碑与监管路径
Radiant 在 2025-2026 年的推进节奏重要,因为核能初创公司通常受制于测试基础设施、许可和燃料采购,而不只是融资。DOE 于 2025 年 7 月有条件选择 Radiant 参加 Idaho National Laboratory 首个 DOME 微反应堆测试活动,政府 DOME 材料确认该场地是全球首个专门用于不超过 20 MWt 带燃料实验的微反应堆测试床。2026 年 2 月,DOE 批准 Radiant 的 DARK 提交材料;公司称 DARK 满足 Preliminary Documented Safety Analysis 的意图,并且是启动前所需三阶段安全文件中的第二阶段。2026 年 4 月,Air Force 和 DIU 选择 Radiant 进入 Buckley Space Force Base;2026 年 5 月,NRC 正式受理 Radiant 为 Tennessee R-50 生产设施提交的 Part 70 许可申请。Radiant 还称 Oak Ridge 工厂计划扩大到每年 50 台反应堆。这些都是实质执行证据,但仍属于收入前、启动前、许可完成前的证据。 [CO007, CO022, CO023, CO024, CO025, CO026]
| 日期 | 事件 | 类型 | 金额 / 估值 / 状态 | 参与方 | 含义 |
|---|---|---|---|---|---|
| 2020 | Radiant 在 El Segundo 成立,目标是建造便携式微反应堆 | 成立 | 公司成立 | Doug Bernauer 和早期团队 | 确立创始人主导的使命和商业化投资逻辑 |
| 2025-05-28 | Series C 轮完成 | 融资 | $165M;公开累计融资达到 $225M | Radiant、DCVC、新投资人 | 为 Kaleidos Development Unit 完成和工厂选址工作提供资金 |
| 2025-06-03 | Rita Baranwal 加入,担任首任首席核能官 | 治理 | 高管任命 | Radiant、Rita Baranwal | 为安全和许可工作补足 DOE 与 Westinghouse 可信度 |
| 2025-07-02 | DOE 选择 Radiant 参加首个 DOME 测试活动 | 监管 | 有条件入选;引用 2026 年春季测试启动 | DOE、NRIC、INL、Radiant | 获得用于加燃料演示的旗舰测试基础设施 |
| 2025-08 | HALEU 分配流程凸显燃料依赖 | 反向 | Radiant 位列五家有条件获选方之一;国内 HALEU 仍稀缺 | DOE、Radiant、其他获选方 | 近期测试燃料改善,但结构性燃料风险仍在 |
| 2025-08-14 | Equinix 预订单公布 | 合作 | 20 台反应堆加押金 | Equinix、Radiant | 迄今最强具名商业需求证明 |
| 2025-12-17 | 新融资轮公布 | 融资 | >$300M 新资本;DCVC 后来引用 >$1.8B 估值 | Radiant、Draper Associates、Boost VC、现有投资人 | 支持商业化和工厂建设 |
| 2026-02-09 | DARK 获批 / PDSA 意向里程碑完成 | 监管 | 三项核安全文件阶段中的第二项完成 | DOE、INL、Radiant | 维持 2026 年夏季启动路径 |
| 2026-02-17 | Lockheed Martin Ventures 进行战略投资 | 合作 | 超额认购融资轮扩大 | Lockheed Martin Ventures、Radiant | 补强防务生态验证和战略信号 |
| 2026-04-22 | Buckley Space Force Base 入选 | 合作 | 首批反应堆目标 2028 年 | Department of the Air Force、DIU、Radiant 等参与方 | 创造首条公开军事基地部署路径 |
| 2026-05-05 | NRC 受理 R-50 设施 Part 70 许可审查 | 监管 | 申请已受理;加速审查目标公布 | NRC、Radiant | 迈向 Tennessee 燃料处理生产能力的关键一步 |
作为已审阅来源中浮现的重大公开里程碑唯一时间线,覆盖从成立到 2026 年 5 月 NRC 受理事件。
[CO001, CO009, CO014, CO015, CO018, CO020]Radiant 从 2020 年成立,到融资、DOME 准入、HALEU 依赖,再到 2026 年 Buckley / NRC 里程碑的公开路径。
[CO001, CO009, CO014, CO015, CO016, CO018]1.5 反向因素与剩余尽调缺口
里程碑密度高,不等于商业化已经充分去风险。Radiant 仍未公开披露当前收入、ARR、员工数、完整董事会构成、投资人控制权、准确积压订单、单机定价或毛利率。公开证据点名了需求信号,但没有独立核实合作伙伴页面上“Radiant 到 2030 年几乎售罄”的说法,也没有量化 Equinix 预购和 Buckley 路径之外的积压订单。另一项主要依赖是燃料。DOE 的 HALEU 分配有助于近期测试,但 Utility Dive 报道称美国国内 HALEU 市场仍面临基础设施缺口,UCS 也强调一项与 Science 相关的分析,认为 U-235 丰度高于约 12% 的 HALEU 可能带来比政府和行业公开承认更严重的扩散和恐怖主义风险。落到尽调实践,Radiant 接下来真正硬的证明点不是更多里程碑新闻稿,而是成功的带燃料测试、更清晰的燃料合同和商业指标披露。 [CO029, CO030, CO031, CO032, CO033, CO034]
1.6 图表
02市场分析
2.1 市场边界与受证据约束的规模测算
Radiant 的相关市场不是“整个核能”,甚至也不是整个小型模块化反应堆板块。公开来源一致把微反应堆描述为非常小、可运输、面向去中心化和非常规市场的反应堆;在这些场景里,1 MWe 级资产可以解决单个站点的可靠性或物流问题。对 Radiant 来说,这意味着电网边缘的韧性供电:军事设施、数据中心、偏远工业、备用关键基础设施和孤立社区。它不意味着大电网基荷发电、公用事业机组替换,也不意味着宽泛 300 MWe 及以下 SMR 预测所暗示的全部收入池。 因此,最佳规模测算方法应是多镜头,而不是单一 TAM。DOE/LBNL 和 Belfer 为数据中心提供了可信需求视角,Equinix、Project Pele、Janus 和 Eielson 则为早期采用者提供了具体细分证据。相比之下,供应商版 SMR 市场报告只能作为上限背景,因为它们把远超 Radiant 直接范围的反应堆类别和电网级应用打包在一起。实际结论是:Radiant 的可服务市场(SAM)比通用 SMR TAM 材料暗示的更窄,但也更具体;它位于柴油、弱电网或并网瓶颈让便携式清洁稳定电力在经济或运营上值得买方认真投入的地点。 [CM001, CM002, CM003, CM004, CM013, CM030]
| 细分 / 类别 | 纳入支出 | 排除支出 | 主要买方 / 付款方 | 对 Radiant 的意义 |
|---|---|---|---|---|
| 军事 / DoD 基地电力 | 1-5 MWe 级站点电力、韧性微电网、可运输反应堆硬件、燃料、运营和站点集成 | 大电网发电、海军推进、战术电池,或与基地韧性无关的中央公用事业发电 | 军种 / 基地能源发起方 | 直接核心细分 |
| 数据中心现场 / 表后电力 | 园区电力资产、长期电力合同、站点集成,以及面向数字基础设施的韧性和清洁稳定供应 | 通用 IT 资本开支、无关电网建设,或运营方无法控制的商用批发发电 | 数据中心运营商 / 能源采购职能 | 直接核心细分 |
| 偏远 / 离网工业和采矿电力 | 矿山或工业站点电力和热力、柴油替代、物流减少和微电网集成 | 区域输电建设或大型公用事业级发电 | 工业运营商 / 项目业主 | 直接核心细分 |
| 备用 / 关键基础设施 | 医院、供水系统、海水淡化、灾害响应和其他韧性驱动的微电网应用 | 单独传统备用柴油,或不绑定站点级韧性的公用事业调峰电厂 | 设施业主 / 公共机构 / 韧性发起方 | 相邻但技术上匹配 |
| 偏远社区 / 孤立电网 | 社区微电网、村庄供电系统,以及输配电成本很高的地点 | 大型区域电网加固或完整公用事业机组替换 | 社区公用事业 / 公共发起方 | 有吸引力的长期细分,但政策权重高 |
| 广义 SMR / 电网级模块化反应堆 | >20 MWe 至 300+ MWe 模块化反应堆和多机组基荷电站 | 公用事业公司 / 国家核能项目 | 相邻 TAM,不是 Radiant 直接 SAM |
围绕站点级韧性电力支出定义直接市场,而不是整个核能或 SMR 行业。纳入和排除支出由分析师基于官方用例描述和采购证据构建。
[CM001, CM002, CM003, CM004, CM013, CM029]| 发布方 / 视角 | 年份 | 地区 | 指标 | 数值 | 方法 | 置信度 | 局限 |
|---|---|---|---|---|---|---|---|
| DOE / LBNL 数据中心需求视角 | 2028 | 美国 | 数据中心电力需求 | 325-580 TWh(6.7%-12% 份额) | DOE 引用的自下而上电力需求预测 | 高 | 衡量的是电力需求,不是直接微反应堆收入 |
| Belfer / 机构情景区间 | 2030 | 美国 | 机构数据中心需求情景 | ~200 TWh 至 >1,000 TWh | 政策简报汇总机构来源引用的宽区间预测 | 中 | 情景区间很宽;不是站点级或反应堆级测算 |
| Equinix 核电需求证据 | 2025 | 全球资产组合 | 已公告协议覆盖的先进核电容量 | >774 MWe 合计;Radiant 部分为 20 x 1.2 MWe | 三家开发商已公告的商业协议 | 中 | 容量并非全是微反应堆,投运时间也在较远期 |
| EIA / 军方选址视角 | 2025-2026 | 美国 | 已点名的军事选址机会 | 9 个陆军候选基地 + Eielson 1-5 MWe 试点 + Buckley 路径 | 公开的联邦项目和选址公告 | 中 | 站点数量不等于已签合同或获得经费的部署规模 |
| Precedence Research | 2026-2035 | 全球 | 广义 SMR 市场收入 | $8.16B(2026)至 $17.37B(2035);8.78% CAGR | 厂商市场预测 | 中 | 涵盖的反应堆类别和用例远宽于 Radiant 的直接市场 |
| MarketsandMarkets | 2024-2030 | 全球 | 广义 SMR 市场收入 | 到 2030 年 $7.14B;3.0% CAGR | 厂商市场预测 | 中 | 方法和范围与其他广义 SMR 预测差异很大 |
本表刻意混合收入、电力需求和已点名站点三个视角,因为已审阅公开来源没有单独拆出 1 MWe 可运输微反应堆的可服务市场(SAM)。表格用于在证据受限下估算规模,不是给出虚假的单一 TAM。
[CM006, CM008, CM031, CM032, CM033, CM034]公开审阅的美国数据中心用电需求区间显示,电力稀缺确实会拉动需求,但规划区间依然很宽。
2028 年和 2030 年中点是派生的可视化规划锚点,不是独立发布的中心情景。图中仅用来说明规划区间仍然很宽。
[CM008, CM010, CM036]2.2 买方细分、预算所有者与采购路径
买方地图更多按运营场景切分,而不是按反应堆技术切分。在国防场景,相关客户是 Army、Air Force 或更广泛联邦设施系统内的设施能源或韧性负责人;Project Pele、Janus、Eielson 和 Buckley 都指向这条渠道。在数据中心场景,买方更可能是运营商的能源或采购职能,而不是租户或应用所有者,因为公开证据围绕组合电力战略、首席采购监督和长期电力合同展开。偏远工业、社区和应急响应用例则落在第三类;痛点通常是燃料物流、薄弱输电或高停电成本,而不是纯粹的 AI 增长。 Radiant 的商业模式重要,因为它既支持直接销售机组,也支持类似购电的结构。这种灵活性让付款方因细分市场而变:军种、数据中心运营商、矿主或公共出资方,都可能在不同合同逻辑下为资产付费。Equinix 是最强的公开商业证据,因为它把兴趣转化为带定金的预购;但即便如此,公开记录仍没有完整披露站点经济性、单机定价或按园区展开的公开部署时间表。 [CM005, CM006, CM007, CM026, CM027, CM028]
| 细分市场 | 主要买方 | 主要用户 | 付款方 | 工作流 / 采购路径 | 预算所有者 | 采用触发因素 |
|---|---|---|---|---|---|---|
| DoD 基地供电 | 陆军 / 空军 / 基地韧性赞助方 | 基地运营和任务关键负载 | 联邦拨款和项目预算 | 试点遴选 -> 选址 -> 安全审查 -> 服务或部署合同 | 基地 / 能源与环境负责人 | 电网中断风险和燃料物流负担 |
| 数据中心 | 运营商能源或采购团队 | 园区运营;托管客户间接受益 | 运营商资产负债表、PPA、电价或其他直接电力合同 | 资产组合能源策略 -> 预购 / PPA -> 园区选址和集成 | 首席采购官 / 基础设施负责人 | AI 负载增长、电网稀缺和清洁稳定电力需求 |
| 偏远 / 离网工业 | 矿山或工业站点所有者 | 工厂运营 | 项目资本开支或运营预算 | 可行性 -> 物流与经济性研究 -> 服务或资产采购 | 站点总经理 / 能源经理 | 柴油替代、可靠性和可选工艺热 |
| 备份 / 关键基础设施 | 医院、供水或应急响应赞助方 | 设施运营方 | 公共或私人韧性预算 | 连续性规划 -> 安全审查 -> 长期运营模型 | 设施 / 连续性负责人 | 停电风险和灾难恢复价值 |
| 偏远社区 / 孤立微电网 | 社区公用事业或公共机构 | 公用事业运营方和居民 | 公用事业费率加公共补贴或赠款支持 | 政策背书 -> 社区同意 -> 运营方模型 | 市政公用事业 / 州级赞助方 | 输电成本高、柴油供应昂贵 |
买方、用户和付款方角色,部分根据公开项目结构和电力采购表述推断。公开证据在国防和数据中心细分市场最强; 社区和工业场景的预算归属仍不够透明。
[CM005, CM006, CM011, CM026, CM027, CM028]市场会通过不同买方渠道落地,但所有路径仍要穿过供应商能力、燃料 / 测试基础设施和监管许可。
[CM018, CM026, CM027, CM028, CM029, CM043]2.3 需求驱动因素
最可见的近期驱动因素,是 AI 和超大规模数据中心扩张带来的用电问题。DOE 和 Belfer 都提到美国数据中心用电需求快速上升,2028 年用电量可能远高于 2023 年水平,一些地区已经出现电力稀缺、项目延期或围绕电网的直接合同。这不会自动让微反应堆成为胜出方案,但确实给了数据中心运营商一个可信理由:评估稳定的现场电力,而不是只依赖常规公用事业时间表。 不过,Radiant 的市场不只是 AI 故事。官方微反应堆项目材料反复强调柴油替代、偏远社区、矿业或离网工业、关键基础设施韧性和国防行动。这些细分市场有吸引力,正因为它们不需要数百兆瓦;它们需要小型、可运输、低用水、长换料间隔的资产。脱碳有帮助,但更强的驱动往往是燃料交付成本与运营负担、停电风险或受限输电接入。因此,即便更广泛的核能行业增长缓慢,1 MWe 可运输反应堆也可能拥有真实利基。 [CM008, CM009, CM010, CM011, CM012, CM013]
| 驱动因素 / 约束 | 方向 | 时间 | 对 Radiant 的含义 | 尽调问题 |
|---|---|---|---|---|
| AI 和数据中心负载增长 | 驱动因素(很强) | 2024-2028 | 运营商有了真实理由,在常规公用事业时间表之外评估站内清洁稳定电力 | 识别哪些园区能实际容纳 1-20 MWe 级资产 |
| 国防能源韧性 | 驱动因素(强) | 当前至 2030 年 | 军事项目可能为可运输微反应堆带来最早的标杆客户 | 确认合同金额、拨款和重复采购路径 |
| 柴油替代和燃料物流 | 驱动因素(强) | 当前 | 燃料配送和停电风险都很贵时,偏远或离网站点最匹配 | 按细分市场量化可避免的燃料和物流成本 |
| 脱碳和清洁稳定电力 | 驱动因素(强) | 当前 | 支撑公共部门和数据中心需求,尤其适合买方希望低碳且可靠的场景 | 测试买方如何在核电、地热、燃气和储能方案之间比较 |
| 小电网和高输配电成本适配性 | 驱动因素(中) | 当前 | 让 Radiant 更适合偏远社区和孤立工业地点 | 绘制哪些目标站点无法支撑常规电网扩容 |
| 许可和审批周期 | 约束(重大) | 当前且持续 | 即便需求强劲,也会拖慢从兴趣到投运资产的转化 | 获取测试、选址和运营审批的预期里程碑日历 |
| HALEU 和 TRISO 供应链 | 约束(重大) | 2025-2028 | 买方有兴趣,燃料仍可能卡住部署节奏 | 核实联邦过渡性配额之外的商业燃料合同 |
| 客户信任和公众接受度 | 约束(重大) | 持续 | 安保和承载站点问题可能推迟社区及关键基础设施采用 | 按细分市场评估同意、安保和保险要求 |
| 电网 / 电价 / 所有权复杂度 | 约束(重大) | 持续 | 数据中心部署仍需要跑得通的土地、所有权和并网结构 | 明确每种模型下谁拥有反应堆、站点资产和并网权 |
| 成本竞争力不透明 | 约束(重大) | 持续 | Kaleidos 经济性没有公开数据,客户转化更难做底层测算 | 要求提供站点级经济性,并与柴油、燃气和电网延伸方案对比 |
将结构性需求驱动因素与最可能改变采用节奏的关口摩擦配对。最后两行尤其重要,因为已审阅公开材料里, 需求能见度高于成本能见度。
[CM008, CM010, CM012, CM015, CM022, CM023]2.4 约束、时间线风险与采用摩擦
核心市场约束不是缺少口头需求,而是兴趣转化为运行中资产太慢。DOE 自身微反应堆项目把经济可行性和许可就绪度列为闸门问题,EIA 指出漫长许可和审批历史上限制了核能扩张,Radiant 的 NRC 材料则显示工厂装料、材料许可和监管差距分析等多条工作流并行推进。DOME 具有战略重要性,因为它降低了技术和安全示范风险;但测试床不等于可商业重复的部署路径。 燃料又加上一层闸门。DOE 通过 HALEU 分配衔接近期需求,但 Utility Dive 仍把美国国内民用 HALEU 供应描述为受基础设施缺口制约,UCS 也强调了安全和扩散批评,可能抬高民用部署审查强度。客户采用也有自己的摩擦:数据中心买方必须解决所有权、电价和园区集成问题;军事和偏远站点客户则面对公共采购、信任、安全和主站点接纳问题。结果就是:市场拉力真实存在,但采用周期漫长且非线性。 [CM015, CM018, CM019, CM020, CM021, CM022]
| 细分市场 | 公开证据点 | 最早可信采用窗口 | 关口依赖 | 主要摩擦 | 含义 |
|---|---|---|---|---|---|
| 军事基地 | Project Pele、Janus、Eielson、Buckley 等基地项目 | 2020 年代末 | 联邦选址、安全审查和服务合同转化 | 试点项目必须转成重复采购 | 最可能成为早期标杆细分市场,但仍是项目制 |
| 数据中心 | Equinix 预购加更广泛的核电组合 | 2020 年代末 | 园区选址、公用事业协调和跑得通的合同结构 | 所有权和电价复杂度 | 需求真实存在,但成批部署仍处早期 |
| 偏远 / 离网工业 | DOE、GAIN 和 WNA 瞄准的用例 | 2020 年代末至 2030 年代初 | 站点经济性和物流验证 | 私营项目经济性未公开 | 柴油替代收益足够强时,吸引力可能较高 |
| 备份 / 关键基础设施 | DOE 用例包括医院和救灾 | 2030 年代初 | 安全、安保和公共采购接受度 | 政治和保险摩擦 | 可能慢于国防或超大规模试点 |
| 偏远社区 | GAIN、DOE、EIA 和 WNA 对偏远社区的引用 | 2030 年代初 | 社区同意,加补贴或公共赞助方支持 | 信任和可负担性 | 社会价值很强,但最难快速商业化 |
本表受证据约束,不是确定的销售预测。窗口是基于已审阅公开里程碑组合和各买方渠道成熟度做出的判断。
[CM017, CM029, CM035, CM036, CM038, CM041]漏斗顶部的需求可见度最高;最大流失发生在许可、燃料、签约和站点转化环节。
漏斗数值是序数型分析师估计,用来显示市场在哪些环节收窄;它们不是披露销售管线中的实证转化率。
[CM015, CM018, CM019, CM022, CM038, CM039]2.5 判断与未解市场缺口
最可复用的尽调判断是:Radiant 位于一个真实但仍处商业化前的市场。公开证据支持国防和数字基础设施买方存在有意义的需求拉力,但没有已审阅来源单独测算可运输 1 MWe 反应堆的独立 SAM,也没有已审阅公开来源为 Kaleidos 本身提供经过验证的相对柴油或相对电网成本基准。因此,宽泛 SMR 市场报告只能提供方向性背景;不能单独用于客户转化或估值逻辑。 后续客户与估值工作,正确做法是从证明点和转化障碍出发测算市场:候选站点有多少,哪些站点能承受核能交付周期,能够支持什么合同模式,以及燃料、许可和选址约束能多快清除。这样形成的市场逻辑比泛泛的“大核能 TAM”更窄,但更可信。它也保留了最大的未知数,而不是把问题抹平:经济性、积压订单转化时间,以及买方从预购或试点话术走向已安装运行机组的真实速度。 [CM030, CM031, CM032, CM033, CM034, CM035]
2.6 图表
03竞争格局
3.1 竞争集合与细分
Radiant 的正确竞争框架比“所有 SMR”更窄。本文审阅的行业资料把微反应堆视为 SMR 市场中低于 20 MW 的边缘段,优化方向是工厂制造、选址灵活性、偏远社区、离网工业、数据中心或微电网用例,而不是替代公用事业规模发电。在这个更大族群里,Radiant 已披露的产品包异常小,也异常围绕物流展开:官方和监管材料都把 Kaleidos 定位在一个约 1 MWe、用于替代柴油、大小接近海运集装箱的任务上。 这让 Westinghouse eVinci 处在最接近的直接赛道,因为它同样可运输、工厂组装,并瞄准偏远社区、采矿和数据中心。BWXT 的 Project Pele 背景也重要,因为它验证了军方对可运输反应堆的需求,哪怕它不是广泛商业产品,也可能影响国防采购路径。Oklo 和 NANO 只是部分重叠:Oklo 有 1-5 MWe 国防试点,也有大得多、面向超大规模云厂商的商业叙事;NANO 具备公开市场可见度,但概念范围更早、更宽。相比之下,Last Energy、NuScale、X-energy 和 Kairos 是相邻的分布式核能基准:它们带来融资、许可或工业热可信度,但公开标准产品从 20 MWe 起步,或明显高于这一水平。 替代品集合同样重要。对最小型任务而言,真正的在位方案仍是柴油加微电网或弱电网连接,而不是另一家反应堆供应商。对园区级负载而言,更大型反应堆 PPA、公用事业合同、燃气或电网延伸策略,都可能在不要求买方接受 1 MWe 便携式机组的情况下解决同一个可靠性问题。因此,当买方任务是机动性、低土建和长时离网韧性时,Radiant 才会赢;一旦买方需要工业蒸汽、很大的园区负载,或需要公开公司 / 公用事业支持平台的融资深度,Radiant 的相关性就会下降。[CP001, CP003, CP007, CP010, CP013, CP014]
| 竞争者 | 类别 | 规模 / 资本信号 | 目标细分市场 | 差异化 | 局限 |
|---|---|---|---|---|---|
| Radiant Kaleidos | 直接竞品:便携式微反应堆 | 已披露风险融资 $225m;Equinix 预购 20 台 | 偏远站点、国防、医院、数据中心备份 | 约 1 MWe 级集装箱式 HTGR;次日部署;工厂回收换料闭环 | 商业化前;公开口径 1.0 与 1.2 MWe 不一致;燃料供应仍是关口 |
| Westinghouse eVinci | 直接竞品:微反应堆 | Westinghouse 公司平台;首批 DOME 队列 | 偏远社区、采矿、数据中心 | 5 MWe 热管反应堆;完全工厂组装;8+ 年堆芯 | 对最小型柴油替代任务来说,比 Radiant 更大 |
| BWXT / Project Pele 参照情境 | 国防可运输微反应堆 | DoD 支持项目;首批燃料于 2025 年交付 | 军事发电 | 同类里与政府项目最贴合 | 证明的是国防采购路径,不是开放商业产品 |
| Oklo Aurora | 部分重叠微反应堆 | 公开申报公司;Eielson 试点;与 Meta 绑定的 1.2 GW 园区协议 | 国防试点、数据中心园区 | 商业所有权模型;超大规模云厂商拉动叙事强 | 钠冷;商业故事偏向园区级能源,不是集装箱式便携部署 |
| NANO Nuclear | 开发阶段直接 / 相邻同行 | 上市微反应堆公司(NASDAQ: NNE) | 便携式和固定式微反应堆;燃料物流 | 覆盖 KRONOS、ZEUS、LOKI 和 HALEU 物流的广泛组合 | 公开客户证据薄,已披露近期测试里程碑较少 |
| Last Energy PWR-20 | 相邻分布式核电 | 20 MWe 产品;全服务 IPP/PPA 模型 | 数据中心和工业本地基荷 | 工厂化模块建筑方案,配全服务交付 | 站点和负载假设远大于 Radiant |
| NuScale Power Module | 相邻 SMR | 上市公司;获美国批准的 SMR 模块 | 公用事业、园区、数据中心、工艺热 | 同组里美国许可和融资菜单最成熟 | 77 MWe 模块远高于便携式微电网用例 |
| X-energy Xe-100 | 相邻 HTGR | DOE ARDP 支持;Dow Long Mott 申报 | 重工业、蒸汽、大型园区 | 80 MWe TRISO / 氦冷 HTGR,提供工业蒸汽 | 工业级建设,不是便携部署 |
| Kairos KP-FHR | 相邻先进反应堆 | 3 张获批 NRC 建造许可证;Google 到 2035 年最高 500 MW | 电网、数据中心、工业用户 | 熔盐 / TRISO 设计,支持在线换料和模块化扩展 | 150 MWe 最小配置远高于 1 MWe 利基 |
| 柴油 + 微电网现状 | 既有替代方案 | 已安装既有基础设施 | 偏远和备份电力 | 短期最快可用,运营方基础成熟 | 燃料物流、排放和较低能量密度 |
| 电网延伸 / 燃气 / 公用事业 PPA | 既有替代方案 | 成熟融资和公用事业渠道 | 园区和工业站点 | 没有新型反应堆许可负担 | 前置周期、拥堵和较弱的离网韧性 |
规模 / 资本单元格混合了已披露融资、上市公司状态、政府项目支持和点名客户证据,因为各厂商的精确累计融资额 并没有可比的公开口径。
[CP006, CP010, CP013, CP014, CP015, CP018]基于证据的序数图,对比 Radiant 最相关的直接和相邻竞争对手在便携性适配与商业或监管成熟度上的位置。
[CP010, CP013, CP014, CP020, CP021, CP022]3.2 直接便携式同业
Radiant 对直接同业最强的公开论点从物流开始。它自己的材料强调卡车或飞机交付、无需场地开挖、次日启动、现场不用水,以及五年或更久之后回厂补充燃料的闭环。按这组证据,Radiant 最贴合“在受限站点替代柴油”的任务。Westinghouse eVinci 是最可信的已披露商业对手,因为它也提供工厂组装的可运输系统,但功率提高到 5 MWe,并宣传八年以上堆芯寿命。这让 eVinci 在单箱需要更大输出的站点更强,却不如 Radiant 严丝合缝地匹配最小偏远负载用例。 国防项目让格局更复杂。Project Pele 表明美国政府愿意资助可运输微反应堆开发,EIA 的 2026 年概览也显示,国防和联邦试点项目可以容纳 BWXT、Oklo、Radiant、Westinghouse、X-energy、Kairos 等多家供应商。这有利于品类验证,却不利于排他性。如果军事设施成为最早的真实买方,渠道权力可能掌握在政府项目和主承包商关系手中,而不是技术上最整洁的商业产品包手中。 Oklo 和 NANO 重要,但原因不同。Oklo 在 Eielson 有 1-5 MWe Aurora 试点,但其 2026 年最醒目的商业证据是与 Meta 签订的俄亥俄 1.2 GW 协议,更指向园区级清洁稳定供应,而不是海运集装箱式便携性。相比之下,NANO 已上市,并在 KRONOS、ZEUS、LOKI 和 HALEU 物流上野心很大,但已审阅公开材料读起来仍像开发阶段的组合搭建,而不是近期客户转化。Radiant 还需要注意一项额外风险:公开来源并不完全一致地描述 Kaleidos 的尺寸。公司和 NRC 材料指向约 1 MWe,而围绕 Equinix 的外部和合作伙伴导向报道描述为 1.2 MWe。这个差异不会推翻便携利基逻辑,但很重要,因为尺寸等级正是 Radiant 区别于 eVinci 和更大型相邻 SMR 的核心。[CP001, CP002, CP003, CP004, CP007, CP008]
| 采购标准 | Radiant | Westinghouse eVinci | Oklo Aurora | NANO Nuclear | Last Energy | NuScale | X-energy | Kairos |
|---|---|---|---|---|---|---|---|---|
| 已披露电功率 | ~1 MWe 级 | 5 MWe | 1-5 MWe 试点 | 概念组合;此处未披露单一主力商业额定功率 | 20 MWe | 每模块 77 MWe | 每单元 80 MWe | 最小电站 150 MWe |
| 可运输性 / 站点负担 | 集装箱;卡车或飞机运输;无需站点开挖 | 可通过铁路、驳船和卡车进行集装箱运输 | 商业试点;便携性不是 2026 年商业故事的核心 | 声称有便携式和固定式概念 | 工厂建造模块,但属于园区电站,不是便携箱体 | 工厂建造模块;大型电站土建范围仍在 | 靠近需求侧选址,但仍是多单元工业电站 | 模块化电站,不是集装箱式便携 |
| 燃料 / 冷却剂组合 | TRISO + 氦冷 HTGR | 热管微反应堆 | 钠冷 Aurora | KRONOS HTGR 加 ZEUS/LOKI 概念 | 压水堆 | 压水堆 | TRISO + 氦冷 HTGR | TRISO 环形球床燃料 + Flibe 熔盐 |
| 换料 / 服务模型 | 5+ 年后返厂;PPA 或直接销售 | 8+ 年后换料 | 已披露商业持有并运营的试点;此处未抓取详细周期 | 已披露燃料子公司和运输布局;运营周期仍早期 | 全服务交付和 PPA 模型 | 通过 ENTRA1 提供 PPA、租赁或客户所有权 | 在线换料;工业项目交付 | 在线换料和多单元模块化扩展 |
| 最佳适配客户 | 偏远韧性和柴油替代 | 偏远社区、采矿、数据中心 | 国防试点和大型数据中心园区 | 投资者和未来微反应堆采用者,而非已披露锚定用户 | 数据中心和工业站点 | 公用事业、园区、工艺热用户 | 重工业和大型园区 | 电网、工业和大型数据中心用户 |
| 具名商业 / 合作伙伴验证 | Equinix 含定金预订单 | DOME 首批入选 | Meta + Eielson 试点 | 抓取材料中没有具名运营客户 | 抓取材料仅有官网披露 | ENTRA1 商业化架构 | Dow Long Mott 项目 | Google 到 2035 年 500 MW |
| 国防适配度 | 强 | 强 | Eielson 场景强 | 有潜力,但公开证据尚未证明 | 弱到中 | 弱到中 | 中 | 中 |
| 工业蒸汽 / 超大型园区适配度 | 低 | 中 | 中 | Unknown | 高 | 高 | 高 | 高 |
| 公开披露深度 | 私营公司,但披露异常具体 | 私营公司产品页 | 公开申报公司 + 重大伙伴公告 | 公开披露的微反应堆公司 | 私营公司官网级披露 | 上市公司级披露 | 私营公司 + DOE/NRC 项目线索 | 私营公司,有详细技术与伙伴页面 |
| 工厂模式清晰度 | 明确装配、装燃料、测试和返厂闭环 | 明确工厂装配和运输 | 披露商业持有模式多于工厂闭环 | 披露垂直整合野心;执行仍早期 | 工厂生产全模块化建筑 | 工厂制造模块 + 项目融资包装 | 标准化单元,但工业项目现场建造 | 自有制造和模块化扩张 |
各单元格是有证据支撑的定性比较。抓取到的公开材料不足以支撑更紧结论时,表中保留“未知”或早期阶段表述。
[CP003, CP004, CP010, CP014, CP018, CP020]能力图显示,Radiant 的便携细分定位与直接和相邻分布式核能产品在买方关心属性上的对比。
[CP003, CP010, CP014, CP018, CP020, CP021]3.3 相邻分布式核能基准
更大型的分布式核能供应商不能直接替代 1 MWe 便携式机组,但它们确实会争夺预算和买方注意力。Last Energy 的公开卖点在合同模式层面最接近:它同样采用 PPA / 全服务交付姿态,但围绕的是面向数据中心和工业站点的 20 MWe PWR-20 电站。NuScale 进一步上探规模阶梯。其 77 MWe 工厂制造压水模块和 ENTRA1 融资菜单表明,更大型 SMR 供应商也能在不要求客户拥有反应堆的情况下提供 electricity-as-a-service,从而削弱“Radiant 独占 PPA 叙事”的论点。 X-energy 和 Kairos 是最重要的相邻技术基准。X-energy 的 Xe-100 与 Radiant 同属 TRISO 和氦冷却家族,但产品是 80 MWe / 200 MWt 工业平台,价值主张包括工业蒸汽,并以 Dow 四机组项目为旗舰部署。Kairos 转向熔盐冷却和 150 MWe 起步的双反应堆电站,把这个规模与在线换料、数据中心和工业定位、三项获批 NRC 建造许可,以及到 2035 年最高 500 MW 的 Google 协议放在一起。相较于 Radiant 已披露内容,这些都是许可成熟度、工业热和超大型园区负载更强的参照。 实际含义是,不应只拿 Radiant 与其他微反应堆初创公司比较。对超大规模云厂商、大型工业客户和大型园区来说,真实选择可能是在小型便携式反应堆、更大型先进反应堆 PPA 与非核现状方案之间切换。这意味着,当机动性、小占地韧性和快速部署占主导时,Radiant 能赢;当买方看重规模、公开公司披露、蒸汽输出,或由公用事业和工业伙伴支持的数百兆瓦项目所带来的信心时,Radiant 会输。Terrestrial Energy 也在更广泛的 DOE 试点项目范围内,但本次获取的公开材料过于稀疏,无法像 NuScale、X-energy、Kairos 或 Last Energy 那样精确画像。[CP006, CP015, CP020, CP021, CP022, CP023]
| 公司 / 选项 | 公开价格或单元信号 | 合同模式 | 包含能力 / 范围 | 未知项 / 限制 | 含义 |
|---|---|---|---|---|---|
| Radiant | 单元价格未披露 | PPA 或直接单元销售 | 便携式反应堆、工厂装燃料、返厂换料、现场供电和部分供热 | 没有公开的特定场址经济性;公开额定功率不一致仍存在 | 买方更看重机动性和低土建、而非规模时,最适配 |
| Westinghouse eVinci | 单元价格未披露 | 抓取材料中仅有商业产品页 | 可运输 5 MWe 反应堆,堆芯寿期 8+ 年,采用集装箱物流 | 未披露客户专属商业条款 | 竞争点是更高单机输出和长燃料周期,而非最小任务适配 |
| Oklo | 单元价格未披露 | 与 Meta 的购电协议包含预付机制 | 园区级清洁稳定电力,以及商业持有试点叙事 | 便携式单元定价未披露;公开验证以大型园区框架为主 | 更适合超大规模园区,而不是柴油替代场址 |
| Last Energy | 电厂价格未披露 | PPA / 全服务 IPP | 现场 20 MWe 电厂,从设计到运营全服务交付 | 园区级场址范围和电厂经济性未在此公开 | 与 Radiant 最重叠的是包装模式,不是规模 |
| NuScale | 抓取材料中未披露电厂价格 | 通过 ENTRA1 提供 PPA、租赁或客户持有 | 77 MWe 模块,配套项目融资、开发和运营支持 | 便携式或边缘场址定价不适用 | 说明更大型 SMR 也能在融资灵活性上追平 Radiant |
| X-energy | 项目定价未披露 | 项目开发 / 工业承购模式 | 多单元电厂同时供电和工业级蒸汽 | 抓取材料中商业条款没有公开到细项 | 更适合蒸汽需求重的工业园区,而非偏远微电网 |
| Kairos | 项目定价未披露 | 机组群开发 / 承购伙伴叙事 | 最低 150 MWe 电厂,可模块化扩容,支持工业热和电网 | 抓取材料中没有公开的小场址商业包 | 争夺大负荷园区和长期企业清洁电力采购 |
| 柴油 / 电网 / 天然气现状方案 | 成熟且有广泛基准 | 设备采购、燃料合同、公用事业电价或 PPA | 采购路径和运营做法已知 | 视方案而定,可能排放更高、燃料物流更重,或韧性更弱 | 即便核能看起来有吸引力,这些仍是许多买方眼前最现实的替代方案 |
先进反应堆公开价格透明度很低,所以本表比较商业包装和买方负担,不假装给出精确的同口径成本表。
[CP005, CP020, CP021, CP034, CP035, CP036]3.4 护城河耐久性与未解问题
Radiant 的护城河真实存在,但很窄。最强的已披露差异化不是泛泛的“核能”,而是约 1 MWe 规模、可运输性、工厂装料和服务、低站点负担,以及明确的柴油替代定位组合。较大型相邻供应商没有一家能同样干净地匹配这套组合。但这条护城河有几部分比营销叙事暗示的更容易复制。Westinghouse 也在营销可运输性和工厂组装。Last Energy 和 NuScale 都把项目打包和买方融资纳入卖点。X-energy 和 Kairos 也证明,基于 TRISO 和模块化制造的叙事并非 Radiant 独有。 更严重的风险来自商业和渠道。Equinix 是客户拉力的最佳公开证据,但同一买方也在同时押注多家核能开发商。DOME 是有意义的里程碑,但首批队列与 Westinghouse 共享,并非 Radiant 独占。政府项目也在扩大竞争场,而不是收窄:EIA 的 2026 年概览列出了一长串正在通过 DOE 和国防路径推进的供应商。在这种环境里,资本深度、许可节奏和伙伴生态的重要性可能不低于原始便携性。 最终判断是,Radiant 很可能拥有该领域最清晰的产品定义之一,但还没有最耐久的竞争地位之一。便携式 1 MWe 利基可信且差异化。仍然缺失的是合同约束力、客户推出时间表、长期燃料可得性,甚至准确商业功率等级的更清晰证据。在这些缺口关闭前,应把 Radiant 视为一个有前景、但仍受证据约束的利基领导者;更大型、资本更充足的先进反应堆供应商可能从上方挤入这个利基。[CP005, CP006, CP008, CP009, CP028, CP029]
| 护城河主张 | 威胁 | 严重性 | 缓释措施 / 尽调问题 |
|---|---|---|---|
| 便携式约 1 MWe 细分市场 | Westinghouse 和国防相关可运输项目可以进攻同一个偏远供电切口 | 高 | 拿实际场址准备负担、物流成本和部署速度,与 eVinci 和 Project Pele 式系统逐项对标 |
| 工厂装燃料和返厂服务闭环 | 工厂化 / 模块化叙事已在 Radiant、Westinghouse、Last Energy、NuScale 和 Kairos 中普遍出现 | 中 | 核实合同下哪家供应商真正控制装燃料、运输、维护和返厂物流 |
| Equinix 预订单作为客户验证 | Equinix 在多家核能供应商之间多线布局,公开条款并非特定场址 | 高 | 要求提供场址清单、交付时间表、定金金额、终止权和里程碑触发条件 |
| DOME 首测位置 | Radiant 与 Westinghouse 共享首批准入,更广泛的 DOE 项目也支持多家对手 | 中高 | 跟踪 DOME 测试究竟带来许可加速,还是只提供技术验证、没有下游排他性 |
| TRISO / HTGR 技术谱系 | 对燃料和材料的依赖与 X-energy、Project Pele 及其他 TRISO 相关项目相同 | 高 | 要求提供长期商业燃料策略,不能只停留在近期项目支持和子公司公告 |
| 资本高效执行 | 上市或联邦支持的同行可以在许可、选址和获客上比 Radiant 花更多钱 | 高 | 把现金跑道、工厂资本开支需求和伙伴深度,与上市公司及 DOE 支持的同行对比 |
| 最小任务适配 | 若买方偏好 5-150 MWe 园区级项目或现状供电方案,该细分市场可能太小 | 中高 | 检验真实目标场址中,有多少确实需要约 1 MWe,且无法用柴油、天然气或更大型 PPA 解决 |
严重性反映竞争、渠道和证据风险,而不是单纯的反应堆安全或工程风险。
[CP028, CP029, CP032, CP033, CP035, CP037]精简指标概括公开证据中 Radiant 竞争位置最强和最弱的部分。
[CP006, CP009, CP029, CP033, CP039]3.5 图表
04财务情况
4.1 收入模式、商业化路径与披露限制
Radiant 的公开变现故事在品类层面清楚,在合同层面不透明。公司网站称客户可以通过直接购买机组或购电协议购买 Kaleidos;这一点具有战略重要性,因为它意味着 Radiant 并未锁死在纯硬件销售模式里。同一批材料把 Kaleidos 定位为面向偏远站点、医院、军事基地和数据中心的柴油替代电力产品;最好的公开需求证据,是 Equinix 对 20 台反应堆支付定金的预购,以及 Department of the Air Force/DIU 旨在军事基地部署的协议。这些事实重要,但并不等于已披露的收入质量。 已审阅公开来源都没有披露单机售价、任何 PPA 电价、Equinix 定金规模或是否可退、里程碑付款安排,或早期客户现金会如何确认为收入。五年以上回厂补充燃料的闭环也意味着服务和物流义务;它可能成为经常性利润引擎,也可能变成经常性成本负担,但公开材料没有拆分这部分经济性。结果是:商业化叙事可信,有具名交易方和付款信号,但证据不足以建模 ASP、贡献利润率,或一次性硬件收入与更长期服务收入之间的拆分。 [CI001, CI002, CI003, CI004, CI005, CI006]
| 收入流 | 机制 | 公开数值 / 状态 | 收入质量 | 证据状态 | 尽调问题 |
|---|---|---|---|---|---|
| 直接单元销售 | 客户购买 Kaleidos 单元 | 价格和质保条款披露前为中等 | 公司官网确认销售路径 | 提供单元报价、交付里程碑和质保 / 服务包 | |
| 购电协议 | Radiant 销售电力产出,而不是硬件所有权 | 若按经常性计则为中等,但电价和期限未披露 | 公司官网确认 PPA 路径 | 提供电价、合同期限、调价机制、最低购电量和运营方责任 | |
| Equinix 预订单 | 20 台反应堆的预订单加定金 | 定金处理和取消权明确前为低到中 | 定金已披露,经济条款未披露 | 提供单台定金金额、可退性、里程碑和场址排期 | |
| 国防基地部署 | ANPI 协议瞄准首个军事基地交付 | 合同金额和付款里程碑公开前为低 | 项目路径公开,经济条款未公开 | 提供合同金额、里程碑付款、验收标准和长尾服务义务 | |
| 换料 / 返厂服务 | 5+ 年后返厂换料 | 可能带来经常性收入,但服务定价未披露 | 生命周期模式已公开 | 提供换料价格、运输责任和毛利率假设 | |
| 供热 / 附加服务 | 1.9 MWth 热输出可能支持设施供热或海水淡化 | 变现路径未公开,因此不清楚 | 能力已披露,价格未披露 | 说明热输出是打包出售、单独销售,还是商业上无关紧要 |
Radiant 披露了变现路径和客户画像,但已审阅来源均未披露标价、实际成交价、PPA 电价或合同金额细节;空值表示抓取材料中未公开披露。
[CI001, CI002, CI003, CI004, CI005, CI006]| 产品 / 条款 | 价格 / 费率 | 合同形式 | 公开客户信号 | 未知项 / 限制 | 来源锚点 |
|---|---|---|---|---|---|
| 直接销售 | 设备销售 | 官网称可直接销售单元 | 未披露单元 ASP、里程碑时间表或支持包 | Radiant 首页 | |
| PPA | 经常性购电合同 | 官网称可提供 PPA | 未披露每 MWh 价格、期限或运营责任划分 | Radiant 首页 | |
| Equinix 商业交易 | 预订单加定金 | 已披露 20 台反应堆和定金 | 定金金额、可退性和收入确认处理未披露 | Newswire / World Nuclear News | |
| ANPI / 空军路径 | 协议旨在向基地交付反应堆 | 已披露 2028 年交付目标 | 合同金额、联邦义务和付款时间未披露 | Newswire / World Nuclear News | |
| 返厂换料 | 生命周期服务闭环 | 公开材料将其列为产品模式的一部分 | 未公开披露服务费、物流成本或利润率 | Radiant 首页 | |
| 客户集中度 | 具名需求验证 | 一个数据中心交易对手和一条国防路径已公开 | 未披露更完整的积压订单名单或逐场址管线 | Marketplace / 公司新闻稿 |
本表有意区分已披露合同结构和未披露经济性。空值表示抓取来源确认产品存在,但未确认商业价格或会计处理。
[CI001, CI004, CI005, CI006, CI009, CI035]公开证据支撑 Radiant 收入模式的结构,但不支撑把该结构转成可承销现金流所需的价格或会计处理。
该流程是合同和运营路径,不是确定会计结论。公开来源确认路径要素,但未披露订金处理、ASP、PPA 电价或收入确认政策。
[CI001, CI003, CI004, CI005, CI006, CI039]4.2 已融资本、估值标记与表观充足性
公开资本形成是 Radiant 财务记录里最强的一部分。2025 年 5 月 Series C 以 $165M 完成,并明确把总风险融资推至 $225M。2025 年 12 月 Series D 又增加超过 $300M,意味着仅凭公司披露即可确认至少 $525M 的累计公开风险资本。这些轮次都明确关联 Kaleidos 开发、商业化、DOME 时间表和 Tennessee 工厂项目。Tennessee 官员另称 Oak Ridge 项目是一项 $280M 制造和研发投资,将创造 175 个岗位;这是一条有用的支出信号,尽管并不等于已经花出的现金。 即便如此,已融资本不等于资本充足。获取到的公开来源没有披露 Series D 后的现金余额、当前月度烧钱速度或经过核实的现金跑道计算。广泛流传的超过 $1.8B 估值标记来自 DCVC 这家既有投资人,而不是监管文件或市场出清的老股交易,因此应视为方向性而非确定性。公开证据因此支持这样的判断:Radiant 已吸引足够资本,能继续推进 2026 年重大里程碑;但它不能精确回答当前流动性是否足以建完工厂、执行测试活动、吸收延误,并撑到有意义的客户现金产生。 [CI010, CI011, CI012, CI013, CI014, CI015]
| 项目 | 公开金额 / 状态 | 证据质量 | 融资含义 | 尽调问题 |
|---|---|---|---|---|
| Series C | 165 | 高 | 以百万美元计,为 2025 年开发和早期工厂工作提供资金 | 确认扣费后净募资额和当前未使用余额 |
| Series C 后公开累计融资 | 225 | 高 | 以百万美元计,给出 2025 年 12 月轮次前的可见资本基线 | 与股权结构表以及任何过桥或补助资金流入核对 |
| Series D | 300 | 高 | 公开确认超过 $300 million;此处以 USD 百万计显示下限 | 提供准确总募资额、净募资额和证券类型 |
| 已确认公开累计融资下限 | 525 | 高 | 以百万美元计,可从公开披露确认的最低风险资本 | 提供生命周期准确融资额,包括补助、客户现金和战略工具 |
| 估值标记(USDm) | 1800 | 中 | 有用的情绪信号,但来源是投资人,不是监管文件 | 提供定价轮投后估值、股份数和清算优先堆叠 |
| Oak Ridge 项目投资(USDm) | 280 | 高 | 显示商业化资本开支和制造野心 | 提供支出分期、土地 / 建筑所有权和应急储备 |
| Tennessee Nuclear Energy Fund 支持 | 已使用;抓取到的直接来源未说明公司特定金额 | 中 | 州政府支持存在,但确切补贴金额仍不清楚 | 提供补助金额、提款条件和追回条款 |
| DOME 测试 | 申请方自行出资;实验窗口最长六个月 | 高 | 测试项目会在收入到来前消耗资本,且仍受里程碑约束 | 提供内部 DOME 预算和下行情景下的延误假设 |
| 燃料供应 | DOE 有条件承诺 HALEU;国内供应目前尚不可得 | 高 | 燃料仍是外部依赖,可能改变时间表和营运资金需求 | 提供已签燃料合同、时间安排和备用采购方案 |
| Part 70 审查 | 已受理;公司申请按 2026 年 11 月 1 日目标加速审查 | 高 | 许可进度可能加快或推迟支出转化为收入 | 提供许可关键路径和按概率加权的延期情景 |
| 手头现金 | 中 | 现金跑道分析缺少关键输入 | 提供不受限现金、受限现金和债务余额 | |
| 月度烧钱速度 / 现金跑道 | 中 | 公开来源无法验证 | 提供当前烧钱速度、2026-2028 年预期烧钱爬坡和融资触发点 |
已确认融资金额与缺乏支撑的估值标记分开处理。除非另有说明,数字行均为百万美元;估值行只是投资人 / 新闻给出的方向性标记,不是有备案支撑的公允价值结论。
[CI010, CI012, CI014, CI015, CI017, CI018]公开材料中只有少数财务数值可获数字支撑;承销真正需要的其他信息仍未披露。
相同的低/高值表示单点公开披露或已确认最低下限,而非模型分布。估值项仅作为中等置信度的方向性标记展示。
[CI014, CI015, CI018, CI020]Radiant 当前融资叙事是一系列资本输入,先投向制造、测试和许可闸门,之后才可能出现公开收入披露。
此图展示顺序而非规模。公开记录给出了真实里程碑和支出信号,但没有内部现金瀑布或应急储备。
[CI016, CI018, CI019, CI022, CI025, CI027]4.3 成本结构、规模化支出与单位经济代理指标
Radiant 的公开披露给出了真实支出信号,但没有给出成型的单位经济。最清晰的资本开支代理指标是计划中的 $280M Oak Ridge 工厂;最清晰的近期运营支出代理指标是招聘页面:66 个开放岗位覆盖工程、运营、财务、供应链和监管工作,最高薪资区间达到 $314,475 加股权。这些岗位让人很难说公司处在低烧钱阶段。它们也显示,这个收入前组织正在同时为产品工程、许可、制造就绪和公司基础设施付费。 其他成本信号进一步强化了资本密集度。DOE 称 DOME 实验由申请方自筹资金,并依赖燃料可得性和监管就绪度。Part 70 申请显示,工厂模式包括装料、环境审查、材料控制、实体安全和加速监管工作流。独立报道还提到一份单独的 Urenco HALEU 相关协议,说明除工厂资本开支外,燃料采购支出也存在。合在一起,公开记录支持把 Radiant 看作一家制造加服务企业;在商业收入披露前,它已承担有意义的监管、供应链和薪酬义务。记录不支持的是任何关于客户价格、毛利率、CAC、回本周期,甚至总员工数的硬公开数字,因此公开单位经济分析只能基于代理指标。 [CI018, CI020, CI021, CI022, CI023, CI024]
| 指标 | 公开数值 / 状态 | 置信度 | 重要性 | 尽调问题 |
|---|---|---|---|---|
| 已确认公开风险融资 | 525 | 高 | 为外部可见资本基础设下以百万美元计的下限 | 与实际到账现金以及任何老股或非新股资金流入核对 |
| 计划中的 Tennessee 工厂投资(USDm) | 280 | 高 | 与商业化相关的最大已披露资本开支式承诺 | 提供支出计划、资金来源拆分和应急预算 |
| 开放职位 | 66 | 高 | 招聘规模可代理近期运营开支和组织搭建 | 提供当前员工数、已填补与开放岗位对比、职能结构 |
| 已披露最高薪酬区间(USD) | 314475 | 高 | 显示许可密集型扩张中专业人才成本 | 提供全成本薪酬、股权消耗和招聘计划 |
| 单台反应堆客户价格 | 中 | 建模收入规模和回本周期必需 | 提供当前报价、预期折扣和客户落地负担 | |
| PPA 电价 | 中 | 用于对比柴油或电网替代方案 | 提供 $/MWh、指数化机制、期限和最低电量承诺 | |
| 单台反应堆定金金额 | 中 | 决定客户现金到底能为营运资本降多少风险 | 提供单元定金、可退性和各里程碑应付余款 | |
| 毛利率 | 中 | 制造和服务质量的核心决定因素 | 提供毛利率桥,拆分燃料、装配、物流和现场支持 | |
| 月度烧钱速度 | 中 | 把融资额换算成现金跑道时必需 | 提供过去六个月现金消耗,以及按职能拆分的烧钱结构 | |
| 手头现金 | 中 | 资本充足性分析必需 | 提供非受限现金,以及任何债务或受限现金余额 | |
| 现金跑道(月) | 中 | 用于测试融资依赖能否顶住进度滑坡 | 提供基准和下行情景下的现金跑道假设 | |
| 总员工数 | 中 | 用于标准化薪酬和人效 | 按场址和职能提供当前员工数与招聘计划 |
各行混合了硬公开数字和空值披露缺口。按约定规则,数字值保留为数字;空值表示抓取来源未公开披露该指标,而不是零。
[CI014, CI018, CI028, CI029, CI036, CI037]Radiant 的公开记录足以识别支出和商业化信号中的成本栈,但不足以计算利润率或回本周期。
这座桥有意区分已披露资本和支出信号,以及缺失的运营输出。它是公开证据图谱,不是已求解的单位经济模型。
[CI018, CI025, CI027, CI028, CI029, CI036]4.4 融资含义、反向证据与承销判断
核心融资含义很直接:Radiant 在风险投资口径下看起来可融资,但在项目或经营口径下还不可承销。风险投资人已经大规模投入,公司公开记录也显示州政府支持、DOE 参与、具名客户,以及比许多先进核能初创公司更具体的 NRC 材料许可流程。但这些正面因素旁边还有几项未解依赖。DOE 称 HALEU 仍无法从国内供应商处获得,UCS 强调了 HALEU 使用相关的扩散和安全担忧,DOME 活动则需要自筹资金且取决于里程碑。这意味着燃料合同、测试或许可一旦延误,Radiant 可能在收入到来前消耗更多资本。 反向来源同样重要,因为它们指出了融资公告天然会弱化的风险。Beyond Nuclear 关注围绕回厂补充燃料模式的社区反对、运输、废料和纳税人责任问题;American Bar Association 则指出,尽管政策势头强,北美真实 SMR 部署先例仍非常少。公开记录仍缺少合同金额、积压订单转化条款、定金机制、现金余额、烧钱速度和利润率数据。因此,正确的财务判断不是 Radiant 缺乏动能,而是公司仍是一家依赖里程碑融资、披露较轻的企业;面对进度滑坡和商业化风险,主要缓冲仍是未来融资,而不是当前已报告经营。 [CI017, CI025, CI027, CI031, CI032, CI033]
| 缺失指标 | 重要性 | 当前公开替代指标 | 精确尽调路径 |
|---|---|---|---|
| 收入 / ARR | 无法判断客户活动是否已经转化为确认收入 | 公开信息只有定金和协议 | 提供月度收入桥、积压订单,以及 ARR 或合同价值披露 |
| 毛利率 | 无法判断工厂和燃料模型能否跑出健康的单位经济 | 工厂、燃料和物流义务已公开,但没有量化 | 按硬件、服务和换料组件提供毛利率桥 |
| 现金、烧钱速度和现金跑道 | 无法用进度延期情景测试资本充足性 | 公开轮次显示融资能力,但不显示当前流动性 | 提供过去十二个月现金流、当前现金和下行情景现金跑道 |
| 客户价格 / PPA 费率 | 无法将 Kaleidos 经济性与柴油或电网替代方案比较 | 变现路径已公开;定价未公开 | 提供报价单机销售价格、PPA 电价和任何最低采购量条款 |
| 定金机制 | 无法判断已宣布定金是否真是降低风险的客户现金 | 已披露定金,但未披露金额或可退款性 | 提供单机定金、托管或退款条款和里程碑时间表 |
| 按站点拆分的积压订单和取消权 | 无法评估转化概率或集中度风险 | 公开具名的只有 Equinix 和军方路径 | 提供站点清单、承诺机组数、终止触发条件和最终期限 |
| 员工数和职能结构 | 无法规范化薪酬支出或生产率 | 招聘页面只显示空缺岗位 | 提供当前员工数,按地点和职能拆分,并附招聘计划 |
| 工厂营运资金 / 库存计划 | 装料和返厂物流可能需要大量库存和运输资本 | Part 70 申请确认了装料范围,但没有确认资产负债表处理 | 提供库存政策、燃料所有权路径,以及返厂换料的营运资金假设 |
这些具体缺失输入阻碍了标准财务承销。每个缺口都足够具体,可以映射为尽调请求,而不是泛泛要求更多细节。
[CI006, CI007, CI008, CI017, CI030, CI035]4.5 图表
05产品与技术
5.1 产品定义与客户工作流
Radiant 销售 Kaleidos 的定位是便携式核发电机,而不是常规现场建造电站。从主页、Buckley 公告、Equinix 预购新闻稿到 Marketplace 资料,公司一直把该机组定位为偏远设施、医院、军事设施和数据中心的柴油发电机替代品。外部来源部分印证了这种定位:NRC 把 Kaleidos 描述为装在单个海运集装箱内、可运输、约 1 MWe / 3 MWth 的微反应堆;Equinix 相关报道也确认,至少有一个严肃商业买方把该产品视为现场数字基础设施电力,而不是抽象的未来反应堆概念。公开证据暗示的工作流在核能领域异常以工厂为中心:机组在场外建造、装料和测试,用卡车或飞机运到客户处,快速安装,运行约五年后补充燃料,然后返回集中服务。这个工作流有吸引力,因为它尽量减少现场施工和本地核处理;但它也意味着商业产品不只是反应堆堆芯,而是完整物流链:工厂装料、运输批准、现场集成、远程监控,以及机组返回补充燃料或翻新时的周转能力。[CE001, CE002, CE003, CE007, CE008, CE009]
| 模块 / 资产 | 主要用户 | 当前成熟度 | 有证据支撑的差异化 | 主要尽调缺口 |
|---|---|---|---|---|
| Kaleidos 反应堆和屏蔽包 | 远程站点运营方 / 基地设施团队 | 高级原型;目标 2026 年完成装料测试 | 单箱体便携式 HTGR,定位为柴油替代微型反应堆 | 未披露装料运行记录或现场可用率 |
| 氦主回路和石墨 / TRISO 堆芯 | Radiant 反应堆工程团队和未来站点运营方 | NRC 和行业报道已公开描述设计 | 使用已知 HTGR 材料组合,而非新型燃料形态 | 装料条件下的集成性能尚未发布 |
| 干式冷却和被动空气夹套系统 | 需要无水选址灵活性的客户 | 宣称的设计特性;尚未经过现场验证 | 无现场用水要求扩大了选址范围 | 独立冷却和瞬态数据未公开 |
| 工厂装料和 R-50 生产线 | Radiant 制造和装料运营 | 商业化前;Part 70 审查进行中 | 集中装料和翻新支撑快速部署模型 | 商业经济性取决于许可获批和产能吞吐 |
| 远程机队监控 / 运行规程 | Radiant 集中运营团队 | 招聘信息显示正在建设 | 支撑自主管理机队概念和场外专家模型 | 尚无公开证据验证控制室或操作员表现 |
各行只反映公开披露的模块;内部子系统边界、供应商名单和未发布的工程接口无法从公开证据获得。
[CE001, CE007, CE009, CE019, CE023, CE031]| 用户任务 | 当前工作流 | Kaleidos 方案 | 有证据支撑的收益 | 限制 / 注意事项 |
|---|---|---|---|---|
| Buckley 军事基地韧性供电 | 电网加柴油备用,支撑关键任务负载 | 部署一台或多台 Kaleidos 机组,提供持续在线的韧性基地电力 | 美国空军部具名部署路径已出现 | NEPA、选址,以及产品到项目的功率匹配仍未解决 |
| 数据中心边缘或园区供电 | 电网电力、备用发电机和依赖输电扩容的增长 | 用现场微型反应堆供电,支撑数据中心韧性和增长 | Equinix 已签定金和 20 台机组预订 | 具体部署站点和多机组架构未披露 |
| 医院或灾害响应备用电力 | 柴油发电机叠加燃料物流和排放负担 | 便携式无水微型反应堆,提供长时备用电力 | 公司和第三方来源反复提到医院 / 灾害用例 | 未公开医院部署或医疗监管工作流 |
| 远程工业或社区电力 | 柴油运输和微电网平衡 | 单箱体核发电机,续航多年 | 便携形态和无水宣称契合远程 / 离网叙事 | 尚无公开商业参考站点 |
| 热电联产用于供热或海水淡化 | 独立柴油或燃气发电,再叠加热能设备 | 同一设备包同时输出电力和外送热能 | 公司和行业来源提到供热和海水淡化应用 | 未有公开客户展示热电联供用法 |
工作流同时包含具名客户和公开宣称的目标客群;在已审阅证据中,只有 Buckley 和 Equinix 是外部具名交易对手。
[CE002, CE004, CE026, CE028, CE029, CE043]产品是工厂到站点再回工厂的服务闭环,不是现场建造电站交付。
[CE002, CE008, CE009, CE010, CE011, CE028]5.2 核心架构、功率转换与热设计
在架构本身上,现在最强的独立佐证来自 NRC 预申请页面和独立行业报道,而不只是 Radiant 自己的营销文案。NRC 称 Kaleidos 是高温气冷反应堆,使用 TRISO 燃料、氦气冷却剂和棱柱石墨块,并全部装在单个海运集装箱内。DOE 更广泛的 HTGR 说明和 INL 技术材料页面支持这样的判断:TRISO 燃料、氦冷却和石墨慢化是该反应堆类别的成熟特征,因此 Radiant 并不是从零发明一种新燃料或慢化剂系统。更未被证明的是围绕这些已知材料的一体化产品包。主页声称,机组借助风扇和被动空气夹套冷却实现现场零用水,交付后可快速满功率启动,还能提供电力加热输出组合,支持供热或海水淡化。这些说法被行业文章方向性重复,但还没有带燃料运行的公开运营数据支撑。结果是一幅混合技术图景:材料栈看起来符合类别,但产品包层面的逻辑取决于能否在一个工厂制造箱体里,把干式冷却、屏蔽、可运输性和功率转换硬件整合跑通。[CE001, CE004, CE005, CE006, CE007, CE008]
| 层级 / 组件 | 作用 | 关键依赖 | 主要风险 |
|---|---|---|---|
| 棱柱石墨堆芯中的 TRISO 燃料 | 在 HTGR 架构中提供裂变热和被动燃料稳健性 | 合格 HALEU TRISO 供应和石墨制造 | 类别层面的材料已知,但 Radiant 特定堆芯性能未验证 |
| 氦主冷却剂回路 | 将热量从堆芯导出,避免水系统的水化学或冷却剂活化问题 | 主回路密封、换热器和透平机械集成 | 泄漏管理和集成热性能没有公开运行数据 |
| 功率转换透平机械 | 将反应堆热量转为 1-1.2 MWe 输出,并支持灵活负载运行 | 可靠旋转设备、控制系统和维护间隔 | 未披露公开效率、负载周期或寿命数据 |
| 被动空气夹套和风扇式干式冷却 | 无需现场用水即可移除余热和运行热 | 对流散热路径和环境站点条件 | 被动安全宣称主要来自公司描述,而非独立验证 |
| 运输和屏蔽包 | 支持卡车或飞机运输,并快速安装 | 运输批准、屏蔽裕度和站点接口 | 快速部署承诺尚未在商业规模经过测试 |
| 集中监控和运行规程 | 支撑自主管理机队模型和标准化运营 | 操作员培训、GUI 成熟度和书面规程 | 公开证据显示公司在招人,而非已验证的运行表现 |
本表捕捉的是集成产品栈,而不只是反应堆堆芯;风险按设备包层面呈现,因为 Kaleidos 卖的是可部署发电机,不只是容器设计。
[CE001, CE005, CE006, CE007, CE009, CE031]从运输外壳到换料周期,分层展示已公开披露的 Kaleidos 方案。
[CE001, CE005, CE006, CE007, CE009, CE010]5.3 测试路径、制造模式与部署顺序
判断 Radiant 的技术成熟度,主要应看 DOME 和 R-50 两条路径,因为它们是把便携式反应堆概念转化为可制造产品的两条公开轨道。DOE 记录显示,Radiant 于 2024 年完成前端工程和实验设计阶段,正在准备在 INL 的 DOME 设施开展带燃料 Kaleidos 实验,春季或 2026 年中测试被反复引用为目标窗口。与此同时,NRC 记录和 2026 年 5 月 Part 70 审查报道显示,Radiant 已从泛泛的预申请沟通,进入针对 Oak Ridge 装料大楼的具体许可工作。这很重要,因为公司不只是提议在 Tennessee 组装反应堆;它提出的是工厂装料和生产模式,其经济性取决于对特殊核材料的集中处理。当前招聘岗位反映的技术组织也支持这一判断。公开招聘需求覆盖流体系统、机构、结构、反应堆运行、中子学、瞬态分析、规程、操作员培训和控制室界面,说明一体化示范活动正在建设中,而不是一个冻结设计只等许可。这个产品因此处在后期原型和首个工厂产品之间,而不是广泛商业就绪阶段。[CE013, CE014, CE015, CE016, CE019, CE020]
| 日期 / 阶段 | 里程碑 | 状态 | 含义 | 来源锚点 |
|---|---|---|---|---|
| October 2022 | NRC 预申请活动在案卷 99902106 下启动 | 已完成历史里程碑 | 说明商业化推进前,公司已较早接触监管 | SE006 |
| 2024 | DOME 实验规划的 FEEED 阶段已完成 | 已完成历史里程碑 | 设计、进度、预算和测试规划已越过概念阶段 | SE007 |
| February 2026 | DOE 批准 PDSA/DARK 提交材料 | 已完成当前里程碑 | 启动活动前,DOE 三道安全文件关口已通过第二道 | SE004 |
| 2026 年春季 / 夏季目标 | DOME 装料启动 / 首次满功率实验 | 进行中 | 第一个真实产品验证点;截至运行日期仍不是已完成里程碑 | SE015 |
| 2026 | R-50 Part 70 审查和 Oak Ridge 工厂建设启动 | 进行中 | 制造路径与测试路径并行推进 | SE005 |
| 2028 目标 | 首批客户部署、Buckley 交付目标和首台 R-50 产出 | 目标 / 尚未实现 | 最重要的商业期限,也是承销中的关键二元变量 | SE002 |
未来行是公司或行业媒体给出的目标,而非已经完成的结果;本表把已完成的安全和许可步骤,与仍面向未来的部署承诺拆开。
[CE013, CE014, CE015, CE017, CE019, CE023]Radiant 的产品路径需要 DOME 测试、NRC 许可、工厂装料和已点名交易对手同步推进。
[CE014, CE016, CE019, CE021, CE023, CE024]5.4 安全设计、质量信号与监管路径
Radiant 的公开安全论证,在 DOE 或 NRC 已触及设计的地方最强;在公司要求投资人或客户基于信任接受产品包层面安全主张的地方最弱。DOE 授权路径已经显著推进:公司和行业来源都认为 Radiant 通过了 Safety Design Strategy 和 PDSA/DARK 里程碑,独立报道也把这些批准描述为带燃料 DOME 启动前分阶段安全审查的一部分。NRC 侧记录也是真实的,而非愿景式说法。Kaleidos 页面列出多项差距分析以及已受理或正在审查的文件,Part 70 申请和相关报道也明确表明,Oak Ridge 装料大楼已进入正式审查。但投资人不应把这些步骤误认为已完成的产品安全论证。公开来源没有显示被动空气夹套冷却概念的独立运行记录、经外部审计的控制室或操作员培训包,也没有显示针对回运服务模式的透明寿命末期废料与运输计划。公司比许多微反应堆同业拥有更多监管动作,但面向重复商业部署的安全边界,仍依赖尚未发生的里程碑。[CE015, CE016, CE017, CE018, CE019, CE020]
| 控制 / 信号 | 当前状态 | 范围 | 未解缺口 |
|---|---|---|---|
| DOE DOME 授权路径 | SDS 和 PDSA/DARK 里程碑已完成;后续安全步骤仍待推进 | 国家实验室装料实验测试授权 | 不等同于已完成商业反应堆许可 |
| NRC 预申请案卷 | 2022 年 10 月以来持续活跃,已有多份缺口分析和获受理的材料许可证文件 | 反应堆和工厂接口的商业许可准备 | 预计未来仍需单独提交反应堆申请 |
| R-50 Part 70 申请 | 已受理,将在 2026 年进入详细审查 | Oak Ridge 的装料建筑和工厂生产运营 | 批准仍未落地,并直接卡住工厂装料模型 |
| 操作员规程和控制室就绪度 | 公开招聘显示规程、培训和 GUI 工作正在搭建 | 示范运营和未来机队运营 | 未公开运行就绪度验证或审计 |
| 生命周期废物和返厂责任 | 公司模型是集中返厂服务 | 乏燃料处理、运输和翻新工作流 | 未披露详细公开时间表、产能或责任计划 |
这里的合规证据只覆盖公开表层;没有披露运行记录或第三方审计,不应被误读为负面发现,也不应被视为健康背书。
[CE015, CE016, CE017, CE018, CE019, CE020]Kaleidos 的不同模块成熟度差异很大;材料栈证据更扎实,整机集成运行仍缺证据。
[CE015, CE016, CE019, CE021, CE031, CE032]5.5 执行风险与仍未验证的环节
核心投资判断是:尽管公开证据比一年前丰富得多,Kaleidos 仍没有在现场条件下证明自己。它还没有加燃料后的运行记录、可用性历史、服务周转证据,也没有公开客户现场正在发电。时间表风险因此不是单线风险,而是会相乘:DOME 必须启动,实验必须产出可信的安全和性能数据,产品和工厂的 NRC 许可必须持续推进,公司还要把一个以演示为中心的组织,改造成可复制的制造和服务体系。两个悬而未决的问题尤其关键。第一,公开材料没有解释 1-1.2 MWe 的 Kaleidos 机组如何对应 ANPI 项目披露的 3-10 MWe 目标,因此军事部署到底需要多台机组、更高额定功率版本,还是分阶段站点架构,仍不清楚。第二,返厂服务模式是 Radiant 价值主张的核心,但公开文件仍很薄:运输周转、备用机组覆盖、退役流程和长期废弃物责任,都没有达到投资级披露深度。产品逻辑足够可信,值得严肃尽调,但还没有成熟到可按已去风险基础设施定价。[CE027, CE030, CE037, CE038, CE039, CE040]
06客户情况
6.1 具名证明与目标细分之间的差距
Radiant 公开客户证据的宽度,远窄于官网列出的终端市场。综合 Radiant 官网、Marketplace 走访、Equinix 公告以及 Air Force / DIU 材料,真正高于泛化营销的需求锚点只有两个:商业侧的 Equinix,防务侧的 Department of the Air Force / DIU Buckley 路径。其他内容——医院、偏远村庄、偏远工业场地、抽象的军事基地、海水淡化和微电网——更像目标细分或用例营销,而不是已披露的签约交易对手。需求证明层级很重要:带押金的具名预订单,强于官网用例;NEPA 仍未完成的政府选址,强于一个 logo,但明显弱于正在运行的客户。因此本章把 Radiant 视为已拿到真实早期需求证明,但尚未拥有多元化装机客户基础。[CU001, CU008, CU009, CU014, CU015, CU016]
| 分群 / 交易对手 | 买方 / 用户 / 付款方 | 用例 | 公开规模 | 收入 / 战略价值 | 缺口 |
|---|---|---|---|---|---|
| Equinix | 买方 / 付款方:Equinix;终端用户:数据中心运营团队和客户 | 现场数据中心供电和韧性 | 20 台 Kaleidos 机组预订,已支付定金 | 定金和机组数公开,是最强商业证据 | 未披露站点清单、定价、可退款性或推出节奏 |
| 美国空军部 / DIU / Buckley SFB | 买方 / 赞助方:DAF 和 DIU;终端用户:Buckley 任务运营方 | 面向空间监视和雷达负载的关键任务基地韧性 | 具名站点配对;目标是到 2030 年或更早落地首台 DAF 先进反应堆 | 战略上重要的联邦验证,且可能成为旗舰防务参考案例 | 未披露合同价值、所有权模型、机组数或最终批准 |
| 其他数据中心运营商(推断分群) | 潜在买方:超大规模、托管和边缘运营商 | 受电网约束或 AI 驱动的数据中心园区,需要稳定电力 | 已审阅证据中没有其他具名商业交易对手 | 如果 Equinix 成为参考销售,可能带来高 ACV | Equinix 之外的商业客户名单未披露 |
| 更广泛的军方 / 远程设施买方(推断分群) | 潜在买方:美国及盟军设施 | 停电风险影响关键任务时,需要持续在线的韧性电力 | 公开分配给 Radiant 的只有 Buckley | 防务采购可能带来规模大、粘性高的合同 | 转化取决于联邦采购、NEPA、NRC 路径和具体站点容量匹配 |
| 医院 / 远程社区 / 工业微电网(营销用例) | 潜在商业或公共部门买方 | 替代柴油,并为不可靠电网提供备用电力 | 营销材料和访谈中具名,但未披露已签交易对手 | 支撑广阔可服务市场叙事 | 未披露现场部署、客户名称或经济性 |
本表把具名交易对手与营销或推断的买方原型拆开。公开规模基于披露的机组数量或项目里程碑;经济性和客户名单的空白是有意保留,因为已抓取来源没有披露这些信息。
[CU014, CU015, CU016, CU017, CU018, CU020]| 客户 / 交易对手方 | 细分市场 | 部署 / 用例 | 生产级 / 试点 | 结果 | 限制 |
|---|---|---|---|---|---|
| Equinix | 商业数据中心买家 | 为支持 AI 的数据中心扩张和韧性提供现场供电 | 带押金的预订单;未披露已上线部署 | 公开预订 20 台 Kaleidos 并支付押金;这是最强商业证明 | 未披露部署站点、定价、合同期限或现场结果 |
| Department of the Air Force / DIU 政府路径 | 联邦买家 / 采购发起方 | 借 ANPI 路径为军事设施提供韧性电力 | 协议 + 入选开发商路径;尚未运行 | 首个旨在向美国军事基地交付量产微反应堆的协议;Radiant 后来与 Buckley 配对 | 未披露合同金额、所有权结构或收入确认节奏 |
| Buckley Space Force Base | 具名基地 / 终端用户 | 支撑雷达和空间监视系统的关键任务基地韧性 | 已选定站点;环境和许可审查待完成 | 具名站点和任务背景让证据强于泛泛的国防用例 | 仍未部署;社区沟通、NEPA 和最终批准仍未完成 |
这张表有意把付费商业买家与联邦和站点级交易对手放在一起,因为 Radiant 在审阅证据中只披露了一个商业买家和一条国防路径。交易对手披露设备数量、押金、站点配对或任务背景时,公开证据质量最强。
[CU001, CU008, CU009, CU010, CU020, CU021]公开证据显示,旅程从营销场景走到具名交易对手、演示关口,最后还要补上扩张证据。
这些阶段概括的是公开证据,不是 CRM 阶段数据。该图是证据质量旅程图,不是披露的销售运营漏斗。
[CU016, CU017, CU020, CU021, CU025, CU043]6.2 Equinix 是最强商业信号,但周期仍很长
Equinix 是最清晰的商业证明点,因为双方公开确认了 20 台机组预订单和押金,这比泛化备忘录或参考 logo 更强。同样重要的是,Equinix 说明了需求动因:AI 驱动的用电需求压迫电网和电力采购,公司在寻找稳定的全天候电力。但 Equinix 自身披露也提醒投资人不能过度解读。Radiant 只是 Equinix 更大能源组合的一部分,该组合还包括 Oklo、ULC-Energy / Rolls-Royce SMR、Stellaria、Bloom 燃料电池、天然气容量和电网升级。Equinix 告诉 Data Center Knowledge,许多下一代核技术距离部署仍有数年,选址还在早期,仍需要政策变化、更快许可和熟练劳动力。公开材料披露了机组数量和押金存在,但没有披露押金规模、是否可退、站点清单、合同期限、单机经济性,也没有说明 Equinix 未来核电负荷中 Radiant 实际能拿下多少。[CU001, CU002, CU003, CU004, CU005, CU006]
| 指标 | 数值 | 日期 | 来源 | 置信度 | 含义 | 缺失分母 |
|---|---|---|---|---|---|---|
| Equinix 预订 + 定金 | 20 台 Kaleidos 机组,已支付定金 | 2025-08-14 | Radiant + Equinix 公告 | 高 | 最强的公开商业需求证据 | 定金金额、可退款性和付款里程碑未披露 |
| Buckley 站点配对 | ANPI 将 Radiant 匹配至 Buckley SFB | 2026-04-22 | DAF + Radiant + ANS/WNN | 高 | 最强的公开联邦需求证据 | 未披露合同价值或最终批准 |
| ANPI 供应商池资格 | 8 家公司具备 OT 奖项资格 | 2025-04-10 | DIU | 高 | Radiant 已从广泛供应商池推进到具名站点合作方 | 供应商未披露资金金额和授标时间 |
| 首批客户部署 | 2028 年开始 | 2025-2026 年声明 | Radiant 公司声明 | 中 | 划定最早公开商业化时间线 | 取决于 DOME、许可和采购能否成功 |
| DAF 运行目标 | 到 2030 年或更早,在一个 DAF 设施部署至少一台先进反应堆 | 2026-04-22 | 美国空军 | 高 | 防务路径横跨多年,受里程碑驱动 | 未说明 Radiant 会成为首个投运机组 |
| 美国数据中心电力需求 | 预计到 2028 年翻倍或增至三倍 | 2024-12-20 | DOE | 中 | 解释了为何数据中心买方在探索新的稳定电力来源 | DOE 不会把该需求直接映射为 Radiant 订单 |
| 公开客户数 | 未披露 | 已审阅的公开来源 | 高 | 无法衡量管线宽度或 ACV 分散度 | 付费客户确切数量和积压订单仍未公开 | |
| 已投运客户站点 | 未披露 | 已审阅的公开来源 | 高 | 尚无现场运行可用率或交付电量的公开证据 | 需要站点级投运数据和运行可用率历史 |
各行把已披露里程碑与显式 null 放在一起,标出公开记录沉默的位置。「未披露」指抓取到的证据集中没有该信息,不代表数值为零。
[CU001, CU004, CU009, CU010, CU013, CU022]| 交易对手 / 主题 | 公开内容 | 未公开内容 | 为何重要 | 尽调追问 |
|---|---|---|---|---|
| Equinix 预订单经济性 | 20 台预订单和押金 | 押金规模、是否可退、单台定价、里程碑节奏、服务定价 | 决定该预订单能否切实降低营运资金风险,还是更多只是象征性 | 提供商业条款清单或已签订单摘要 |
| Equinix 部署架构 | 设备计划用于数据中心 | 哪些站点、每站点多少台、并网范围、备用架构 | 需要判断订单是边缘、备用,还是主电源规模 | 提供候选站点清单和首个设施的预期架构 |
| Buckley 商业结构 | ANPI 下的具名站点配对和联邦韧性理由 | 合同金额、所有权、登记运营方、退役责任、付款时间 | 需要判断 Buckley 是设备收入、长期服务收入,还是第三方 PPA 式结构 | 提供 OTA / 授标摘要和预期商业模型 |
| 更广客户名单 / 积压订单 | Equinix 和 Buckley 已具名 | 总客户数、分阶段管线、复购订单、更多具名客户 | 集中度和销售可复制性分析的核心输入 | 按细分、阶段和预计成交日期提供管线 |
| 留存 / 满意度证据 | 没有公开上线客户站点,也没有公开流失或 NPS 数据 | 续约、满意度、运行可用率、SLA 表现、可供访谈的部署 | 缺少这些指标,公开客户故事仍停在投运前 | 首个送电后提供客户推荐访谈和早期运营 KPI |
这张表把缺失的商业条款视为核心分析发现,而不是格式遗漏。每个缺失字段都直接影响集中度、营运资金或转化风险分析。
[CU005, CU006, CU023, CU028, CU031, CU043]公开证据漏斗:从广泛营销的需求,收窄到具名交易对手、明确商业 / 采购里程碑和已投运现场。
阶段计数是依据已抓取证据得到的证据数量:该漏斗衡量公开可见的交易对手,而不是内部销售管线规模。
[CU001, CU005, CU009, CU021, CU022, CU023]6.3 Buckley 验证了防务需求,但转化仍要跨过采购和选址关口
Air Force / DIU 的 Buckley 路径有价值,因为它把抽象的军事用途营销,变成了具名联邦项目、具名基地和具名任务集。Radiant 先披露了 2025 年 7 月与 DIU 和 Department of the Air Force 的协议,随后 2026 年 4 月的 Air Force 官方材料又把 Radiant 与 ANPI 下的 Buckley Space Force Base 配对。这是很强的类客户需求证明,但不等于已入账、可重复的运营收入。公开记录仍显示它处在部署前:选址和环境分析必须按 NEPA 推进,许可仍未完成,具体商业结构没有披露。Partnership for Global Security 对军事微反应堆采购的描述表明,这类项目可能采用长期第三方所有权和合同模式,而不是简单硬件销售。防务信号因此真实且具战略意义,但投资人仍需尽调:谁拥有反应堆,谁承担退役责任,Buckley 需要多少台机组,客户现金何时真正开始流入。[CU008, CU009, CU010, CU011, CU012, CU013]
| 扩张驱动因素 | 集中度风险 | 影响 | 尽调路径 |
|---|---|---|---|
| Equinix 将预订单转为逐站点铺开 | 公开信息中只有一个具名商业买家 | Equinix 若规模化铺开,会验证商业可复制性;若放慢,公开需求证据会明显变弱 | 获取站点清单、部署节奏、付款里程碑,以及任何独家条款或上限 |
| Buckley 从入选站点转为运行反应堆 | 公开信息中只有一条具名国防路径 | Buckley 成功送电会验证军方采用;若延迟,国防收入会进一步后移 | 要求提供 NEPA 时间线、NRC 路径、合同业主 / 运营方结构和验收标准 |
| Buckley 之后的后续 DAF 或 DoD 基地 | 未公开披露 Radiant 后续基地订单 | 会降低单站点风险,并证明联邦采购内部可复制 | 要求提供额外基地管线、期权结构和预算授权 |
| Equinix 之外更广泛的商业数据中心买家 | 没有额外具名主机托管或超大规模买家 | 会分散 ACV,降低对单一旗舰客户的依赖 | 要求按细分和阶段提供合格管线,以及任何已签 LOI 或押金 |
| 数据中心和国防之外目标细分的转化 | 医院、偏远村庄和微电网仍是营销用例 | 可能打开多元化、小批量渠道,但目前更多是叙事而非证据 | 要求实际试点、客户名称,以及旗舰垂直以外存在采购周期的证据 |
| 国防负载的规模和运营模型 | ANPI 公开材料讨论 3–10 MWe 设施,而 Kaleidos 约为 1–1.2 MWe | 如果需要多台机组或定制架构,合同复杂度和时间都可能上升 | 要求 Buckley 系统架构、机组数量、冗余设计和服务义务 |
这张表关注当前具名交易对手能否变成可复制、多元化收入,而不是产品听起来是否有吸引力。公开合同经济性未披露,因此影响只能定性。
[CU019, CU023, CU031, CU037, CU038, CU039]矩阵对比 Equinix、Buckley 防务路径和未具名营销场景的证据质量。
公开记录稀疏且多为定性信息,因此单元格用文本概括公开证据质量,而不是给数值分。
[CU020, CU021, CU032, CU037, CU041, CU042]6.4 留存不透明和集中度主导投资判断
公开客户投资判断仍主要受 Radiant 未披露内容支配。所有已审阅来源都没有给出 NRR、GRR、流失、续约率、合同长度、满意度分数、客户现场已装机可用性,或更广的付费客户数量。也没有来源显示某个客户现场已经在发电。公开叙事因此仍是一笔带押金的预订单、一条具名联邦路径,以及一大圈可能以后转化的营销用例。落实到投资风险,集中度很高:如果 Equinix 放慢、重定价或缩小采购,或者 Buckley 在 NEPA、许可或采购上滑期,公开需求叙事会变薄得多。更广客户名单缺失,也很难判断 Radiant 是在搭建可复制销售机器,还是只拿下少数旗舰关系。在部署站点、定价、合同期限和复购证据披露之前,客户质量仍是由少数有前景锚点支撑的投资逻辑,而不是广泛商业基础支撑的事实。[CU022, CU023, CU027, CU028, CU037, CU038]
| 指标 | 数值 / null | 细分 | 置信度 | 尽调追问 |
|---|---|---|---|---|
| 净留存率(NRR) | 未披露 | 全部 | 高 | 任何部署开始产生收入后,提供过去 12 个月 NRR |
| 总留存率(GRR) | 未披露 | 全部 | 高 | 提供 GRR,把流失和扩张拆开看 |
| Logo 流失 | 未披露 | 全部 | 高 | 按年份提供流失交易对手或取消项目 |
| 合同续约率 | 未披露 | 商业 + 国防 | 高 | 提供续约率,并说明订单是一次性、PPA 还是服务合同 |
| 平均合同长度 / 期限 | 未披露 | 商业 + 国防 | 高 | 提供合同期限、续约选项和终止权 |
| 客户满意度 / NPS | 未披露 | 全部 | 高 | 部署出现后,提供客户访谈、调研数据或案例研究 |
| 已安装客户运行可用率 | 未披露 | 全部 | 高 | 首个客户送电后,提供站点级运行可用率和停机历史 |
| 复购订单 / 扩张订单 | 未披露 | 商业 + 国防 | 高 | 提供后续设备订单、额外基地或站点扩张证据 |
公开客户记录还太早、太薄,撑不起传统留存分析。「未披露」本身就是主要发现:尚无公开的流失、续约、满意度或运行可用率证据。
[CU022, CU023, CU027, CU043]6.5 图表
07风险
7.1 许可、环境审查和授权风险
Radiant 的授权路径比一年前具体得多,但仍是层层堆叠,并未完成。R-50 Part 70 审查把工厂加燃料概念推进到正式 NRC 审查里,但已受理申请覆盖的是加燃料建筑,而不是 Kaleidos 反应堆本身,NRC 也已表示仍预计需要单独的反应堆申请。部署侧,Air Force 与 Buckley 配对是真实需求验证,但官方下一步是 NEPA 下的选址和环境分析,而不是施工授权。NRC 拟议的 Part 57 框架最终可能让合格设计的重复微反应堆许可更便宜、更快,但它仍是拟议规则,不是可融资的权利。DOE 新的先进反应堆类别排除可以加快 DOE 控制的行动,但法律分析明确指出,它不会取消 NRC、州、部落或地方审查。投资人的核心结论是,许可风险仍有多道关:工厂批准、反应堆批准、运输处理和站点接受都必须通过。[CR001, CR002, CR003, CR004, CR005, CR006]
| 风险或关口 | 司法辖区 / 监管机构 | 当前状态 | 可能性 | 严重性 | 缓释状态 | 剩余暴露 | 尽调路径 |
|---|---|---|---|---|---|---|---|
| R-50 Part 70 审查已受理,但详细安全和环境审查仍未完成 | NRC / Tennessee / 联邦 | 2026 年 5 月受理审查;评估仍在进行;预计还会有单独的 Kaleidos 申请 | 高 | 关键 | 进行中——正式审查已开始,但尚未获批 | 关键——工厂装料是商业化时间表的核心 | 要求提供 NRC 往来函件、RAI 节奏,以及 R-50 和 Kaleidos 两项申请的目标日期 |
| Buckley 部署仍取决于选址和 NEPA 分析 | DAF / Space Force / 本地利益相关方 | 项目推进到具名基地,但站点级分析仍在后面 | 高 | 高 | 部分——政府赞助和任务理由清晰 | 高——部署时间仍可能显著滑后 | 要求提供 Buckley 环境审查范围、社区沟通计划和里程碑日历 |
| Part 57 仍是拟议规则,尚未最终定稿 | NRC / Federal Register | 2026 年评论期和规则制定仍在进行 | 中 | 高 | 部分——规则背后政策动能强 | 高——融资假设可能跑在最终文本之前 | 按 Parts 50/52/53 和现行实践建模,不要假设 Part 57 最终条款 |
| DOE CATEX 不能替代 NRC、州、部落或地方审查 | DOE / NRC / 州和部落主管机构 | 仅对 DOE 行动生效,并有逐案限制 | 高 | 中 | 部分——符合条件项目的 DOE 侧时间表可改善 | 中——非 DOE 审批仍可能主导时间表 | 为每个站点梳理完整许可栈,而不只是面向 DOE 的审批 |
| 装燃料微反应堆运输和退役要求仍是活跃监管议题 | NRC / DOT / 未来东道州 | 2026 年框架仍在搭建 | 中 | 高 | 低——有政策工作,但没有成熟机队先例 | 高——返程物流经济性可能在后期恶化 | 获取运输容器假设、退役资金计划和乏燃料返回路径 |
| 未来站点的公众干预或法律挑战风险 | 地方 / 州 / 社区流程 | Buckley 担忧和 Wyoming 先例显示社会许可风险仍在 | 中 | 中 | 低——尚无公开证据显示 Radiant 已有扎实社区沟通体系 | 中——即便联邦政策保持友好,单个站点仍可能放慢 | 审查利益相关方地图、本地外联计划和争议站点法律预算假设 |
各行按剩余严重性和可能性排序。本登记表聚焦会直接影响 Radiant 2026-2028 商业化路径的审批和公共流程关口,而不是泛泛的核监管。
[CR001, CR004, CR005, CR006, CR008, CR009]剩余严重性矩阵显示,许可 / 授权、燃料供应和融资不透明风险落在最高的可能性-影响组合区间;社区反对和替代电源压力也重要,但中心性略低。
[CR001, CR013, CR022, CR031, CR044, CR051]7.2 燃料、工厂和首台套执行风险
燃料仍是 Radiant 故事里最难的外部依赖。DOE 继续表示国内供应商无法供应 HALEU,甚至向 Radiant 分配材料的项目也明确把自己定位为私人供应出现前的过渡桥梁。2026 年给 Radiant 的原料交付对决策有用,因为它证明公司有拿到首炉堆芯的路径,但还不能证明批量部署级供应。Centrus 仍受 DOE 指令和拨款影响;BWXT 已交付的 TRISO 堆芯绑定 Project Pele,而不是广泛商业队列;BWXT 与 Kairos 的合作本身也仍是在建设未来商业产能。测试项目同样没有消除这一风险:DOME 由自筹资金支持,取决于燃料可得性和监管计划,只证明一次边界清晰的测试活动,而不是多年客户可用性。同时,Radiant 还想把这条测试路径转化为 $280 million 工厂,并长期做到每年 50 台反应堆。这是一次首台套制造跃迁,QA、招聘和物流风险都很明显。[CR013, CR014, CR015, CR016, CR017, CR018]
| 失效模式 | 可能性 | 严重性 | 缓释成熟度 | 剩余暴露 | 未解缺口 |
|---|---|---|---|---|---|
| 美国国内 HALEU 供应仍稀缺,靠 DOE 过桥 | 高 | 关键 | 低——DOE 配额存在,但私人市场仍薄 | 关键——延迟可能卡停测试并拖慢机队计划 | 首个启动堆芯之外的燃料没有公开可见度 |
| Radiant 当前原料路径看起来只针对首个堆芯,而非机队规模 | 高 | 高 | 低——2026 年有 Standard Nuclear 加工路径 | 高——扩产经济性仍不透明 | 需要多堆芯合同、备选加工方和交期可见度 |
| 商业 TRISO 制造仍在超越国防探路项目的扩产阶段 | 中 | 高 | 低-中——BWXT 已交付 Pele 燃料,并在建设未来产能 | 高——商业排队深度仍不确定 | 需要商业产出、产能和定价细节 |
| DOME 测试有边界且自筹资金,不等同于多年现场可靠性 | 高 | 高 | 中——测试床和时间表真实存在 | 高——DOME 之后长期运行可用率仍未证实 | 需要 DOME 后现场测试计划、服务假设和备用机组策略 |
| R-50 工厂从原型爬坡到每年 50 台反应堆,带来 QA 和良率风险 | 中 | 关键 | 低——资金和州支持存在,但没有核机队制造先例 | 关键——良率不达标或 QA 漏检会同时冲击成本和进度 | 需要产能爬坡模型、QA 人员计划和供应商认证状态 |
| 返厂物流和换料闭环尚未在运营中跑通 | 中 | 高 | 低——商业模式有描述,但未演示 | 高——停机、运输或周转失败会损害客户经济性 | 需要端到端物流计划、运输假设和换料周转目标 |
这张表关注从 2026 年测试成功到可复制商业交付体系之间的执行风险。多数缓释仍是计划或政策过桥,不是成熟运营证据,因此剩余暴露偏高。
[CR013, CR014, CR015, CR016, CR017, CR018]关键依赖把 DOE 燃料项目、当前 TRISO 转化路径、NRC 审批、Tennessee 工厂,以及 Radiant 两个标杆交易对手连在一起。
[CR015, CR016, CR018, CR021, CR022, CR024]7.3 客户集中、防务依赖和替代电源竞争
Radiant 的公开需求证明仍高度集中。商业侧仍由 Equinix 主导,防务侧仍由 Buckley 路径主导。集中度之所以重要,是因为两个交易对手都具战略意义,但没有任何一方公开披露足以让外部投资人测算经常性现金流的经济条款。同样重要的是,Equinix 并不是只押注 Radiant。它的公开能源战略包括 Bloom 燃料电池、电网升级、备用基础设施,以及多个反应堆形态和时间线不同的核能开发商。这种多元化符合 Equinix 的理性选择,但也意味着 Radiant 不仅要和其他微反应堆供应商竞争,还要和可更早部署的现场燃料电池及其他电力可用性替代方案竞争。Bloom 自己的 2026 年调查指向一个越来越重视交付确定性和现场电力、而不是特定发电技术的市场。防务依赖也有同样的问题:DIU 和 Air Force 是有意义的验证者,但这个项目也是产业政策工具,目标是刺激整个供应商基础,而不是给单一供应商未来订单担保。[CR027, CR028, CR029, CR030, CR031, CR032]
| 依赖项 | 在投资逻辑中的角色 | 集中度 / 可替代性 | 主要失败场景 | 严重性 | 缓释状态 | 剩余暴露 |
|---|---|---|---|---|---|---|
| DOE HALEU 计划 | 为首批部署弥合燃料缺口 | 集中度极高;联邦过桥 | 配额放慢、优先级变化或政策支持减弱 | 关键 | 低——当前支持强,但只是临时性的 | 关键 |
| Standard Nuclear 和当前原料转化路径 | 把 DOE 提供材料转成 Radiant 燃料 | 集中度高;公开记录中仅限首个堆芯路径 | 加工或认证延迟打断 2026 年启动 | 高 | 低——只看到一条路径,替代方案未公开 | 高 |
| NRC Part 70 + 未来 Kaleidos 许可 | 支撑工厂装料和反应堆部署 | 集中度高;没有替代监管机构 | 审查滑出 2028 计划或施加设计变更 | 关键 | 中——正式审查已开始 | 关键 |
| Equinix | 最强公开商业需求锚点 | 集中度高;对方自身客户多元 | 预订单缩小、延迟,或在其他能源选项面前优先级下降 | 高 | 中——关系真实,但经济性未披露 | 高 |
| DAF / DIU / ANPI | 最强公开国防路径和可信度信号 | 集中度高;项目有竞争性且受政策驱动 | 预算、NEPA 或采购变化推迟 Buckley | 高 | 中——官方分配已经存在 | 高 |
| Tennessee / Oak Ridge 生态 | R-50 的州支持、核劳动力和场地背景 | 集中度中;一旦投入,搬迁成本高 | 建设、许可或劳动力爬坡不及预期 | 中 | 中——州支持可见 | 中 |
最关键的依赖不是普通供应商,而是项目对手方和监管方;它们可能同时牵动多项风险。Radiant 缺少公开替代方案或合同细节的地方,剩余敞口最高。
[CR003, CR015, CR016, CR018, CR027, CR033]| 选项 | 公开部署状态 | 买方可能偏好的原因 | 为什么仍给 Radiant 留出空间 | Radiant 风险 |
|---|---|---|---|---|
| Bloom 燃料电池 | 现在已达到 Equinix 规模(75 MW 运行;另有 30 MW 在建) | 现在可用、可在现场部署,并已在客户环境验证 | 续航弱于多年核燃料循环;仍暴露在燃料和设备经济性风险下 | 近期竞争风险高 |
| 电网现代化与公用事业升级 | 已进入客户计划,但关键枢纽进展慢于预期 | 沿用熟悉的许可路径和公用事业对手方 | 交付时间往往跟不上 AI 增长节奏 | 进度缺口仍在,因此风险中等 |
| 天然气与现场混合发电 | 属于 Equinix 更广泛电力组合讨论的一部分 | 部署快、可调度 | 暴露于燃料价格和排放风险;脱碳叙事较弱 | 中等风险 |
| 其他先进核能厂商(Oklo、ULC-Energy、Stellaria、Westinghouse、Antares) | 公开合作或基地分配已经可见 | 客户可以分散反应堆技术和地理风险 | 如果 Radiant 率先落地,仍能守住便携式柴油替代场景的小生态位 | 战略风险高 |
| 等电网 / 推迟场址策略 | 始终可作为默认选项 | 避开首台同类反应堆执行风险 | 让客户继续暴露在电力稀缺和增长损失中 | 中等风险 |
这张表不是 LCOE 对比,而是基于已审阅来源做出的决策时点对比。到 2026 年,Radiant 的主要竞争威胁不是另一款完全相同的便携反应堆,而是能更早解决电力稀缺的替代方案。
[CR027, CR028, CR029, CR030, CR031, CR032]7.4 公众接受度、运输、废弃物和扩散担忧
Radiant 的便携式反应堆模式没有绕开社会许可问题,反而放大了它。围绕 Buckley 的公开报道已经显示,至少一些附近居民对公告感到意外,并立即关注废弃物和长期环境影响。Wyoming 事件也基于同一个原因重要:当地公共阻力不只是抽象的反核意识形态,还来自问责、废弃物和出事后谁承担下行风险等问题没有答案。便携部署也改变了风险形态。NRC 和行业评论现在明确区分新燃料机组运输,以及把已辐照机组运回换料或寿期末处理这一更难问题。这是 Radiant 商业模式的核心,不是边缘问题。最后,HALEU 不只是供应瓶颈,也是政治和安全风险。UCS 和其他批评者认为,HALEU 安全应比当前公开叙事所暗示的更严肃地处理。如果这种批评转化为更严格的保障措施或出口管制,Radiant 的成本和物流优势会收窄。[CR039, CR040, CR041, CR042, CR043, CR044]
授权、燃料、社会许可、替代电源和政策风险,如何传导为进度滑坡、客户转化减弱和估值压缩。
[CR001, CR016, CR027, CR031, CR041, CR051]7.5 融资不透明、执行人员配置和失败条件监控
Radiant 看起来不是缺钱,而是披露不足。公开记录支撑一笔很大的 2025 年融资、一个大型 Tennessee 工厂项目,以及足以让故事继续融资的政治和客户牵引力。但这些记录仍不足以支撑常规经营层面的投资判断。公开材料没有披露当前现金、烧钱速度、现金跑道、毛利率、单台反应堆经济性或客户付款条款。因此投资人无法仅凭公开证据判断,当前资本基础能否同时稳妥覆盖 DOME、R-50、许可延迟和营运资本需求。公司还必须同时搭建几块执行肌肉:核级制造 QA、材料管控、返厂服务模式下的现场物流,以及横跨监管和商业里程碑的项目控制。正确的风险缓释姿态因此应以里程碑为基础。在 Radiant 证明加燃料测试成功、首炉堆芯之外燃料可持续取得、许可继续推进,并更清楚披露现金需求和旗舰客户经济性之前,它仍是一个有前景但仍靠里程碑融资的基础设施投资逻辑。[CR049, CR050, CR051, CR052, CR053, CR054]
| 角色或职能 | 依赖或缺口 | 发生概率 | 严重性 | 当前缓释措施 | 尽调路径 |
|---|---|---|---|---|---|
| 许可与监管项目管理 | 必须并行推进 Part 70、未来 Kaleidos 许可和具体场址审批 | 中 | 关键 | Part 70 审查获受理,已能证明部分进展 | 索取整合后的许可时间表、人员配置和外部律师支持 |
| 特殊核材料与燃料循环运营 | 工厂模式量产前,必须先跑通材料管控、装料和处理纪律 | 中 | 高 | 早期 R-50 审查和 DOE 燃料项目提供探路支持 | 索取运营组织架构、培训计划和材料核算控制 |
| 核级制造 QA | 年产 50 台反应堆的目标,需要比原型阶段更强的 QA 体系 | 中 | 关键 | 州政府支持的工厂建设和招聘计划已经可见 | 索取供应商认证清单、QA 框架和预期一次通过率目标 |
| 现场物流与返厂服务 | 便携模式需要运输、停机响应和周转能力;这些能力尚未公开验证 | 中 | 高 | 目前还没有公开的服务网络证据 | 索取运输合作伙伴、服务覆盖假设和周转 KPI |
| 资本规划与项目控制 | DOME、许可、R-50 和客户项目都会在收入公开前消耗资本 | 高 | 关键 | 公司披露了大额融资,但没有烧钱速度或现金跑道细节 | 索取月度烧钱速度、下行情景现金跑道、应急预算和里程碑挂钩支出计划 |
这张表强调执行职能,而不是具体个人,因为公开披露更多落在项目里程碑,较少覆盖管理层的日常运营节奏。若一个能力缺口会同时冲击许可、成本和客户信任,严重性就高。
[CR024, CR050, CR051, CR053, CR054, CR055]| 风险 | 可监测触发项 | 阈值 / 事件 | 行动含义 | 当前状态 |
|---|---|---|---|---|
| 首炉芯之后的燃料供应 | 公开披露第二炉芯或机队燃料合同 | 首次 DOME 启动前仍未披露可信的启动后燃料路径 | 将 2028 年部署时间视为缺乏支撑 | 未解决 |
| DOME 执行 | 测试时间表、启动和六个月试验活动完成情况 | 明显滑期,或试验活动数据实质性不完整 | 重新承销产品成熟度,并推迟客户时间表 | 未关闭,且高度可监测 |
| R-50 与 Kaleidos 许可 | NRC 里程碑、RAI,以及工厂与反应堆审批之间是否脱钩 | 工厂审查推进,但反应堆申请明显滞后 | 假设工厂无法按当前时间表转化为销售 | 未关闭 |
| Buckley 路径 | NEPA 进展、公开会议和基地层面时间表 | 环境程序停滞,或社区阻力升级且没有缓释措施 | 下调国防转化概率和战略溢价 | 未关闭 |
| Equinix 承诺质量 | 场址披露、部署里程碑和合同细节 | 替代方案持续扩张,但没有新的具体场址进展 | 将商业需求标记为可选项,而非锚定需求 | 未关闭 |
| 工厂 capex 与劳动力爬坡 | R-50 建设进展,以及相对 175 个岗位计划的招聘情况 | 建设延期,或出现可见的 QA / 劳动力瓶颈 | 上调成本超支和进度滑期假设 | 未关闭 |
| 财务不透明 | 管理层披露烧钱速度、现金跑道和经济性 | 又一轮融资发生,但基本运营透明度仍缺失 | 将资本可得性视为唯一真实缓冲,并下调估值信心 | 未解决 |
| 政策与安全环境 | HALEU 保障、出口管制或联邦项目支持发生变化 | 安全收紧或政策反转抬高成本,或移除进度支持 | 一次性重估可触达市场和时间表 | 未关闭 |
触发项选择可由外部公开证据监测的信号,而不是依赖管理层叙事。这里一旦触发投资逻辑破裂,新增里程碑标题就不应再替代更彻底的重新承销。
[CR022, CR024, CR027, CR044, CR045, CR050]7.6 图表
08估值
8.1 价格信号与证明缺口
Radiant 的估值讨论必须先区分已确认事项和只是被标记的事项。已确认的内容很可观:公司公开披露 $165 million Series C 轮,使累计风险融资达到 $225 million;随后又披露 2025 年 12 月超过 $300 million 的融资轮。外部人因此看到高于 $525 million 的公开资本底线,也让“Radiant 已吸引严肃后期风险资本支持”有了真实依据。未被同样确认的是有监管文件支撑的投后估值。最常被引用的 >$1.8 billion 数字来自 DCVC 和媒体报道,而不是公开证券文件或独立清算的二级市场成交价。估值来源差异很重要,因为同一公开记录仍没有披露收入、毛利率、烧钱速度、现金跑道、客户定价或清算优先权。换句话说,价格信号很强,但支撑它的证据栈仍主要是融资动能、具名客户和里程碑进展,而不是经营经济性。[CV001, CV002, CV003, CV004, CV005, CV006]
| 维度 | 评估 | 证据基础 | 决策含义 |
|---|---|---|---|
| 建议 | 继续研究 / 观察 | 融资和客户信号强,但缺少运营财务承销 | 没有更多证据或更好入场价前,不应把当前估值标记当作可行动信号 |
| 信心 | 中 | 多数核心事实已公开,但决定性经济性仍是私有信息 | 在私有数据补上最大缺口前,把确信度压在买入阈值以下 |
| 风险评级 | 高 | 燃料、许可、制造复制和披露缺口都可能冲击估值 | 先承销下行;假设仍可能需要更多资本 |
| 估值立场 | 当前 >$1.8B 标记下偏高,最好也只是合理 | 当前信号位于宽泛估算基准区间的中部附近 | 宁可等待,不追逐动量 |
| 偏好的入场纪律 | 等更多证据,或等更低价格 | 证据 = 测试成功、许可进展、客户转化和经济性披露 | 只有证据改善快于价格时,才积极重启评估 |
| 可能退出姿态 | 仅靠公开数据,现在还太早承销清晰 IPO 或战略退出回报 | 商业交付指引仍在 2028 年及以后 | 把退出测算视为高度依赖情景,而非可入账收益 |
这些评估是定性的,并且明确对价格敏感。表格把已确认融资信号,与收入转化、利润率和稀释方面未获支撑的假设分开。
[CV003, CV004, CV006, CV016, CV040, CV041]| 视角 | 当前判断 | 支撑证据 | 什么会改变判断 |
|---|---|---|---|
| 资本可得性 | 投资逻辑 | >$525M 公开融资下限和新近 >$300M 融资显示融资需求强 | 平轮 / 下轮融资,或惩罚性优先股堆叠,会很快削弱投资逻辑 |
| 客户证据 | 带保留的投资逻辑 | Equinix 订金和 Buckley 路径构成有意义的早期验证 | 详细订金条款、铺开时间表或场址承诺,会让该信号更可入账 |
| 运营证据 | 反向逻辑 | 没有公开收入、利润率或烧钱速度披露支撑当前估值标记 | 若披露收入桥和单位经济性材料包,确信度会实质性改善 |
| 监管进展 | 带保留的投资逻辑 | DOME 路径、已受理的 Part 70 审查和 DOE 支持显示实际推进 | 反应堆许可滑期,或延后到请求时间表之后,会压缩溢价 |
| 板块背景 | 混合 | 公私市场同业显示,在现金流出现前,投资人也会为核能期权价值付费 | 如果泡沫担忧消退,且同业在首批测试后仍守住价值,Radiant 的溢价会更容易辩护 |
| 估值标记可靠性 | 反向逻辑 | >$1.8B 数字来自投资人 / 新闻报道,而非独立确认 | 经审计的股权结构表、新融资文件,或可信独立估值标记,会增强价格信心 |
这张表刻意区分两类论据:一类支撑公司质量,另一类支撑今天的价格。收入前反应堆开发商不能把两者混为一谈。
[CV003, CV004, CV008, CV016, CV017, CV018]建议先承认真实融资、客户和监管进展,再扣除经济性缺口、燃料依赖和投机价格风险,最终落到继续研究 / 观察。
这是定性决策图,不是加权评分模型。
[CV003, CV008, CV012, CV016, CV017, CV040]8.2 可比锚点和按里程碑调整的估值逻辑
公私市场核能可比公司解释了 Radiant 为什么能拿到十亿美元叙事、尽管还没有证明商业反应堆收入;它们也说明投资人为什么要抵制伪精确。公开市场可比公司是最清晰的警示标签。Oklo 的公开估值远高于 Radiant 据报道的标记,尽管收入仍列为 n/a;NuScale 估值低于 Oklo,但收入有限且稀释很重,仍以数十亿美元估值交易。NANO Nuclear 说明,即便是更早期、收入前的微反应堆故事,只要投资人为期权价值定价,市值也能超过 $1 billion。私募轮也强化同一教训:X-energy 的 $700 million 融资、Helion 的 $5.425 billion 投后估值,以及 Last Energy 的 $100 million Series C 轮,都显示资本正在追逐先进核能里程碑。但这些可比公司都无法完全解决 Radiant 的估值问题,因为每家公司在反应堆尺寸、监管成熟度、客户类型或披露质量上都不同。正确方法因此是按里程碑调整的比较,而不是用电子表格假装 Radiant 已经有可知现金流。[CV017, CV018, CV021, CV022, CV023, CV024]
| 可比对象 | 阶段 / 指标 | 估值或状态 | 为何与 Radiant 相关 | 主要局限 |
|---|---|---|---|---|
| Radiant Nuclear | 私有微反应堆开发商;无公开收入披露 | >$300M Series D 轮;>$1.8B 投资人 / 新闻报道估值标记 | 直接研究对象;检验商业收入出现前,当前市场愿意为数据中心 + 国防叙事支付多少 | 当前价格信号没有监管文件支撑,经济性仍是私有信息 |
| Oklo | 上市先进核能 / 微反应堆开发商;收入 n/a | 2026-05-21 市值 ~$11.33B;EV 约 $9.00B | 显示反应堆收入出现前,公开市场可能为 AI 相关核能期权价值支付高价 | 公开流动性、Sam Altman 光环和更大规模,让它成为激进溢价可比对象 |
| NuScale | 上市 SMR 开发商,有少量收入且大幅稀释 | 2026-05-21 市值 ~$4.14B;滚动收入 ~$18.67M | 为核能估值提供一个收入更多、监管更重的基准 | 更大的公用事业式模块、不同客户群和重大股份稀释降低可比性 |
| NANO Nuclear | 上市微反应堆开发商;收入前 | 市值 ~$1.31B;SEC 称截至报告日尚未产生收入 | 显示即使早期微反应堆股权,也能达到十亿美元级期权价值 | 阶段更早、规模更小,商业验证也弱于 Radiant |
| X-energy | 私有 HTGR / 燃料开发商 | 2025 年 11 月 Series D 轮 ~$700M;估值未披露;声称订单簿 >11 GW | 可作为资本强度、燃料和商业化叙事的私有参照 | 反应堆更大、工业客户更大,且披露了订单簿;这些都不等同于 Radiant 当前状态 |
| Helion | 私有聚变开发商 | Series F 轮,投后估值 $5.425B | 提醒上限:前沿核能估值可以远超当前收入 | 聚变不是裂变,技术与政策时间线也不同 |
| Last Energy | 私有微反应堆 / 小型反应堆开发商 | >$100M Series C 轮;估值未披露 | 可作为工厂建造核能故事中较小规模私有融资参考 | PWR 架构、监管路径不同,且未披露估值 |
这组可比对象刻意混合而非穷尽:Radiant 仍处收入前且披露较少,所以这里把公开收入前可比公司和私有里程碑融资放在一起。估值是市场快照或最近披露的轮次数据,不是标准化运营价值倍数。
[CV004, CV022, CV023, CV024, CV025, CV026]示意性的中点估值估计显示,最大驱动因素不是表格倍数,而是 Radiant 在下一轮融资前能过多少里程碑关口。
数值为估计的百万美元,依据里程碑调整后的可比公司比较,不是 DCF 输出或管理层指引。
[CV037, CV038, CV039, CV043, CV044]8.3 当前标记下的乐观、基准和悲观情景
Radiant 的情景分析应被框定为一组宽区间,并明确里程碑假设,而不是一个细到小数点的目标价。乐观情景需要的不只是正面标题:2026 年 DOME 活动成功、NRC 在工厂路径上持续推进、HALEU 可持续取得,并证明 Equinix 和防务路径正在转成付费部署,而不只是战略验证。基准情景更温和:Radiant 保持动能、仍可融资,并因资本获取和旗舰客户得到信用,但运营数据披露仍不足以让当前标记显得明显便宜。悲观情景不只是“核能很难”,而是多个依赖同时滑期,公司在商业收入出现前还必须穿越延期继续融资。在这种框架下,目前据报道的 >$1.8 billion 价格信号已经位于宽泛估计基准区间的中部附近,这意味着仅靠公开证据,并不能支持以今日标记进入是有吸引力的。投资人需要更低价格、更好的私有证据,或两者兼具。[CV008, CV009, CV010, CV011, CV012, CV013]
| 情景 | 核心假设 | 估算估值区间(USDm) | ~1.8B 入场下的回报逻辑 | 概率信号 |
|---|---|---|---|---|
| 悲观 | DOME 或燃料滑期,许可漂移,客户转化仍不透明,且收入证据出现前还需要再融资 | 900–1300 | 以当前入场价大概率亏损;即使 1.0x 也难支撑 | 如果两条或更多里程碑链条一起滑期,概率就有实质分量 |
| 基准 | Radiant 保持融资通道,达成部分里程碑,但仍缺少足够经济性披露,无法支撑溢价重估 | 1400–2000 | 相比当前标记水平,大致持平到温和上行 | 从今天公开证据看,这是最可能的情景 |
| 乐观 | 2026 年测试成功,NRC 继续推进,燃料路径更稳,旗舰需求明显转化为付费部署 | 2100–3000 | 存在上行,但从当前标记看,仍不是典型风投式爆发回报 | 需要多个证据闸门放行,而不是一个标题就够 |
区间是明确估算,不是 DCF 输出。它们是与公开收入前同业和私有先进核能融资轮对比后,按里程碑调整得出的结果;由于 Radiant 不披露做更精细估值所需的运营指标,不确定性区间很宽。
[CV037, CV038, CV039, CV041, CV042, CV043]悲观、基准、乐观估值区间显示,当前标记已接近估计基准情景中部,并非处在明确安全边际折价上。
区间是依据上市和私营核能可比公司推导出的估计百万美元,并明确提示 Radiant 运营指标缺失。
[CV041, CV042, CV043, CV044]IC 式评分显示,Radiant 在资本获取和市场需求上最强,客户与监管证据居中,经济性披露和估值支撑最弱。
评分为分析师依据本章审阅的公开证据给出的 1–5 分判断。
[CV016, CV031, CV032, CV036, CV041, CV042]8.4 打破投资逻辑的触发条件和仍缺失的证据
现在最有用的估值工作,不是猜 2029 年 EBITDA 利润率,而是识别哪些证据会真正改变判断。公开证据中仍有几个问题明显悬而未决。Equinix 预订单真实存在,但押金规模、是否可退、排他性和积压订单转化并未披露。Radiant 的融资动能真实存在,但烧钱速度、现金跑道、当前现金和优先股堆叠不公开。监管进展真实存在,但 Part 70 提交并被受理,不等于商业反应堆运营许可。燃料获取正在改善,但 DOE 仍称国内供应商无法供应 HALEU。上述缺口直接影响估值,因为它们决定当前标记只是激进,还是确实缺乏支撑。正确尽调姿态因此应以触发条件为基础:如果 DOME 滑期、许可节奏漂移、下一轮融资暴露昂贵结构,或旗舰客户经济性在下一个里程碑周期后仍不透明,估值标记就应压缩。反过来,如果相反情况发生,估值论据可能很快变强。[CV019, CV020, CV034, CV035, CV036, CV038]
| 触发项 | 阈值 / 事件 | 如何传导到投资逻辑 | 行动含义 |
|---|---|---|---|
| DOME 时间滑期 | 测试看起来不再是 2026 年事件 | 移除支撑今天溢价的最近硬证据点 | 在新的已验证时间表出现前,转入明确下行情景 |
| 燃料脆弱性 | HALEU 获取恶化,或仍只覆盖首炉芯 | 把时间风险变成融资风险,并推迟付费部署 | 假设需要更多资本;下调估值区间 |
| 许可漂移 | Part 70 审查明显晚于请求时间,或反应堆许可路径变弱 | 降低工厂投入能按期转化为可部署产品的信心 | 压缩情景区间,并下调基准情景 |
| 客户转化失败 | Equinix 订金条款变弱、节奏滑期,或预订未转成付费铺开 | 削弱当前估值标记背后最强的商业证据 | 在重新证明前,把客户逻辑视为宣传信号 |
| 结构化融资冲击 | 下一轮暴露重度优先权、重大稀释或下轮定价 | 显示当前估值标记没有以干净条款通过市场检验 | 在堆叠经济性明确前,把建议重置为回避 |
| 重大里程碑后披露仍缺席 | 测试或许可进展之后,公司仍不披露收入、现金和单位经济性 | 表明管理层想拿估值信用,却不给承销透明度 | 没有数据室证据前,不支付溢价倍数 |
这些是可监测的估值触发项,不是泛泛的公司风险。每一项都会直接改变商业化概率,或改变企业价值中现实可分给新投资人的份额。
[CV010, CV011, CV013, CV018, CV034, CV035]| 主题 | 缺失证据 | 重要性 | 负责人或尽调路径 |
|---|---|---|---|
| 股权结构表与优先权 | Series D 后股数、证券条款、清算堆叠,以及任何投资人保护 | 决定今天的头条估值标记是否能转化为真实普通股价值 | 向法律顾问或 CFO 索取股权结构表、条款清单摘要和优先权瀑布 |
| 现金、烧钱速度与现金跑道 | 当前非受限现金、月度烧钱速度,以及滑期情景下的下行现金跑道 | 单靠已融资本身,不能证明足以覆盖测试、许可和工厂支出 | 索取最新资产负债表、预算和董事会融资触发备忘录 |
| 订金经济性与积压订单转化 | 单台订金规模、可退款性、里程碑、取消权和排他条款 | 用来区分真实商业去风险和战略信号 | 索取已签署的 Equinix 条款清单和管理层收入确认备忘录 |
| 单位经济性 | 反应堆报价、PPA 电价、毛利率桥,以及服务 / 换料成本堆叠 | 没有这些,任何精确回报或退出模型都不可信 | 索取客户定价材料、带成本的 BOM 和毛利率敏感性分析 |
| 独立估值标记支持 | 任何支持当前定价的第三方 409A、老股交易、要约收购或外部估值标记 | 区分实时市场清算估值和投资人叙事 | 索取最新董事会估值标记包和任何老股交易证据 |
| 客户部署经济性 | 旗舰客户的场址时间表、电网对比经济性和付款时点 | 决定客户证据能否真正以可接受利润率转化为现金 | 索取逐场址管线复盘,包含合同状态和预期现金里程碑 |
每一行都对应一个缺失输入,足以让估值从偏高变成可辩护,或从激进变成缺乏支撑。承销新资金前,这些是最高优先级的尽调问题。
[CV034, CV035, CV036, CV040, CV045]8.5 图表
免责声明
本报告是基于公开证据的尽调快照,不构成投资建议。重要财务、法律、技术和合同事实仍未公开;作出任何投资决定前,应直接向管理层和一手文件核验。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Radiant was founded in 2020. | 高 | SO002, SO005 |
| CO002 | Radiant is headquartered in El Segundo, California. | 高 | SO002, SO007 |
| CO003 | Radiant is a private company developing portable, mass-produced nuclear microreactors intended to replace diesel generators. | 高 | SO001, SO018 |
| CO004 | Company materials describe Kaleidos as a 1 MWe microreactor that uses TRISO fuel and helium gas coolant. | 高 | SO001, SO018 |
| CO005 | Kaleidos uses passive air cooling without on-site water and is designed to travel in a containerized format by truck or aircraft. | 高 | SO001, SO018 |
| CO006 | Radiant says a Kaleidos unit can run for five or more years before refueling and support a 20-year product lifetime via factory refueling cycles. | 高 | SO001, SO018 |
| CO007 | Radiant says the first fueled Kaleidos test is targeted for 2026 at DOME, with initial customer deployments beginning in 2028. | 高 | SO003, SO005 |
| CO008 | Doug Bernauer founded Radiant after a long SpaceX engineering career and remains the company's CEO. | 中 | SO015, SO016 |
| CO009 | Rita Baranwal joined Radiant in June 2025 as its first Chief Nuclear Officer after senior roles at DOE and Westinghouse. | 高 | SO004, SO017, SO026 |
| CO010 | Mike Starrett was publicly identified as Radiant's Chief Revenue Officer by late 2025 and represented the company in the April 2026 Buckley announcement. | 高 | SO003, SO007 |
| CO011 | Tori Shivanandan was publicly identified as Radiant's Chief Operating Officer in February 2026. | 中 | SO009 |
| CO012 | Radiant said in December 2025 that it had added almost a dozen VP- and director-level hires across engineering, manufacturing, and supply chain over the prior six months. | 高 | SO003, SO015 |
| CO013 | Radiant's public narrative remains tightly centered on Bernauer, creating material key-person dependence around fundraising, engineering credibility, and external relations. | 中 | SO014, SO015, SO016 |
| CO014 | Radiant closed a $165 million Series C in May 2025, bringing total venture funding to $225 million, with DCVC leading the round. | 高 | SO002, SO019 |
| CO015 | Radiant announced a new funding round of more than $300 million in December 2025 led by Draper Associates and Boost VC. | 高 | SO003, SO013, SO018 |
| CO016 | DCVC said the December 2025 round valued Radiant above $1.8 billion, but that figure came from an existing investor rather than a neutral market source or filing. | 中 | SO014 |
| CO017 | Series D materials disclosed follow-on participation from Founders Fund, ARK Venture Fund, Chevron Technology Ventures, and other existing investors. | 高 | SO003, SO013 |
| CO018 | Lockheed Martin Ventures made a strategic investment in Radiant in February 2026, adding a named defense-industry investor to the cap-stack narrative. | 高 | SO009, SO003 |
| CO019 | Publicly disclosed rounds imply Radiant had raised more than $525 million in venture funding by December 2025. | 高 | SO002, SO003 |
| CO020 | Equinix signed a preorder agreement and paid deposits in August 2025 for 20 Kaleidos microreactors. | 高 | SO010, SO011, SO012 |
| CO021 | The Equinix deal is the clearest public commercial demand signal for Radiant, but deployment sites and unit economics were not publicly disclosed. | 中 | SO010, SO011 |
| CO022 | The Department of the Air Force and the Defense Innovation Unit selected Radiant in April 2026 for a microreactor project at Buckley Space Force Base, with first reactors targeted for 2028. | 高 | SO007, SO018 |
| CO023 | DOE conditionally selected Radiant in July 2025 for the first DOME microreactor test campaign at Idaho National Laboratory, with testing slated to begin as early as spring 2026. | 高 | SO005, SO022 |
| CO024 | DOE approved Radiant's DARK submission in February 2026 as the second of three safety-document phases needed before startup at DOME. | 高 | SO006, SO020 |
| CO025 | The NRC formally accepted for review Radiant's 10 CFR Part 70 license application in May 2026 for the R-50 production facility in Tennessee and said it would pursue an accelerated review timeline. | 中 | SO008 |
| CO026 | Radiant says the R-50 factory in Oak Ridge, Tennessee is intended to scale to 50 reactors per year within a few years of 2028 production start. | 高 | SO003, SO013 |
| CO027 | DOME is a government-run microreactor test bed at Idaho National Laboratory that allows fueled experiments up to 20 MWt, making access strategically valuable to Radiant's licensing and demonstration path. | 高 | SO022, SO023, SO025 |
| CO028 | DOE conditionally committed HALEU fuel to Radiant as one of five advanced nuclear awardees, but the program draws on scarce federal stockpiles as a near-term bridge. | 高 | SO002, SO021 |
| CO029 | Utility Dive reported that U.S. civilian HALEU production remains constrained by market and infrastructure gaps, implying continued fuel-supply risk for developers including Radiant. | 中 | SO021 |
| CO030 | UCS publicized a Science-related analysis arguing that HALEU above roughly 12% U-235 can present larger proliferation and terrorism risks than previously acknowledged. | 中 | SO024 |
| CO031 | HALEU supply scarcity and proliferation concerns are external risks that could slow Radiant's regulatory, procurement, or public-acceptance timeline even if the company executes technically. | 中 | SO021, SO024 |
| CO032 | Radiant's current revenue or ARR is not publicly disclosed in the reviewed chapter sources. | 低 | |
| CO033 | Radiant's current headcount is not publicly disclosed in the reviewed chapter sources. | 低 | |
| CO034 | Reviewed public sources do not disclose a full board roster, ownership percentages, or investor control rights for Radiant. | 低 | |
| CO035 | Public sources do not disclose exact backlog beyond the Equinix preorder and Buckley pathway, nor do they reveal reactor pricing or gross margin structure. | 低 | |
| CO036 | Radiant positions Kaleidos for military installations, data centers, remote industry, hospitals, disaster response, and remote communities. | 高 | SO001, SO003, SO010 |
| CO037 | Radiant says customers can buy Kaleidos through either power purchase agreements or direct unit sales. | 中 | SO001 |
| CO038 | Radiant and trade press describe Kaleidos as deployable within days or roughly 48 hours, with factory assembly and testing before shipment. | 高 | SO001, SO018 |
| CO039 | Partner portfolio pages characterize Radiant as an execution leader with prepaid demand, but claims such as being sold out through 2030 are not independently corroborated in reviewed public sources. | 中 | SO015, SO016 |
| CO040 | Public evidence supports strong momentum across capital, government testing access, defense selection, and named commercial interest, but Radiant still has no public record of completed reactor operations or disclosed commercial financial performance. | 中 | SO003, SO007, SO010, SO025 |
| CM001 | Reviewed official and regulatory sources define microreactors as a subset of SMRs generally at 20 MWe or less, while Radiant's Kaleidos is framed around roughly 1 to 1.2 MWe and about 3 MWth. | 高 | SM003, SM010, SM015, SM014 |
| CM002 | DOE program materials describe microreactors as transportable reactors for non-conventional markets including remote communities, mining, defense, backup generation, and emergency response. | 高 | SM005, SM003 |
| CM003 | EIA says SMRs and microreactors are under consideration for AI and data centers, industrial activities, and remote areas or communities with high transmission and distribution costs. | 中 | SM015 |
| CM004 | Radiant publicly positions Kaleidos as a diesel-replacement and resilient-power product for remote villages, hospitals, data centers, military installations, and other distributed use cases. | 高 | SM001, SM002, SM003 |
| CM005 | Project Pele, Janus, Eielson, and Buckley together show that military installation power is one of the earliest concrete microreactor demand segments in the United States. | 高 | SM007, SM009, SM015 |
| CM006 | Equinix signed a preorder agreement and paid deposits for 20 Kaleidos reactors, making data-center power the clearest named commercial segment for Radiant in public sources. | 高 | SM016, SM017, SM018 |
| CM007 | Equinix's 2025 nuclear announcements covered more than 774 MWe across Radiant, ULC-Energy/Rolls-Royce SMR, and Stellaria, implying a portfolio approach rather than a single-reactor bet. | 中 | SM017, SM018 |
| CM008 | DOE and Belfer both cite LBNL work showing U.S. data centers used about 176 TWh in 2023 and could reach 325 to 580 TWh by 2028, equal to roughly 6.7% to 12% of U.S. electricity consumption. | 高 | SM022, SM023 |
| CM009 | DOE says data-center load growth has tripled over the past decade. | 中 | SM022 |
| CM010 | Belfer says AI-driven electricity demand is already outpacing available capacity in some regions, leading project delays, direct power contracting, and temporary gas-generator workarounds. | 中 | SM023 |
| CM011 | Belfer says data-center financing relies mainly on parent-company balance sheets, corporate bonds, and incentives, while power procurement increasingly includes PPAs, availability payments, upfront capital payments, and co-location structures. | 中 | SM023 |
| CM012 | DOE explicitly lists onsite generation and storage, transmission expansion, innovative rate structures, advanced nuclear, geothermal, and long-duration storage as part of the response to data-center demand growth. | 中 | SM022 |
| CM013 | World Nuclear Association and EIA both frame SMRs and microreactors as better suited than large reactors to small grids, remote communities, industrial sites, and locations where siting flexibility matters. | 高 | SM014, SM015 |
| CM014 | EIA says high-temperature gas reactor and other advanced designs can serve industrial processes and high-heat applications as well as electricity generation. | 中 | SM015 |
| CM015 | DOE's microreactor program says broad deployment depends on improving both economic viability and licensing readiness, indicating that buyer demand alone is insufficient for market takeoff. | 中 | SM005 |
| CM016 | Project Pele is a 1 to 5 MWe transportable high-temperature gas reactor designed to move in 20-foot containers, showing that the defense segment's target power range overlaps Radiant's class of site-level assets. | 高 | SM007, SM008 |
| CM017 | INL says the Army's Janus Program is intended to follow Project Pele with affordable, reliable commercial nuclear power for critical infrastructure when the electric grid is disrupted. | 高 | SM009, SM015 |
| CM018 | DOME is positioned as the world's first microreactor test bed and as a place to lower development risk for fueled experiments up to 20 MWt. | 高 | SM003, SM004, SM006, SM019 |
| CM019 | NRC pre-application materials show Radiant is pursuing multiple parallel workstreams around regulatory engagement, factory fueling, and materials licensing rather than a single simple approval path. | 中 | SM010 |
| CM020 | EIA says high capital costs and lengthy licensing and approval processes have historically limited nuclear expansion in the United States. | 中 | SM015 |
| CM021 | EIA and DOE describe HALEU as important to many advanced reactor designs because its higher enrichment can support smaller footprints, better performance, and different operating characteristics than standard reactor fuel. | 高 | SM011, SM015 |
| CM022 | DOE conditionally committed HALEU to five advanced-reactor developers including Radiant, using federal material to meet near-term fuel needs. | 高 | SM011, SM012 |
| CM023 | Utility Dive reports that U.S. civilian HALEU production remains constrained by market uncertainties and infrastructure gaps, so domestic supply is still a bottleneck. | 中 | SM012 |
| CM024 | UCS publicized a Science-based critique arguing that HALEU above about 12% U-235 may present greater proliferation and terrorism risk than previously acknowledged. | 中 | SM013 |
| CM025 | INL says Project Pele's recent shipment was the first TRISO microreactor fuel delivered at its final destination, underscoring how early the specialized fuel supply chain still is. | 中 | SM009 |
| CM026 | Radiant says Kaleidos can be sold through either PPAs or direct unit sales, implying that the payer can vary by segment and contract structure. | 中 | SM001 |
| CM027 | For data-center deployments, the buyer is likely the operator's energy or procurement organization rather than the IT tenant, because public evidence is framed through operator-level procurement and power strategy. | 中 | SM017, SM018, SM023 |
| CM028 | For military-site deployments, the buyer and sponsor are federal installation-energy programs and service branches pursuing resilience rather than merchant power customers. | 高 | SM007, SM008, SM015 |
| CM029 | Remote communities, mining sites, remote defense bases, and disaster-response settings are repeatedly cited as early use cases because microreactors are transportable, self-regulating, and suitable for decentralized generation. | 高 | SM005, SM003, SM014 |
| CM030 | Broad SMR market forecasts are much larger than Radiant's direct niche and therefore overstate any immediate SAM for a 1 MWe transportable high-temperature gas microreactor. | 高 | SM014, SM015, SM024, SM025 |
| CM031 | Precedence Research estimates the global SMR market at $8.16 billion in 2026 and $17.37 billion by 2035, a reported 8.78% CAGR. | 中 | SM024 |
| CM032 | MarketsandMarkets projects the small modular reactor market to reach $7.14 billion by 2030 at a reported 3.0% CAGR. | 中 | SM025 |
| CM033 | The gap between Precedence's and MarketsandMarkets' forecasts shows that even broad SMR TAM estimates vary materially by definition and method. | 中 | SM024, SM025 |
| CM034 | No reviewed public source isolates a dollar SAM specifically for 1 MWe transportable microreactors serving defense, data centers, remote industry, backup power, and remote communities. | 低 | |
| CM035 | Evidence-constrained near-term demand is clearer in segment proof points—Equinix units, military pilot programs, and named Army and Air Force sites—than in vendor TAM decks. | 高 | SM015, SM016, SM017, SM018 |
| CM036 | Data-center power scarcity is a strong demand driver, but adoption still depends on site-level interconnection, financing, ownership, and regulatory design rather than on load growth alone. | 高 | SM022, SM023, SM017 |
| CM037 | Microreactor demand is also pulled by diesel replacement, resilience, smaller-grid compatibility, and decarbonization rather than by hyperscale AI alone. | 高 | SM001, SM014, SM015 |
| CM038 | Licensing, testing, factory-fueling approval, and fuel availability make commercialization timelines slower than power-demand curves suggest. | 高 | SM005, SM010, SM012, SM018, SM021 |
| CM039 | Radiant's public schedule still centers on a 2026 DOME test and initial deployments beginning in 2028, so the market remains pre-commercial even where demand signals exist. | 高 | SM002, SM003, SM021 |
| CM040 | No reviewed public source quantifies Kaleidos's delivered cost or levelized-cost advantage versus diesel, gas, or grid extension. | 低 | |
| CM041 | No reviewed public source converts Equinix, Buckley, Janus, or Eielson interest into a dated revenue or backlog schedule for Radiant. | 低 | |
| CM042 | Factory fabrication, shipping-container transport, semi-autonomous operation, and small footprint are core market attributes because they reduce siting and deployment friction relative to conventional reactors. | 高 | SM005, SM007, SM010, SM014 |
| CM043 | Equinix and Belfer both point to direct power contracting as a relevant adoption path for data-center customers, not just standard utility supply. | 中 | SM018, SM023 |
| CM044 | Remote and critical-infrastructure adoption may still face trust and security scrutiny even when the technical use case is compelling because public HALEU criticism focuses on theft, proliferation, and terrorism risk. | 中 | SM013, SM022 |
| CM045 | EIA's 2026 survey shows the military segment is already progressing from general interest to specific programs through Advanced Nuclear Power for Installations, Janus, nine Army candidate bases, and Eielson's planned 1-5 MWe pilot. | 中 | SM015 |
| CP001 | Radiant markets Kaleidos as a portable nuclear microreactor intended to replace diesel generators for remote villages, hospitals, datacenters, and military installations. | 中 | SP001 |
| CP002 | Radiant says Kaleidos delivers 1 MWe and 1.9 MWth while using air cooling and zero on-site water. | 中 | SP001 |
| CP003 | NRC pre-application materials describe Kaleidos as a high-temperature gas-cooled reactor using TRISO fuel, helium gas coolant, prismatic graphite blocks, and a single shipping-container package generating about 3 MWth and approximately 1 MWe. | 高 | SP001, SP004 |
| CP004 | Radiant says Kaleidos is assembled, fueled, and tested in the factory, can be delivered by truck or aircraft, reaches full power the next day without site excavation, and can be returned after five or more years for refueling. | 中 | SP001, SP009 |
| CP005 | Radiant says customers can buy Kaleidos through PPAs or direct unit sales. | 中 | SP001 |
| CP006 | Radiant’s May 2025 Series C added $165 million and brought disclosed venture funding to $225 million, with proceeds directed in part to a factory expected to produce up to 50 reactors per year. | 中 | SP002 |
| CP007 | Radiant’s own homepage and NRC materials frame Kaleidos as roughly a 1 MWe-class reactor with about 3 MWth output. | 高 | SP001, SP004 |
| CP008 | Independent and partner-oriented coverage around the Equinix deal describes Kaleidos as a 1.2 MWe and 3 MWth reactor, creating a public rating mismatch with Radiant’s 1 MWe framing. | 中 | SP008, SP025 |
| CP009 | Radiant announced that Equinix signed a preorder agreement with deposits for 20 Kaleidos microreactors, while DatacenterDynamics reported the combined Equinix nuclear agreements across three vendors total 774 MWe. | 中 | SP003, SP025 |
| CP010 | Westinghouse says eVinci is a fully factory-assembled, shipping-container transportable microreactor that produces 5 MWe from a 15 MWth core and can run eight or more full-power years before refueling. | 高 | SP007, SP008 |
| CP011 | DOME materials say Westinghouse’s first test article is a one-fifth-scale eVinci reactor that will produce 1 MWe from a 3 MWth core, while the commercial product targets remote communities, mining operations, and data centers. | 中 | SP008, SP009 |
| CP012 | Compared with eVinci, Radiant is closer to the 1 MWe buyer job and publicizes faster site deployment, while eVinci offers more power per unit and a longer disclosed refueling interval. | 中 | SP001, SP007, SP009 |
| CP013 | DOE and INL describe Project Pele as a mobile microreactor effort for electricity production using TRISO fuel, with site work underway and first fuel delivered to Idaho in 2025. | 中 | SP006, SP010 |
| CP014 | EIA says the Air Force selected Oklo’s sodium-cooled Aurora design for Eielson Air Force Base, with a commercially owned and operated 1 MW to 5 MW pilot targeted for 2027. | 中 | SP011 |
| CP015 | Oklo and Meta announced an agreement supporting a 1.2 GW nuclear energy development in southern Ohio, and Meta said its broader 2026 nuclear agreements unlock up to 6.6 GW by 2035. | 中 | SP012, SP013 |
| CP016 | Oklo’s most visible 2026 commercial proof is campus-scale data-center power rather than a shipping-container portable product, so it overlaps with Radiant on clean-firm-power demand more than on package size. | 中 | SP011, SP012, SP013 |
| CP017 | Oklo has an SEC EDGAR entity page, making its disclosure posture more comparable to NuScale and NANO than to private Radiant, Westinghouse, Last Energy, X-energy, or Kairos. | 中 | SP014 |
| CP018 | NANO says it is the first nuclear microreactor company publicly listed in the U.S. and is developing KRONOS, ZEUS, and portable LOKI concepts plus HALEU transport and fuel subsidiaries. | 中 | SP015, SP016 |
| CP019 | NANO’s public materials show a broad development portfolio but do not provide named operating customers or a public test milestone comparable to Radiant’s DOME schedule. | 中 | SP005, SP015 |
| CP020 | Last Energy markets a 20 MWe PWR-20 with a full-service delivery model, factory-produced fully modular building approach, and power-purchase-agreement structure for on-premises baseload at data centers and industrial sites. | 中 | SP022 |
| CP021 | NuScale’s factory-built pressurized-water module produces 77 MWe, scales to a 924 MWe twelve-module plant, and can be financed or supplied under PPA or lease structures through ENTRA1. | 中 | SP017 |
| CP022 | X-energy’s Xe-100 is an 80 MWe and 200 MWt high-temperature gas-cooled reactor that uses TRISO fuel and helium coolant and is designed for four to twelve units per site. | 高 | SP018, SP019 |
| CP023 | DOE and POWER describe X-energy’s first flagship buyer as Dow’s Long Mott project, a four-unit 320 MWe and 800 MWth industrial plant filing for construction permitting in Texas. | 中 | SP019, SP020, SP021 |
| CP024 | Kairos describes KP-FHR as a two-reactor 150 MWe fluoride-salt-cooled high-temperature reactor with TRISO annular pebble fuel, online refueling, and modular scaling for grid, data-center, or industrial users. | 中 | SP023 |
| CP025 | WNA and EIA place microreactors below about 20 MW, emphasize factory fabrication and siting flexibility, and frame data centers, remote communities, and off-grid industry as natural use cases. | 中 | SP011, SP024 |
| CP026 | On that industry definition, NuScale, X-energy, Kairos, and Last Energy are adjacent distributed-nuclear competitors rather than direct portable 1 MWe peers because their standard public offerings start at 20 MWe or far above. | 中 | SP017, SP018, SP022, SP023, SP024 |
| CP027 | Radiant’s closest disclosed direct commercial peer is Westinghouse eVinci, while Oklo, Project Pele, and NANO only partially overlap because they differ on coolant, procurement channel, maturity, or product scope. | 中 | SP007, SP011, SP015 |
| CP028 | DatacenterDynamics reported that Equinix signed deals not only with Radiant but also with ULC-Energy and Stellaria, so Radiant’s best public data-center proof still comes from a buyer deliberately spreading exposure across multiple nuclear vendors. | 中 | SP025 |
| CP029 | DOME selected Radiant and Westinghouse for the first tests, so DOE-backed access to the U.S. microreactor test bed is currently shared rather than exclusive. | 中 | SP005, SP008, SP026 |
| CP030 | DOE, NRC, and NRIC evidence makes Radiant’s licensing and test path more concrete than NANO’s public materials but less mature than NuScale’s approved module and Kairos’s NRC-permitted commercial demonstration sequence. | 中 | SP004, SP015, SP017, SP023 |
| CP031 | Radiant’s fuel and coolant stack is technologically closer to X-energy and Project Pele’s TRISO-centered high-temperature designs than to Oklo’s sodium-cooled Aurora, NuScale’s pressurized-water module, or Kairos’s fluoride-salt reactor. | 中 | SP004, SP010, SP011, SP017, SP018, SP023 |
| CP032 | Radiant’s biggest disclosed competitive advantage is logistics because truck or aircraft transport, no site excavation, next-day startup, and factory refueling align tightly with diesel-replacement jobs that larger SMRs do not fit well. | 中 | SP001, SP017, SP018, SP022, SP023 |
| CP033 | Radiant’s biggest disclosed disadvantages are smaller capital depth than many public or federally backed peers, shared HALEU and TRISO dependence, and the fact that public pricing remains mostly undisclosed across the field. | 中 | SP002, SP011, SP015, SP017, SP020 |
| CP034 | Radiant and Last Energy both publicize PPA-style commercial models, but Last Energy pairs that model with utility-sized 20 MWe plants and a full-service IPP posture rather than portable unit sales. | 中 | SP001, SP022 |
| CP035 | NuScale’s ENTRA1 structure shows that larger SMR vendors can also sell electricity without forcing buyers to own reactor operations, reducing the distinctiveness of Radiant’s PPA narrative outside the portable niche. | 中 | SP017 |
| CP036 | X-energy and Kairos look stronger than Radiant where the buyer needs industrial steam or a very large campus block, but weaker where the job is replacing diesel or energizing a remote microgrid with minimal civil work. | 中 | SP001, SP018, SP021, SP023 |
| CP037 | Public price transparency is too thin for a like-for-like cost table, so contract packaging, site burden, fuel cycle, and customer type are more defensible comparison axes than quoted LCOE or unit price. | 中 | SP001, SP017, SP022, SP024 |
| CP038 | EIA’s 2026 market overview lists BWXT, Kairos, Oklo, Radiant, Westinghouse, X-energy, Last Energy, and Terrestrial among vendors advancing through DOE or defense microreactor pathways, indicating a crowded field with multiple government-backed routes to relevance. | 中 | SP011 |
| CP039 | Because size class is central to the buying job, the most defensible current framing is that Radiant is a 1 MWe-class product until company and partner disclosures reconcile the 1.0 versus 1.2 MWe distinction. | 中 | SP001, SP004, SP025 |
| CP040 | Public sources do not yet show whether the Equinix preorder is site-specific, fully binding, or timed to a disclosed delivery schedule. | 低 | |
| CP041 | Public sources do not yet reconcile long-term commercial HALEU and TRISO availability across Radiant, X-energy, Project Pele-linked supply, and NANO beyond near-term program or subsidiary announcements. | 低 | |
| CP042 | Kairos says it has three approved NRC construction permits and a Google agreement to develop up to 500 MW of clean electricity by 2035. | 中 | SP023 |
| CP043 | EIA also lists Terrestrial Energy among DOE pilot-program-eligible vendors, but the fetched public material here was not detailed enough to profile it as tightly as NuScale, X-energy, Kairos, or Last Energy. | 中 | SP011 |
| CI001 | Radiant publicly says Kaleidos can be sold through either direct unit sales or power purchase agreements. | 中 | SI001 |
| CI002 | Radiant markets Kaleidos as a diesel-replacement power source for remote sites, hospitals, military installations, and data centers. | 高 | SI001, SI014 |
| CI003 | Radiant's public product model includes five-plus-year operation before refueling and a return-to-factory refueling loop, which implies a post-deployment service and logistics obligation. | 高 | SI001, SI006 |
| CI004 | Equinix signed a preorder agreement and submitted deposits for 20 Kaleidos microreactors. | 高 | SI006, SI011 |
| CI005 | Radiant signed the first public agreement designed to deliver a mass-manufactured nuclear microreactor to a U.S. military base under the ANPI pathway. | 高 | SI008, SI012 |
| CI006 | The fetched public sources do not disclose unit price, PPA tariff, deposit size, or contract value for either the Equinix deal or the military-base agreement. | 中 | SI006, SI008, SI011 |
| CI007 | The fetched public sources do not disclose current revenue or ARR for Radiant. | 中 | SI003, SI009, SI014 |
| CI008 | The fetched public sources do not disclose gross margin, revenue mix, or revenue-recognition mechanics for Radiant's early customer agreements. | 中 | SI001, SI003, SI014 |
| CI009 | Radiant's public commercial proof is concentrated in one named data-center counterparty and one named defense deployment pathway rather than a diversified disclosed revenue base. | 中 | SI006, SI008, SI014 |
| CI010 | Radiant's Series C closed at $165 million on May 28, 2025 and brought total venture funding to $225 million. | 高 | SI002, SI026 |
| CI011 | Radiant said Series C proceeds would complete the Kaleidos Development Unit and fund factory siting and early construction for facilities expected to produce up to 50 microreactors per year. | 高 | SI002, SI026 |
| CI012 | Radiant announced on December 17, 2025 that it had raised more than $300 million in a new funding round. | 高 | SI003, SI009 |
| CI013 | Radiant said Draper Associates and Boost VC led the December 2025 round with added commitments from current investors. | 高 | SI003, SI009 |
| CI014 | Combining the disclosed $225 million post-Series C total with the later >$300 million Series D implies a publicly confirmed cumulative funding floor above $525 million. | 高 | SI002, SI003 |
| CI015 | DCVC said Radiant's Series D valued the company at over $1.8 billion, but that valuation marker is investor-sourced rather than filing-backed. | 中 | SI029 |
| CI016 | Public descriptions of the December 2025 financing link the round to commercialization, the DOME schedule, and R-50 factory buildout. | 中 | SI003, SI009, SI029 |
| CI017 | The fetched public financing sources do not disclose post-round cash on hand, monthly burn, or runway months. | 中 | SI003, SI009, SI029 |
| CI018 | Tennessee state and local sources describe Radiant's Oak Ridge project as a $280 million manufacturing and R&D investment that will create 175 jobs. | 高 | SI019, SI027, SI028 |
| CI019 | Public sources say Radiant is using Tennessee's Nuclear Energy Fund and that the statewide fund totals $70 million, but the fetched direct sources do not disclose Radiant's individual grant amount. | 中 | SI020, SI027, SI013 |
| CI020 | Radiant says construction beginning in early 2026 should support first factory-built deliveries in 2028 and eventual scale to 50 reactors per year within a few years. | 高 | SI004, SI010, SI013 |
| CI021 | Data Center Dynamics reported that Radiant also signed a £4 million ($5.1 million) HALEU-related agreement with Urenco, indicating dedicated supply-chain spend beyond the factory headline. | 中 | SI010 |
| CI022 | Radiant's February 2026 Part 70 filing applies for a greater-than-critical-mass license for an Oak Ridge production site that will support fueling of Kaleidos microreactors. | 高 | SI018, SI017 |
| CI023 | The public filing includes an environmental report, material control and accounting plan, physical security plan, safety summary, and a request for expedited review and reduced NEIMA hourly rates. | 中 | SI018 |
| CI024 | The filing says Radiant expects to submit an operating license application for the Kaleidos reactor in the fourth quarter of 2026 and asked NRC to approve the Part 70 application by November 1, 2026. | 中 | SI018 |
| CI025 | NRC's Kaleidos page shows the R-50 materials-license application entered accepted status in May 2026. | 高 | SI017, SI025 |
| CI026 | DOE said DOME experiments are self-funded by applicants, can operate for up to six months, and depend on fuel availability and regulatory plans. | 中 | SI015 |
| CI027 | DOE made a conditional HALEU commitment to Radiant in April 2025 and also said HALEU is not currently available from domestic suppliers. | 中 | SI016 |
| CI028 | Radiant's public jobs board listed 66 open positions across engineering, operations, finance, regulatory, and supply-chain roles. | 中 | SI021 |
| CI029 | Radiant's jobs board shows salary bands up to $314,475 plus equity for some roles, implying a meaningful pre-revenue payroll ramp. | 中 | SI021 |
| CI030 | Factory assembly, factory fueling, delivery logistics, centralized fleet monitoring, and factory return refueling imply a service stack with working-capital and operating-cost obligations beyond one-time reactor manufacturing. | 中 | SI001, SI018, SI021 |
| CI031 | UCS highlighted scientific criticism that HALEU above about 12% uranium-235 can pose practical weapons and terrorism risks. | 中 | SI022 |
| CI032 | DOE and UCS together indicate that HALEU availability and handling standards remain major external constraints on advanced-reactor commercialization. | 高 | SI016, SI022 |
| CI033 | Beyond Nuclear argued that community opposition, waste-accountability concerns, and transport/refueling logistics can create local permitting and reputational risk for Radiant's model. | 低 | SI023 |
| CI034 | The American Bar Association noted that despite policy momentum there is only one commercial North American SMR under construction, underscoring thin deployment precedent for FOAK underwriting. | 中 | SI024 |
| CI035 | Public sources do not disclose backlog by site, cancellation rights, or conversion milestones for the Equinix deposits or the defense agreement. | 中 | SI006, SI008, SI014 |
| CI036 | Public sources do not disclose CAC, payback period, contribution margin, total headcount, or function-by-function staffing. | 中 | SI001, SI021, SI003 |
| CI037 | Because the DOME campaign is self-funded, the factory carries a $280 million headline investment, and fuel remains DOE-linked, Radiant appears likely to remain financing-dependent until testing and early deliveries convert milestones into cash generation. | 中 | SI015, SI016, SI019 |
| CI038 | Radiant's strongest public financial positives are large venture backing, deposit-bearing customer proof, and state-supported factory siting rather than proven recurring revenue or margin disclosure. | 中 | SI002, SI003, SI006, SI027 |
| CI039 | Revenue quality cannot yet be underwritten because the public record does not say whether early customer cash receipts are refundable deposits, milestone payments, PPA prepayments, or recognized sales. | 中 | SI001, SI006, SI008 |
| CI040 | The public financial verdict is therefore 'capitalized but still ununderwritten': enough financing and spend signals exist to support continued execution, but not enough disclosed operating data exists to model durable economics or near-term self-sufficiency. | 低 | SI014, SI018, SI019 |
| CE001 | NRC describes Kaleidos as a transportable microreactor designed to generate about 1 MWe and 3 MWth using TRISO fuel, helium coolant, prismatic graphite blocks, and a single-shipping-container package. | 高 | SE006, SE029 |
| CE002 | Radiant publicly markets Kaleidos as a portable diesel-replacement generator for remote facilities, hospitals, data centers, and military installations rather than as a site-built plant. | 中 | SE001, SE024 |
| CE003 | Public sources round Kaleidos electric output differently, with some describing a 1 MW unit and others a 1.2 MWe unit, so the product rating should be treated as roughly 1.0-1.2 MWe in public diligence materials. | 中 | SE001, SE007, SE018, SE029 |
| CE004 | Company and trade sources also describe Kaleidos as able to provide roughly 1.9 MW of thermal output for facility heating or desalination. | 中 | SE001, SE018, SE019 |
| CE005 | Radiant says helium transfers heat from the core and does not become radioactive, which is part of its argument for leak tolerance and simplified plant operation relative to water-cooled systems. | 中 | SE001, SE009 |
| CE006 | Radiant claims Kaleidos can operate with zero on-site water use because fans and an air jacket remove heat through natural convection. | 中 | SE001, SE019 |
| CE007 | Regulatory and independent sources corroborate that Kaleidos is designed around a single-container transport concept intended for truck or aircraft delivery and rapid deployment. | 中 | SE006, SE014, SE024 |
| CE008 | Radiant alone claims that Kaleidos requires no site excavation and can reach full power the day after delivery, but no independent fueled operating record yet validates that installation promise. | 低 | SE001 |
| CE009 | Radiant positions the product as factory assembled, factory fueled, and factory tested before shipment, making offsite production a core part of the offering rather than an implementation detail. | 中 | SE001, SE005 |
| CE010 | DOE and multiple trade sources say Kaleidos is designed to operate for about five or more years before refueling. | 高 | SE007, SE008, SE015, SE017 |
| CE011 | The 20-year product-life formulation in public materials comes from the company ship-back model, under which a container is returned for refueling and reused across multiple cycles. | 低 | SE001, SE019 |
| CE012 | Public sources do not disclose transport duration, refurbishment turnaround, or spare-unit policy for the factory-return servicing model. | 低 | SE001, SE025, SE030 |
| CE013 | DOE says Radiant completed the FEEED phase in 2024, including schedule, budget, design, and test planning for the DOME experiment. | 中 | SE007, SE017 |
| CE014 | DOE and POWER reported that Radiant is one of the first fueled microreactor experiments selected for INL’s DOME facility, with testing targeted as early as spring or mid-2026. | 高 | SE007, SE015, SE018 |
| CE015 | Radiant says DOE approved its PDSA/DARK submission in February 2026, closing the second of three safety-document stages in the DOE authorization pathway and enabling startup preparation. | 中 | SE004, SE018, SE019 |
| CE016 | Independent reporting says SDS and PDSA approval are only early steps in the DOME safety process, with later conceptual-safety, final-safety, readiness, and startup steps still required before a fueled run. | 高 | SE004, SE012, SE018, SE020 |
| CE017 | NRC pre-application activities with Radiant began in October 2022 under Docket 99902106. | 中 | SE006 |
| CE018 | The NRC public docket lists multiple Kaleidos gap analyses, factory fueling documents, a nonpublic fleet-operator training white paper, and accepted materials-license items, showing the program has moved beyond a single introductory conversation. | 中 | SE006 |
| CE019 | Radiant’s R-50 Part 70 application was accepted in 2026 and covers the Oak Ridge fueling building needed for factory fueling operations rather than the commercial reactor deployment itself. | 高 | SE006, SE010, SE030 |
| CE020 | ANS reports that the NRC still expects a separate Kaleidos reactor application later, so success on the Part 70 factory path does not by itself complete product licensing. | 中 | SE010, SE026 |
| CE021 | World Nuclear News and Nuclear Engineering International both say DOME testing is intended to generate safety and performance evidence that supports future NRC commercial licensing. | 高 | SE008, SE012, SE018 |
| CE022 | Policy and legal coverage describe DOE pilot testing as a separate federal path that can accelerate first operation at a national lab before the standard NRC process for customer-deployed units is complete. | 中 | SE019, SE020, SE026 |
| CE023 | Radiant’s Oak Ridge factory plan calls for construction beginning in early 2026 and scale toward about 50 reactors per year. | 高 | SE005, SE010 |
| CE024 | Because the R-50 facility is intended to support fueling of Kaleidos microreactors, factory-fueling approval is a direct manufacturing dependency rather than a peripheral permitting task. | 高 | SE010, SE030 |
| CE025 | Marketplace reported from Radiant’s headquarters that a physical demonstration unit existed in El Segundo and was expected to ship to Idaho for testing in 2026. | 中 | SE024 |
| CE026 | Buckley selection gives Radiant a named military deployment path, but the next public steps remain site-specific siting and NEPA environmental analysis rather than immediate installation. | 高 | SE002, SE011, SE016 |
| CE027 | ANS and POWER describe the ANPI objective as at least one 3-10 MWe reactor on a military installation by 2030, while Radiant’s public product materials still describe Kaleidos as a 1-1.2 MWe unit. | 中 | SE002, SE011, SE016 |
| CE028 | Radiant announced that Equinix signed a preorder and paid deposits for 20 Kaleidos units, creating a concrete data-center commercial proof point rather than a purely hypothetical use case. | 中 | SE027, SE028, SE029 |
| CE029 | Independent customer-facing reporting describes Kaleidos as on-site data-center power that can be installed in days and operate independently of the grid. | 中 | SE027, SE028, SE029 |
| CE030 | Radiant’s public target remains initial customer deployments beginning in 2028, making the DOME result and later licensing milestones immediate gating items rather than distant roadmap steps. | 中 | SE002, SE015, SE018 |
| CE031 | Current hiring shows Radiant is staffing fluid systems, structures, mechanisms, reactor operations, procedure writing, and control-room interface work for the Kaleidos Development Unit. | 中 | SE022, SE023 |
| CE032 | The public reactor-operations role requires neutronic and transient analysis, procedure development, operator training, and GUI work for the 2026 demonstration, implying operating readiness is still being built now. | 中 | SE023 |
| CE033 | The fluid-systems role ties KDU work to design, assembly, validation, and troubleshooting of subsystems, which suggests balance-of-plant integration remains a live engineering risk area. | 中 | SE022 |
| CE034 | DOE’s HTGR explainer and INL ART technical materials show that TRISO fuel, helium coolant, and graphite structures are established features of the HTGR class, so Kaleidos’ materials stack is class-consistent. | 中 | SE009, SE021 |
| CE035 | INL and ORNL technical materials show that TRISO fuel qualification and graphite behavior are subjects of long-running national-lab programs, indicating that the material basis is more mature than Radiant’s packaged system integration. | 中 | SE021, SE031 |
| CE036 | Generic HTGR and TRISO technical literature does not validate Radiant-specific claims about its integrated package, turbomachinery, transport shell, or service model. | 中 | SE009, SE021, SE031 |
| CE037 | As of 2026-05-21, Kaleidos has not completed a fueled test or public field deployment; the most advanced public milestone is startup preparation for DOME. | 高 | SE004, SE007, SE024 |
| CE038 | Because no fueled field run has occurred, there are no public uptime, maintenance, or availability data for Kaleidos in real customer conditions. | 中 | SE001, SE024 |
| CE039 | Public evidence therefore supports viewing Kaleidos as an advanced prototype with named customers and serious regulatory progress, not as a proven deployed generator fleet. | 中 | SE002, SE024, SE028 |
| CE040 | The 2028 first-delivery target depends on timely DOME success, continued NRC progress, and factory execution, so schedule risk compounds across multiple gates. | 中 | SE010, SE012, SE018 |
| CE041 | The factory-return servicing model reduces on-site nuclear handling but transfers operational risk to transport, turnaround capacity, and centralized refurbishment that public sources do not quantify. | 中 | SE001, SE025, SE030 |
| CE042 | Beyond Nuclear criticizes Radiant’s waste and transport narrative as insufficiently specified, which is an advocacy-framed claim but still points to a real public-evidence gap around end-of-life handling. | 低 | SE025 |
| CE043 | Homepage, Marketplace, and Equinix materials all frame Kaleidos as a diesel replacement for remote, hospital, military, and data-center loads, but no public source yet shows the reactor serving any of those loads outside planned tests or preorders. | 中 | SE001, SE024, SE027 |
| CE044 | Public materials do not reconcile whether Buckley will require multiple Kaleidos units, a modified higher-output configuration, or a phased architecture to meet the ANPI objective. | 低 | SE002, SE011, SE016 |
| CE045 | Public sources do not provide independent operating data for passive air-jacket cooldown, load-following behavior, or integrated generator reliability; most of that evidence remains company-described. | 低 | SE001, SE019, SE022 |
| CE046 | The public five-year refuel interval is better corroborated than the full 20-year lifecycle claim, because independent sources repeat the former while the latter remains primarily company-originated. | 中 | SE001, SE007, SE019 |
| CE047 | Equinix reporting confirms a serious data-center demand signal, but deployment sites have not been publicly disclosed, so siting, interconnection, and multi-unit design complexity remain unknown. | 中 | SE028, SE029 |
| CE048 | No public source reviewed for this chapter discloses Brayton-cycle efficiency, turbomachinery lifetime, maintenance intervals, or demonstrated duty-cycle performance for Kaleidos. | 低 | SE001, SE022 |
| CU001 | Equinix publicly signed a preorder for 20 Kaleidos microreactors and submitted deposits. | 高 | SU004, SU005, SU024 |
| CU002 | Equinix describes the Radiant relationship as part of a diversified power strategy meant to mitigate future power constraints at its data centers. | 中 | SU005, SU021, SU022 |
| CU003 | Equinix publicly says it operates a global network of more than 270 data centers. | 中 | SU004, SU005 |
| CU004 | Radiant says initial customer deployments are expected to begin in 2028. | 中 | SU004, SU002, SU003 |
| CU005 | Public Equinix preorder materials disclose deposits but not deposit size, refundability, or milestone mechanics. | 中 | SU004, SU005, SU007, SU024 |
| CU006 | Public sources do not identify which Equinix sites would receive Kaleidos units. | 中 | SU007, SU021, SU022 |
| CU007 | Trade coverage describes each Kaleidos unit in the Equinix deal as about 1.2 MWe and 3 MWth. | 中 | SU007, SU006, SU008 |
| CU008 | Radiant's July 2025 ANPI announcement was framed as the first-ever agreement designed to deliver a mass-manufactured nuclear microreactor to a U.S. military base. | 中 | SU002, SU013 |
| CU009 | In April 2026 the DAF and DIU selected Radiant for Buckley Space Force Base under ANPI. | 高 | SU003, SU010, SU014, SU016 |
| CU010 | ANPI aims to have at least one advanced nuclear reactor operating on a DAF installation by 2030 or sooner. | 高 | SU003, SU010, SU016 |
| CU011 | The Buckley pathway still requires siting and environmental analyses under NEPA before final approval. | 高 | SU003, SU010, SU014, SU023 |
| CU012 | Air Force and Buckley officials describe the project as resilient power for critical national-security loads tied to air and space operations. | 中 | SU003, SU010, SU011, SU023 |
| CU013 | DIU's 2025 ANPI program made Radiant one of eight companies eligible for Other Transaction awards before the 2026 site pairing. | 中 | SU012, SU018 |
| CU014 | Reviewed public commercial demand proof is limited to one named commercial counterparty: Equinix. | 中 | SU001, SU004, SU009 |
| CU015 | Reviewed public defense demand proof is limited to one named Air Force pathway centered on Buckley Space Force Base. | 中 | SU003, SU010, SU012 |
| CU016 | Radiant markets hospitals, datacenters, military installations, remote villages, and similar off-grid sites as target users. | 中 | SU001, SU009 |
| CU017 | Equinix's interest reflects the need for firm round-the-clock power as AI-driven data-center demand strains grids and utilities. | 中 | SU005, SU020, SU025 |
| CU018 | DAF and DIU frame microreactors as resilience assets against constrained grids, natural disasters, and physical or cyber attacks. | 中 | SU010, SU012, SU023 |
| CU019 | ANPI public materials emphasize 3–10 MWe installation goals while Radiant publicly describes Kaleidos as roughly 1–1.2 MWe, leaving open whether Buckley needs multiple units or a tailored architecture. | 中 | SU019, SU018, SU003 |
| CU020 | Equinix is Radiant's strongest public commercial proof point because the counterparties disclosed a unit count and confirmed deposits. | 中 | SU004, SU005, SU024 |
| CU021 | Buckley is a selected pre-deployment pathway rather than an operating customer site already taking power. | 高 | SU003, SU010, SU023 |
| CU022 | No reviewed public source documents a Radiant customer site already generating electricity. | 中 | SU004, SU009, SU010 |
| CU023 | No reviewed public source discloses a broader paying-customer count, backlog count, or diversified commercial roster beyond the named Equinix and Buckley demand anchors. | 中 | SU001, SU004, SU009, SU010 |
| CU024 | Equinix told Data Center Knowledge that many next-generation nuclear technologies are still several years away from deployment and that site selection is still early. | 中 | SU021 |
| CU025 | Equinix told Data Center Knowledge that the 2030s are the earliest period when nuclear technology can be deployed at scale. | 中 | SU021, SU020 |
| CU026 | Equinix's public energy strategy pairs Radiant with Oklo, ULC-Energy / Rolls-Royce, Stellaria, Bloom, grid upgrades, gas, and fuel cells rather than making Radiant its sole future power provider. | 中 | SU005, SU007, SU022 |
| CU027 | No public NRR, GRR, churn, renewal-rate, contract-length, or customer-satisfaction metrics for Radiant were found in the reviewed evidence. | 中 | SU004, SU009, SU010 |
| CU028 | No public source in the reviewed set discloses contract value, per-unit pricing, or PPA tariff for either Equinix or the Buckley pathway. | 中 | SU004, SU005, SU002, SU010 |
| CU029 | Equinix's own forward-looking release flags risks around sourcing power and land, supply-chain constraints, financing execution, and decline in business from key customers. | 中 | SU005 |
| CU030 | Partnership for Global Security says the Air Force's separate microreactor pilot at Eielson targets a 30-year third-party PPA, while ANPI uses OTA and milestone-based contracting through commercial and NRC pathways. | 中 | SU018, SU012 |
| CU031 | The public procurement structure suggests Buckley conversion into revenue depends on federal contracting, licensing, ownership, operations, and decommissioning terms not yet disclosed for Radiant. | 中 | SU018, SU010, SU002 |
| CU032 | Axios reported that Buckley neighbors said the April 2026 announcement came without warning and that no community meetings were yet scheduled. | 中 | SU023 |
| CU033 | Buckley's commander said strong community trust and continued engagement are important to the project's success. | 中 | SU003, SU014 |
| CU034 | Goldman says data-center power demand is on track to grow more than 160% by 2030 versus 2023, increasing the appeal of firm baseload power. | 中 | SU020, SU025 |
| CU035 | Goldman also says permitting, specialized labor, and uranium supply make new nuclear harder to build in the short term, limiting near-term customer conversion. | 中 | SU020, SU021 |
| CU036 | Equinix told Data Center Knowledge that policy changes, faster permitting, and skilled workforce availability are needed to deliver nuclear-backed projects on time. | 中 | SU021 |
| CU037 | With one named commercial buyer and one named defense pathway, Radiant's public customer evidence appears highly concentrated. | 中 | SU004, SU010, SU009 |
| CU038 | If Equinix or Buckley slips, the public demand narrative weakens materially because no broader roster or repeat orders are publicly disclosed. | 中 | SU004, SU010, SU009 |
| CU039 | Equinix's portfolio approach reduces the odds that Radiant alone captures all future nuclear capacity Equinix may procure. | 中 | SU005, SU007, SU022 |
| CU040 | Data Center Frontier describes Equinix's near-term capacity additions as fuel cells, grid upgrades, and other alternative-energy projects, with advanced nuclear positioned on a longer 2029–2035 horizon. | 中 | SU022, SU005 |
| CU041 | Breaking Defense says critics argue ANPI's benefits are illusory and come with too much cost and risk. | 中 | SU015 |
| CU042 | Axios shows Buckley community sentiment is mixed supportive-but-cautious rather than fully settled. | 中 | SU023, SU010 |
| CU043 | The absence of public deployment-site, pricing, contract-tenor, and satisfaction data keeps contract-quality uncertainty high for both named demand anchors. | 中 | SU004, SU005, SU002, SU010 |
| CU044 | Equinix says it is pursuing 100% clean and renewable energy across its global portfolio by 2030 and had reached 96% renewable coverage in 2024. | 中 | SU005, SU024 |
| CU045 | Radiant's homepage says customers can buy Kaleidos through either direct unit sales or power purchase agreements. | 中 | SU001 |
| CU046 | The gap between Radiant's broad marketing segments and its narrow disclosed counterparties means target-segment breadth should not be mistaken for a diversified customer base. | 中 | SU001, SU009, SU004, SU010 |
| CU047 | Radiant's service-heavy model includes factory assembly, transport, refueling, and centralized 24/7 fleet monitoring, but no public customer SLA or service-performance evidence is disclosed. | 中 | SU001, SU009 |
| CU048 | Radiant claims installation can happen in days and full power can follow the next day, but no public customer case study verifies those field outcomes. | 中 | SU001, SU004, SU010 |
| CU049 | DOE said in late 2024 that U.S. data-center electricity use is expected to double or triple by 2028, reinforcing why buyers like Equinix are searching for new firm power sources. | 中 | SU025, SU005 |
| CR001 | Buckley Space Force Base remains in the siting and environmental-analysis phase, so Radiant has not yet cleared the site-level NEPA gate for deployment. | 高 | SR005, SR007 |
| CR002 | The Air Force’s ANPI objective is to have at least one advanced reactor operating on a DAF installation by 2030 or sooner, embedding schedule pressure into Radiant’s defense timeline. | 中 | SR005 |
| CR003 | Radiant’s defense path is competitive rather than exclusive because ANPI initially named eight eligible companies and later assigned only three companies to three bases. | 高 | SR005, SR006 |
| CR004 | The NRC’s active microreactor work program still treats security, safeguards, emergency preparedness, transportation, siting, and decommissioning funding as live regulatory topics. | 中 | SR001 |
| CR005 | NRC Part 57 was proposed in 2026 and remains a not-yet-final rulemaking rather than a settled licensing shortcut available on fixed terms today. | 中 | SR002, SR020 |
| CR006 | Part 57 is designed to allow joint applications, fleet approvals, and manufacturing-license reuse for qualifying low-consequence designs, potentially compressing review timelines. | 中 | SR002, SR020 |
| CR007 | Eligibility for Part 57 depends on consequence and fuel-inventory criteria, so a transportable reactor does not automatically qualify just because it is small. | 中 | SR002 |
| CR008 | Radiant’s accepted Part 70 application covers the R-50 fueling building rather than the Kaleidos reactor itself. | 中 | SR004 |
| CR009 | The NRC expects Radiant to submit a separate Kaleidos application, leaving factory authorization and reactor authorization as distinct gates. | 中 | SR004 |
| CR010 | DOE’s 2026 advanced-reactor categorical exclusion can shorten DOE-side environmental review but does not replace NRC, state, tribal, or local approval processes for commercial deployment. | 中 | SR018 |
| CR011 | Legal commentary still describes North America as having only one commercial SMR under construction, underscoring how little deployment precedent advanced-reactor developers can cite. | 中 | SR022 |
| CR012 | Colorado’s nuclear-policy environment is still being updated in 2026, so Buckley-related siting politics sit inside an evolving rather than mature state framework. | 中 | SR019 |
| CR013 | DOE says HALEU is not currently available from domestic suppliers. | 高 | SR010, SR012 |
| CR014 | DOE says gaps in HALEU supply could delay advanced-reactor deployment. | 中 | SR010 |
| CR015 | Radiant was one of only five first-round HALEU allocatees selected from fifteen requesting companies. | 高 | SR012, SR013 |
| CR016 | The HALEU feedstock delivered for Radiant’s 2026 demonstration was described as sufficient for a full first core load for startup, not as proof of fleet-scale fuel availability. | 中 | SR014 |
| CR017 | DOE’s HALEU program is explicitly meant to catalyze private investment and eventually remove the federal government from the role of initial supplier. | 高 | SR010, SR011 |
| CR018 | Centrus’ current HALEU production extension runs through June 30, 2026 and remains DOE-controlled and appropriations-dependent. | 中 | SR015 |
| CR019 | Centrus’ disclosed pilot-scale HALEU quantities remain measured in kilograms to hundreds of kilograms rather than in a clearly commercial fleet-supply cadence. | 中 | SR015 |
| CR020 | BWXT’s successful delivery of Project Pele TRISO fuel shows real TRISO production capability exists, but the associated pathfinder reactor is still pre-formal system test and not yet a proof of broad commercial availability. | 中 | SR016 |
| CR021 | BWXT and Kairos are still exploring a future commercial TRISO fabrication facility, which implies broader commercial TRISO scale-up remains in development. | 中 | SR017 |
| CR022 | DOME experiments are self-funded by applicants and keep their slot only if they satisfy technology-readiness, fuel-availability, and regulatory-plan milestones. | 中 | SR025 |
| CR023 | Each DOME experiment is expected to operate for only up to six months, so DOME is not equivalent to a multi-year customer-service proof point. | 中 | SR025 |
| CR024 | Radiant’s public plan still combines a 2026 test article, a 2028 first-factory target, and a longer-term 50-reactor-per-year ambition, which makes the manufacturing ramp a first-of-a-kind leap. | 中 | SR004, SR032 |
| CR025 | No reviewed public source in this chapter discloses multi-core HALEU purchase commitments, spare-core inventory, or alternate TRISO conversion providers beyond the first-startup path. | 中 | SR010, SR014, SR015, SR017 |
| CR026 | Factory fueling does not eliminate downstream compliance because the NRC is still separately developing approaches for transport, operational oversight, and decommissioning of microreactors. | 高 | SR001, SR003 |
| CR027 | Equinix’s power strategy is diversified across Bloom fuel cells and multiple advanced-nuclear developers, so Radiant is one element of a wider portfolio rather than the sole plan. | 中 | SR027, SR028 |
| CR028 | Equinix already operates about 75 MW of Bloom fuel cells and is constructing another 30 MW, making substitute onsite power live today rather than hypothetical. | 中 | SR027 |
| CR029 | Equinix said many next-generation nuclear technologies remain several years from deployment and that site selection is still early. | 中 | SR028 |
| CR030 | Bloom’s 2026 power survey says utilities’ delivery timelines run roughly 1.5 to 2 years longer than hyperscaler and colo expectations in key hubs. | 中 | SR029, SR030 |
| CR031 | Bloom’s 2026 power survey says roughly one-third of data centers expect to use fully onsite power by 2030. | 中 | SR029, SR030 |
| CR032 | Bloom’s 2026 power survey says onsite power is becoming a top-ranked way for developers to minimize data-center timelines and costs. | 中 | SR029, SR030 |
| CR033 | DIU says ANPI is intended not only to power bases but also to stimulate commercial microreactor technology development and associated U.S. supply chains. | 中 | SR006 |
| CR034 | Marketplace quoted Doug Bernauer saying government support is extremely important for Radiant. | 中 | SR026 |
| CR035 | No public source in this chapter discloses Buckley contract value, Equinix pricing, or the recurring cash-flow terms of either flagship demand pathway. | 中 | SR005, SR006, SR027, SR028 |
| CR036 | Radiant’s public demand case remains concentrated in one marquee commercial relationship and one marquee defense pathway. | 中 | SR005, SR027, SR028 |
| CR037 | Fuel-cell-backed onsite power is already deployed at customer scale while Radiant remains pre-test, so substitute power can win on immediacy even if nuclear wins on eventual endurance. | 中 | SR027, SR029 |
| CR038 | Defense demand is not exclusive to Radiant because Westinghouse and Antares were assigned alternative Air Force bases under the same initiative. | 高 | SR005, SR006 |
| CR039 | Axios found Buckley-area neighbors said the April 2026 announcement came without warning. | 中 | SR009 |
| CR040 | Axios reported that no community meetings with Buckley neighbors were yet scheduled when it published. | 中 | SR009 |
| CR041 | KUNC reported that local unease around Buckley focused on nuclear waste and long-term environmental impact even while expert sources argued the risk is manageable. | 中 | SR008 |
| CR042 | Beyond Nuclear says Wyoming opposition to Radiant’s earlier factory plan centered on unanswered safety, waste, and accountability questions. | 低 | SR024 |
| CR043 | Utility Dive quoted former NRC chair Christopher Hanson flagging end-of-life shipment of irradiated microreactors as a distinct risk category from shipment of fresh-fueled units. | 中 | SR003 |
| CR044 | The NRC’s microreactor agenda explicitly includes transportation of fueled microreactors and decommissioning funding assurance, showing those topics are unresolved at framework level. | 高 | SR001, SR003 |
| CR045 | UCS says HALEU above about 12% uranium-235 could be directly usable in practical nuclear weapons and merits stronger safeguards. | 中 | SR023 |
| CR046 | If HALEU security rules tighten or export controls broaden, Radiant’s portable-reactor logistics and addressable market would become more constrained. | 中 | SR018, SR023 |
| CR047 | INL’s 2026 microreactor market study says public perceptions around uncertainty, waste, and fuel management remain deployment challenges. | 中 | SR021 |
| CR048 | INL’s 2026 market study recommends additional research on public acceptance and cross-jurisdictional regulatory issues, indicating the social-license problem is not solved. | 中 | SR021 |
| CR049 | Radiant disclosed a Series D of more than $300 million in December 2025 and outside coverage said the raise supported commercialization and the R-50 buildout. | 中 | SR031, SR032 |
| CR050 | Tennessee described R-50 as a $280 million project creating 175 jobs, which underscores the scale and capital intensity of the factory program ahead of revenue disclosure. | 中 | SR033 |
| CR051 | Public funding and factory announcements still do not disclose current cash, monthly burn, runway, gross margin, or per-reactor economics. | 中 | SR026, SR031, SR032, SR033 |
| CR052 | Government-backed milestones can support fundraising narratives, but they do not substitute for licensed operation or customer cash generation. | 中 | SR025, SR031, SR032 |
| CR053 | Commercialization requires Radiant to scale licensing work, special-nuclear-material handling, manufacturing QA, logistics, and field service at the same time. | 中 | SR004, SR025, SR033 |
| CR054 | Radiant’s own public milestone set ties capital deployment to DOME testing, factory construction, and future production scale, making program-management slippage financially important. | 中 | SR031, SR032, SR033 |
| CR055 | Public evidence still does not support a precise runway calculation for Radiant despite the size of its fundraising. | 中 | SR026, SR031, SR032 |
| CR056 | Because ANPI, DOME access, HALEU allocations, and DOE permitting reforms are all government-driven, a policy reversal could hit schedule and valuation at the same time. | 中 | SR006, SR010, SR018, SR020 |
| CR057 | The clearest thesis-break triggers are DOME slippage, fuel interruption beyond the first core, licensing delay beyond the current 2028 plan, flagship-customer pullback, and inability to finance the factory ramp. | 中 | SR004, SR014, SR027, SR031, SR032, SR033 |
| CV001 | Radiant said its Series C closed at $165 million and brought total venture funding to $225 million. | 中 | SV002 |
| CV002 | Radiant announced on December 17, 2025 that it had raised more than $300 million in a new funding round. | 高 | SV003, SV011, SV023 |
| CV003 | Combining Radiant’s disclosed $225 million post-Series C total with the later >$300 million Series D implies a publicly confirmed funding floor above $525 million. | 高 | SV002, SV003 |
| CV004 | The >$1.8 billion valuation reference is investor- and media-reported rather than company filing-backed. | 高 | SV012, SV023 |
| CV005 | TechCrunch reported that Radiant’s December 2025 round valued the company at more than $1.8 billion. | 中 | SV023 |
| CV006 | The fetched public record does not disclose Radiant revenue, ARR, gross margin, monthly burn, or runway. | 中 | SV001, SV003, SV015 |
| CV007 | Radiant’s public customer proof is still deposits, agreements, and future availability by 2028 rather than disclosed recognized sales. | 中 | SV008, SV015, SV016 |
| CV008 | Equinix signed a preorder agreement and submitted deposits for 20 Kaleidos microreactors. | 高 | SV008, SV009 |
| CV009 | Marketplace reported that Radiant expects its first commercial microreactors to be available by 2028 at the latest. | 中 | SV015 |
| CV010 | DOE said DOME testing campaigns are self-funded by applicants and depend on fuel availability and a regulatory approval plan. | 中 | SV016 |
| CV011 | DOE said HALEU is not currently available from domestic suppliers even while distributing first amounts to advanced-reactor developers. | 中 | SV017 |
| CV012 | NRC’s Kaleidos page shows Radiant’s R-50 greater-than-critical-mass application was accepted for review in May 2026. | 中 | SV018 |
| CV013 | Radiant’s Part 70 filing seeks a greater-than-critical-mass license for the Oak Ridge production site and asked for approval by November 1, 2026. | 中 | SV019 |
| CV014 | Tennessee and Roane County sources describe Radiant’s Oak Ridge project as a $280 million investment creating 175 jobs. | 高 | SV013, SV014 |
| CV015 | Radiant said the Tennessee factory path is intended to support first factory-built deliveries in 2028 and eventual output of up to 50 reactors per year. | 高 | SV007, SV013 |
| CV016 | Radiant’s valuation support today is milestone-based rather than operating-metric-based because the strongest public evidence is capital, customer names, and regulatory progress rather than revenue disclosure. | 中 | SV003, SV015, SV016, SV018 |
| CV017 | TechCrunch argued the recent string of large nuclear startup rounds makes it reasonable to ask whether the sector is in a bubble. | 中 | SV023 |
| CV018 | TechCrunch warned that startups may reach first criticality yet still stumble when they try to replicate designs through mass manufacturing. | 中 | SV023 |
| CV019 | UCS argued that HALEU above about 12% uranium-235 could create practical proliferation and terrorism risks. | 中 | SV020 |
| CV020 | Beyond Nuclear criticized unresolved transport, waste, and community-accountability questions in Radiant’s return-to-factory model. | 低 | SV021 |
| CV021 | The American Bar Association wrote that only one commercial North American SMR is currently under construction, highlighting thin deployment precedent. | 中 | SV022 |
| CV022 | On May 21, 2026, Stock Analysis listed Oklo at roughly $11.33 billion market cap and $9.00 billion enterprise value with revenue marked n/a. | 中 | SV024 |
| CV023 | The same Oklo market-data page showed about $2.21 billion of cash and negative free cash flow, indicating investors are paying for option value and balance-sheet runway rather than current reactor revenue. | 中 | SV024, SV025 |
| CV024 | On May 21, 2026, Stock Analysis listed NuScale at roughly $4.14 billion market cap, $3.19 billion enterprise value, and $18.67 million of trailing revenue. | 中 | SV026 |
| CV025 | NuScale’s March 2026 10-Q shows quarterly revenue categories totaling only about $565 thousand and notes no revenue from Fluor in the quarter. | 中 | SV027 |
| CV026 | On May 21, 2026, Stock Analysis listed NANO Nuclear at roughly $1.31 billion market cap and $742.82 million enterprise value with revenue marked n/a. | 中 | SV028 |
| CV027 | NANO Nuclear’s 10-K says it is pre-revenue and had not generated any revenues as of the report date. | 中 | SV029 |
| CV028 | X-energy closed an oversubscribed Series D of about $700 million in November 2025 and said its orderbook exceeded 11 GW representing roughly 144 reactors. | 中 | SV030 |
| CV029 | Helion raised $425 million at a $5.425 billion post-money valuation in January 2025, showing that private advanced-nuclear valuations can stretch far above current operating revenue when milestone optimism dominates. | 中 | SV031 |
| CV030 | Last Energy announced an oversubscribed Series C of more than $100 million and said the round would fully capitalize its DOE pilot project. | 中 | SV032 |
| CV031 | POWER reported that nuclear fission companies had already raised $1.3 billion in equity funding by Q3 2025, the sector’s highest annual total on record. | 中 | SV033 |
| CV032 | Relative to public microreactor peers, Radiant’s >$1.8 billion mark sits well below Oklo’s public valuation but above NANO Nuclear’s while Radiant still lacks disclosed revenue. | 中 | SV012, SV024, SV028 |
| CV033 | Relative to NuScale, Radiant has less disclosed revenue support and less regulatory maturity, so a similar valuation premium would require stronger future milestone delivery than public evidence yet shows. | 中 | SV018, SV024, SV026, SV027 |
| CV034 | No reviewed public source discloses Radiant’s cap-table share count, liquidation preferences, or dilution overhang from the 2025 rounds. | 中 | SV003, SV012, SV023 |
| CV035 | No reviewed public source discloses the size, refundability, or conversion mechanics of the Equinix deposits. | 中 | SV008, SV009, SV010 |
| CV036 | No reviewed public source discloses unit price, PPA tariff, gross margin, or burn well enough to support DCF-style precision. | 中 | SV001, SV003, SV015 |
| CV037 | A defensible base case therefore has to weight milestone completion more than revenue multiples because Radiant’s commercial deliveries remain future-tense and financing is still milestone-funded. | 中 | SV016, SV018, SV019, SV033 |
| CV038 | A reasonable bear case is a chain of DOME, fuel, or licensing slippage that forces another raise before Radiant converts deposits or defense agreements into recognized revenue. | 中 | SV016, SV017, SV018, SV020, SV023 |
| CV039 | A credible bull case requires successful 2026 DOME testing, continued NRC progress, durable HALEU access, and conversion of flagship customer proof into paid deliveries. | 中 | SV004, SV006, SV008, SV016, SV017, SV018 |
| CV040 | Because the current >$1.8 billion figure is investor- and media-reported rather than filing-backed, it should be treated as a directional price signal rather than a precise fair value. | 高 | SV012, SV023 |
| CV041 | At the current reported mark, the public evidence set supports a research-more or track posture rather than a buy call. | 中 | SV003, SV015, SV023 |
| CV042 | The most decision-useful stance is stretched to fair-at-best: Radiant has stronger financing and customer proof than very early peers, but not enough operating disclosure to justify treating the current mark as cheap. | 中 | SV012, SV024, SV026, SV028, SV032 |
| CV043 | An estimated scenario framework of roughly $0.9–1.3 billion bear, $1.4–2.0 billion base, and $2.1–3.0 billion bull is more defensible than a single-point DCF because it anchors on milestone-adjusted peer comparisons and explicit caveats. | 低 | SV012, SV024, SV026, SV028, SV030, SV031, SV032, SV033 |
| CV044 | Downside triggers include DOME slipping past 2026, HALEU supply disruption, Part 70 delay beyond the requested timeline, evidence of Equinix retrading or cancellation, or any new financing with punitive preferences. | 中 | SV016, SV017, SV018, SV019, SV023 |
| CV045 | Evidence still missing to move bullish includes exact post-money share count, liquidation preferences, current cash, monthly burn, deposit terms, unit economics, and independent support for the current mark. | 中 | SV003, SV008, SV012, SV015, SV023 |
| CV046 | Equinix’s broader nuclear portfolio shows that its Radiant preorder is validation but not exclusivity; the customer is actively diversifying across multiple advanced-nuclear pathways. | 高 | SV009, SV010 |
| CV047 | Public market comps show investors will award multi-billion values to nuclear names before commercial reactor revenue, but those same comps also show dilution, volatility, and short interest that warn against treating narrative premiums as stable. | 中 | SV024, SV026, SV028 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | Radiant Nuclear | Radiant | Requiring zero on site water use by leveraging fans for air cooling, Kaleidos can operate anywhere. |
| SO002 | Radiant Nuclear | Radiant closes $165 Million Series C with additional funding from Giant Ventures, StepStone, ARK Venture Fund, and others | Radiant announced today the successful close of its $165 million Series C funding round, bringing the company's total venture funding to $225 million. |
| SO003 | Radiant Nuclear | Radiant raises over $300 million in new funding to mass-produce portable nuclear reactors | Radiant today announced that it has raised more than $300 million in a new round of funding. |
| SO004 | Radiant Nuclear | Dr. Rita Baranwal Joins Radiant as First Chief Nuclear Officer Ahead of Historic Reactor Test | Radiant announced the appointment of Dr. Rita Baranwal as its first-ever Chief Nuclear Officer. |
| SO005 | Radiant Nuclear | Radiant Selected by Department of Energy as First New Nuclear Reactor Design to Be Tested in DOME | Radiant announced today that it has been conditionally selected by the U.S. Department of Energy to conduct the first test of its Kaleidos microreactor in the DOME test bed at Idaho National Laboratory. |
| SO006 | Radiant Nuclear | Radiant receives DOE approval of Preliminary Documented Safety Analysis, latest milestone towards start up of first reactor | The U.S. Department of Energy has granted approval of its DOE Authorization Request for Kaleidos (DARK). |
| SO007 | Radiant Nuclear | Air Force selects Radiant to deliver microreactors to Buckley Space Force Base | The Department of the Air Force, in conjunction with the Defense Innovation Unit, today announced that it has selected Radiant to develop and operate a nuclear microreactor proposed for Buckley Space Force Base. |
| SO008 | Radiant Nuclear | Radiant on Track to Build the First Commercial Nuclear Microreactors Factory in America | The U.S. Nuclear Regulatory Commission has formally accepted for review Radiant's 10 CFR Part 70 license application for its R-50 Production Facility in Tennessee. |
| SO009 | Radiant Nuclear | Radiant adds Lockheed Martin as strategic investor to oversubscribed funding round | Radiant today announced a strategic investment from Lockheed Martin through their investment arm, Lockheed Martin Ventures. |
| SO010 | Newswire.com / Radiant Industries | Radiant Announces Equinix Preorder & Deposits for 20 Kaleidos Microreactors | Equinix has signed a preorder agreement and submitted deposits for the purchase of 20 Kaleidos microreactors. |
| SO011 | World Nuclear News | Equinix signs further agreements with SMR developers | Equinix has signed a preorder agreement for the purchase of 20 of Radiant's Kaleidos microreactors. |
| SO012 | Data Center Dynamics | Equinix signs three major deals with advanced nuclear firms for more than 750MW of power | Equinix will purchase 20 Kaleidos microreactors. |
| SO013 | Data Center Dynamics | Equinix-backed microreactor firm Radiant raises $300m in latest funding round | Radiant Industries has raised more than $300 million as part of its latest funding round. |
| SO014 | DCVC | Radiant raises $300 million+ Series D toward portable nuclear energy at scale | Radiant's $300M+ Series D values the company at over $1.8 billion. |
| SO015 | Giant Ventures | Radiant | When Doug Bernauer started Radiant in 2020, he predicted this test would happen at the end of Q2 2026. |
| SO016 | Inflection VC | Radiant Nuclear | Advanced Microreactor Power for Defense and AI Infrastructure | Radiant is effectively sold out through 2030 with prepaid deposits across defense and commercial buyers. |
| SO017 | University of Michigan News | Powering the next generation of nuclear energy | Rita Baranwal is the first chief nuclear officer at Radiant. |
| SO018 | Nuclear Engineering International | Radiant secures microreactor funding | The reactor has a five-year fuel cycle and a 20-year service life. |
| SO019 | NucNet | Radiant Raises $165 Million For Development Of 'World's First' Mass Produced Nuclear Reactor | Radiant Industries announced it has raised $165m in funding to complete what it says could be the world's first mass-produced portable microreactor design. |
| SO020 | NucNet | Radiant Receives DOE Approval For Kaleidos Microreactor Safety Document | Radiant Industries announced it has received approval from the Department of Energy for a safety document covering its Kaleidos test reactor. |
| SO021 | Utility Dive | DOE agrees to give HALEU to 5 advanced nuclear companies | HALEU remains scarce in the United States, however. Civilian production is held back by market uncertainties and infrastructure gaps. |
| SO022 | U.S. Department of Energy | Demonstration of Microreactor Experiments (DOME) | The DOME facility at Idaho National Laboratory is the world's first microreactor test bed. |
| SO023 | National Reactor Innovation Center | DOME | Experimental reactors can safely go critical for the first time. |
| SO024 | Union of Concerned Scientists | Analysis Published in Science Finds High Assay Low-Enriched Uranium Fuel to be Produced for Small Nuclear Power Reactors Poses a Greater Proliferation Threat than Previously Acknowledged | HALEU fuel used for small nuclear power reactors can be used to produce nuclear weapons, posing terrorism and nuclear proliferation threats. |
| SO025 | American Nuclear Society | NRIC's DOME open for business | NRIC is currently preparing for DOME's inaugural fueled experiment, which is slated to take place later this year. That experiment will focus on Radiant Industries' Kaleidos Demonstration Unit. |
| SO026 | ExchangeMonitor | Radiant appoints Rita Baranwal as its chief nuclear officer | Radiant has appointed former Department of Energy Assistant Secretary for Nuclear Energy Rita Baranwal as its first-ever chief nuclear officer. |
| SM001 | Radiant Nuclear | Radiant | Replacing diesel generators in remote villages and providing resilient backup power for hospitals, datacenters, and military installations |
| SM002 | Radiant Nuclear | Radiant closes $165 Million Series C with additional funding from Giant Ventures, StepStone, ARK Venture Fund, and others | Radiant's Kaleidos 1MW microreactor is designed to replace diesel generators and can be rapidly deployed to provide resilient power for remote villages, emergency response, and military installations. |
| SM003 | U.S. Department of Energy | 3 Microreactor Experiments to Watch Starting in 2026 | Kaleidos is designed to operate for five years before refueling and could provide backup power to hospitals, military installations, data centers, and more. |
| SM004 | U.S. Department of Energy | Demonstration of Microreactor Experiments (DOME) | The DOME facility at Idaho National Laboratory is the world's first microreactor test bed. |
| SM005 | Gateway for Accelerated Innovation in Nuclear | DOE Microreactor Program | Identifying and addressing technology solutions to improve the economic viability and licensing readiness of microreactors |
| SM006 | National Reactor Innovation Center | DOME | Experimental reactors can safely go critical for the first time. |
| SM007 | U.S. Department of Energy | Department of Defense Breaks Ground on Project Pele Microreactor | The reactor is expected to operate for a minimum of three years at the lab and will help demonstrate the use of clean, reliable, and transportable nuclear power to help meet the increasing energy demands at military bases across the country. |
| SM008 | Office of the Under Secretary of Defense for Research and Engineering | Project PELE Mobile Nuclear Reactor | The U.S. military could become the beneficiaries of reliable, abundant, and continuous energy through the deployment of nuclear energy power systems. |
| SM009 | Idaho National Laboratory | INL advances Department of War's Project Pele demonstration microreactor with first TRISO fuel delivery | The Army's Janus Program will follow on to deliver affordable, reliable, commercial nuclear power to ensure that our critical infrastructure has power even if the electric grid is disrupted. |
| SM010 | U.S. Nuclear Regulatory Commission | Kaleidos | Nuclear Regulatory Commission | Radiant is an El Segundo, CA company developing its Kaleidos design, a transportable micro-reactor designed to generate 3MWth and approximately 1MWe. |
| SM011 | U.S. Department of Energy | U.S. Department of Energy to Distribute First Amounts of HALEU to U.S. Advanced Reactor Developers | The U.S. Department of Energy (DOE) today made conditional commitments to provide high-assay low-enriched uranium (HALEU) to five U.S. nuclear developers to meet their near-term fuel needs. |
| SM012 | Utility Dive | DOE agrees to give HALEU to 5 advanced nuclear companies | HALEU remains scarce in the United States, however. Civilian production is held back by market uncertainties and infrastructure gaps. |
| SM013 | Union of Concerned Scientists | Analysis Published in Science Finds High Assay Low-Enriched Uranium Fuel to be Produced for Small Nuclear Power Reactors Poses a Greater Proliferation Threat than Previously Acknowledged | HALEU now being produced with federal subsidies to fuel the next generation of small nuclear power reactors can be used directly to make nuclear weapons. |
| SM014 | World Nuclear Association | Small Modular Reactors - World Nuclear Association | The smaller capacity of SMRs allows for deployment in settings where large plants may not be practical – such as remote communities, industrial clusters, or regions with small electricity grids. |
| SM015 | U.S. Energy Information Administration | Small modular reactors and microreactors under development in the United States | SMRs are under consideration for powering AI, data centers, or other industrial activities where developers may not want or need to connect to the grid. |
| SM016 | Newswire.com / Radiant Industries | Radiant Announces Equinix Preorder & Deposits for 20 Kaleidos Microreactors | Equinix ... has signed a preorder agreement and submitted deposits for the purchase of 20 Kaleidos microreactors. |
| SM017 | Data Center Dynamics | Equinix signs three major deals with advanced nuclear firms for more than 750MW of power | Equinix has expanded its commitment to nuclear energy, announcing three major agreements with advanced nuclear developers that together could provide up to 774MWe of power. |
| SM018 | World Nuclear News | Equinix signs further agreements with SMR developers | Equinix has signed a preorder agreement for the purchase of 20 of Radiant's Kaleidos microreactors and intends to sign a power purchase agreement with Dutch nuclear energy development company ULC-Energy. |
| SM019 | American Nuclear Society | NRIC's DOME open for business | NRIC is currently preparing for DOME's inaugural fueled experiment, which is slated to take place later this year. |
| SM020 | NucNet | Radiant Raises $165 Million For Development Of 'World's First' Mass Produced Nuclear Reactor | Radiant Industries announced it has raised $165m in funding to complete what it says could be the world's first mass-produced portable microreactor design. |
| SM021 | NucNet | Radiant Receives DOE Approval For Kaleidos Microreactor Safety Document | US microreactor developer Radiant Industries announced it has received approval from the US Department of Energy for a safety document covering its Kaleidos test reactor. |
| SM022 | U.S. Department of Energy | DOE Releases New Report Evaluating Increase in Electricity Demand from Data Centers | Data centers consumed about 4.4% of total U.S. electricity in 2023 and are expected to consume approximately 6.7 to 12% of total U.S. electricity by 2028. |
| SM023 | Belfer Center for Science and International Affairs | AI, Data Centers, and the U.S. Electric Grid: A Watershed Moment | AI-driven energy demand is outpacing available capacity, driving companies to delay projects, contract power directly from private producers, and/or install multiple, inefficient reciprocating generators using natural gas. |
| SM024 | Precedence Research | Small Modular Reactor Market Size to Hit USD 17.37 Bn By 2035 | The global small modular reactor market size is calculated at USD 7.49 billion in 2025 and is predicted to increase from USD 8.16 billion in 2026 to approximately USD 17.37 billion by 2035. |
| SM025 | MarketsandMarkets | Small Modular Reactor Market by Type ... Global Forecast to 2030 | The small modular reactor market is projected to reach USD 7.14 billion by 2030 at the CAGR of 3.0% during the forecast period. |
| SP001 | Radiant Nuclear | Radiant | |
| SP002 | Radiant Nuclear | Radiant closes $165 Million Series C with additional funding from Giant Ventures, StepStone, ARK Venture Fund, and others | |
| SP003 | Newswire / Radiant Nuclear | Radiant Announces Equinix Preorder & Deposits for 20 Kaleidos Microreactors | |
| SP004 | U.S. Nuclear Regulatory Commission | Kaleidos | Nuclear Regulatory Commission | |
| SP005 | U.S. Department of Energy | 3 Microreactor Experiments to Watch Starting in 2026 | |
| SP006 | U.S. Department of Energy | Department of Defense Breaks Ground on Project Pele Microreactor | |
| SP007 | Westinghouse Nuclear | eVinci™ Microreactor | Westinghouse Nuclear | |
| SP008 | World Nuclear News | Westinghouse, Radiant to perform first US microreactor tests | |
| SP009 | National Reactor Innovation Center | DOME Demonstrators | |
| SP010 | Idaho National Laboratory | INL advances Department of War’s Project Pele demonstration microreactor with first TRISO fuel delivery | |
| SP011 | U.S. Energy Information Administration | Small modular reactors and microreactors under development in the United States - U.S. Energy Information Administration (EIA) | |
| SP012 | Business Wire / Oklo | Oklo, Meta Announce Agreement in Support of 1.2 GW Nuclear Energy Development in Southern Ohio | |
| SP013 | Meta | Meta Announces Nuclear Energy Projects, Unlocking Up to 6.6 GW to Power American Leadership in AI Innovation | |
| SP014 | Securities and Exchange Commission | EDGAR Entity Landing Page | |
| SP015 | NANO Nuclear Energy | NANO Nuclear Energy Inc. | |
| SP016 | Securities and Exchange Commission | EDGAR Entity Landing Page | |
| SP017 | NuScale Power | The NuScale Power Module | NuScale Power | |
| SP018 | X-energy | Xe-100: High-Temperature Gas-Cooled Nuclear Reactors (HTGR) — X-energy | |
| SP019 | U.S. Nuclear Regulatory Commission | Long Mott Energy, LLC – Long Mott Generating Station Xe-100 Power Reactor Application | |
| SP020 | U.S. Department of Energy | Advanced Reactor Demonstration Projects | |
| SP021 | POWER Magazine | Dow and X-energy Advance Landmark Nuclear Project in Texas With Construction Permit Filing | |
| SP022 | Last Energy | Last Energy | 20 MWe SMR | Fully modular, factory made | |
| SP023 | Kairos Power | Technology | Kairos Power | |
| SP024 | World Nuclear Association | Small Modular Reactors - World Nuclear Association | |
| SP025 | Data Center Dynamics | Equinix signs three major deals with advanced nuclear firms for more than 750MW of power | |
| SP026 | American Nuclear Society | NRIC’s DOME “open for business” | |
| SI001 | Radiant Nuclear | Radiant | Customers can purchase through either Power Purchase Agreements (PPAs) or direct unit sales. |
| SI002 | Radiant Nuclear | Radiant closes $165 Million Series C with additional funding from Giant Ventures, StepStone, ARK Venture Fund, and others | Radiant announced today the successful close of its $165 million Series C funding round, bringing the company's total venture funding to $225 million. |
| SI003 | Radiant Nuclear | Radiant raises over $300 million in new funding to mass-produce portable nuclear reactors | Radiant today announced that it has raised more than $300 million in a new round of funding. |
| SI004 | Radiant Nuclear | Radiant to Build First Portable Nuclear Generator Factory on Tennessee Manhattan Project Site | With construction beginning in early 2026, this agreement keeps Radiant on track to deliver its first mass-produced Kaleidos nuclear generator by 2028, and, within a few years, scale up production to 50 reactors per year. |
| SI005 | Radiant Nuclear | Radiant on Track to Build the First Commercial Nuclear Microreactors Factory in America | The U.S. Nuclear Regulatory Commission (NRC) has formally accepted for review Radiant's 10 CFR Part 70 license application for its R-50 Production Facility in Tennessee. |
| SI006 | Newswire.com / Radiant Industries | Radiant Announces Equinix Preorder & Deposits for 20 Kaleidos Microreactors | Equinix, the world's digital infrastructure company, has signed a preorder agreement and submitted deposits for the purchase of 20 Kaleidos microreactors. |
| SI007 | Newswire.com / Radiant Industries | Radiant on Track to Build the First Commercial Nuclear Microreactors Factory in America | The R-50 Production Facility is designed to support Radiant's mission to deliver reliable, scalable, and innovative nuclear power for use everywhere. |
| SI008 | Newswire.com / Radiant Industries | Radiant Signs Agreement Designed to Deliver Nuclear Microreactor to U.S. Military Base in 2028 | Radiant announced today it has signed an agreement with the Defense Innovation Unit and the Department of the Air Force. |
| SI009 | Data Center Dynamics | Equinix-backed microreactor firm Radiant raises $300m in latest funding round | The latest funding round will support its commercialization efforts as it prepares to break ground on its R-50 factory in Oak Ridge, Tennessee. |
| SI010 | Data Center Dynamics | Equinix-backed microreactor firm Radiant to construct first reactor factory in Oak Ridge, Tennessee | Following this, in August, the company signed a £4 million ($5.1m) deal with nuclear enrichment firm Urenco to deliver high-assay low-enriched uranium (HALEU) fuel into the American market. |
| SI011 | World Nuclear News | Equinix signs further agreements with SMR developers | Equinix has signed a preorder agreement for the purchase of 20 of Radiant's Kaleidos microreactors. |
| SI012 | World Nuclear News | Radiant signs deal to supply microreactor for US military base | The contract was signed with the Department of Defense's Defense Innovation Unit and the U.S. Air Force as part of the Advanced Nuclear Power for Installations program. |
| SI013 | World Nuclear News | Radiant to locate microreactor factory in Tennessee | With construction of the factory set to begin in early 2026, Radiant said this agreement keeps it on track to deliver its first mass-produced Kaleidos nuclear generator by 2028, and, within a few years, scale up production to 50 reactors per year. |
| SI014 | Marketplace | Inside a company making nuclear reactors the size of a shipping container | The data center company Equinix has pre-ordered 20 of Radiant's microreactors. |
| SI015 | U.S. Department of Energy | Energy Department Announces First Microreactor Experiments in DOME Test Bed | The testing campaigns are self-funded by the applicants with the sequencing of experiments based on several criteria, including technology readiness, fuel availability, and a regulatory approval plan. |
| SI016 | U.S. Department of Energy | U.S. Department of Energy to Distribute First Amounts of HALEU to U.S. Advanced Reactor Developers | Many advanced reactors will need HALEU to achieve smaller designs, longer operating cycles, and increased efficiencies over current technologies, but HALEU is not currently available from domestic suppliers. |
| SI017 | U.S. Nuclear Regulatory Commission | Kaleidos | Nuclear Regulatory Commission | Radiant R-50 Microreactor Facility Greater than Critical Mass Application for Kaleidos Reactor Production ... Accepted. |
| SI018 | U.S. Nuclear Regulatory Commission | Radiant - R-50 Special Nuclear Materials License Application | With this letter, R-50, LLC (Radiant) is applying to the U.S. Nuclear Regulatory Commission (NRC) to obtain a greater than critical mass license under 10 CFR Part 70. |
| SI019 | Roane County, Tennessee | Radiant Selects Roane County to Build World's First Mass-Produced Nuclear Generator Factory | Radiant will invest $280 million to build its factory in Roane County, where it will create 175 high-quality jobs. |
| SI020 | WKRN News 2 | Sixth company locates in Tennessee through state's Nuclear Energy Fund | This week, Radiant Industries announced its plans to build the world's first mass-produced nuclear generator factory in Oak Ridge. It's the sixth company to locate in Tennessee through the Nuclear Energy Fund. |
| SI021 | Ashby / Radiant Industries | Radiant Industries Jobs | Open Positions (66). |
| SI022 | Union of Concerned Scientists | Analysis Published in Science Finds High Assay Low-Enriched Uranium Fuel to be Produced for Small Nuclear Power Reactors Poses a Greater Proliferation Threat than Previously Acknowledged | The authors of the Science paper concluded that HALEU above about 12% uranium-235 could be used to make practical weapons. |
| SI023 | Beyond Nuclear | Radiant booted from WY Heads to TN | Radiant's vague and indefinite plan being hauled back across the country and from around the world is not convincing of all points in between. |
| SI024 | American Bar Association | Pressure to succeed: Small modular (nuclear) reactor approvals on the horizon? | Despite all this progress, there is currently only one commercial North American SMR under construction—in Canada. |
| SI025 | American Nuclear Society | NRC to review Radiant R-50 Part 70 license application | Long-term plans call for production of up to 50 reactors per year. |
| SI026 | Nuclear Engineering International | Radiant closes Series C microreactor funding | The funding will primarily be used to complete Radiant's Kaleidos Development Unit, and for factory siting and early construction efforts. |
| SI027 | Tennessee Department of Economic and Community Development | Radiant Selects Tennessee to Build World's First Mass-Produced Nuclear Generator Factory | Radiant will create 175 new jobs in Roane County that will directly support the development and mass production of Radiant's Kaleidos. |
| SI028 | Upper Cumberland Business Journal | Tennesse selected for worlds first Mass-Produced Nuclear Generator Factory | Radiant has selected Tennessee for a $280 million investment that will expand the company's nuclear manufacturing and research and development to Oak Ridge, Tennessee. |
| SI029 | DCVC | Radiant raises $300 million+ Series D toward portable nuclear energy at scale | Radiant's $300M+ Series D values the company at over $1.8 billion. |
| SE001 | Radiant Nuclear | Radiant | |
| SE002 | Radiant Nuclear | Air Force selects Radiant to deliver microreactors to Buckley Space Force Base | |
| SE003 | Radiant Nuclear | Radiant Selected by Department of Energy as First New Nuclear Reactor Design to Be Tested in DOME | |
| SE004 | Radiant Nuclear | Radiant receives DOE approval of Preliminary Documented Safety Analysis, latest milestone towards start up of first reactor | |
| SE005 | Radiant Nuclear | Radiant to Build First Portable Nuclear Generator Factory on Tennessee Manhattan Project Site | |
| SE006 | U.S. Nuclear Regulatory Commission | Kaleidos | Nuclear Regulatory Commission | |
| SE007 | U.S. Department of Energy | Radiant Completes Study for First Kaleidos Microreactor Experiment | |
| SE008 | U.S. Department of Energy | 3 Microreactor Experiments to Watch Starting in 2026 | |
| SE009 | U.S. Department of Energy | Nuclear 101: What Is a High-Temperature Gas Reactor? | |
| SE010 | American Nuclear Society | NRC to review Radiant R-50 Part 70 license application | |
| SE011 | American Nuclear Society | Air Force selects three microreactor developers for ANPI | |
| SE012 | World Nuclear News | US microreactor clears initial stage of pre-testing safety review | |
| SE013 | World Nuclear News | Development and funding milestones for microreactor project | |
| SE014 | World Nuclear News | Westinghouse, Radiant to perform first US microreactor tests | |
| SE015 | POWER Magazine | Westinghouse, Radiant Selected for First Fueled Nuclear Microreactor Tests at INL’s DOME Facility | |
| SE016 | POWER Magazine | Air Force ANPI Picks Put Radiant, Antares, Westinghouse on Track for First On‑Base Microreactors by 2028 | |
| SE017 | Nuclear Engineering International | Radiant completes microreactor design phase | |
| SE018 | Nuclear Engineering International | US DOE approves safety design for Kaleidos microreactor | |
| SE019 | Interesting Engineering | US firm's mass-producible Kaleidos nuclear reactor gets key approval | |
| SE020 | Partnership for Global Security | Status of the DoE Reactor Pilot Project | |
| SE021 | Idaho National Laboratory | INL Advanced Reactor Technologies - ART Program | |
| SE022 | Built In | Mechanical Engineer - 2026 New Grad (Fluid Systems, Mechanisms, or Structures) - Radiant Nuclear | |
| SE023 | Decisive Point Job Board | Reactor Operations Engineer, Nuclear | |
| SE024 | Marketplace | Inside a company making nuclear reactors the size of a shipping container | |
| SE025 | Beyond Nuclear | Radiant booted from WY Heads to TN | |
| SE026 | American Bar Association | Pressure to succeed: Small modular (nuclear) reactor approvals on the horizon? | |
| SE027 | Newswire / Radiant Industries | Radiant Announces Equinix Preorder & Deposits for 20 Kaleidos Microreactors | |
| SE028 | World Nuclear News | Equinix signs further agreements with SMR developers | |
| SE029 | Data Center Dynamics | Equinix signs three major deals with advanced nuclear firms for more than 750MW of power | |
| SE030 | U.S. Nuclear Regulatory Commission / Radiant | Radiant - R-50 Special Nuclear Materials License Application | |
| SE031 | Oak Ridge National Laboratory / NRC document server | SCALE Analyses of Scenarios in the TRISO-based HPMR Fuel Cycle | |
| SU001 | Radiant Nuclear | Radiant | |
| SU002 | Radiant Nuclear | Radiant Signs Agreement Designed to Deliver Nuclear Microreactor to U.S. Military Base in 2028 | This is the first-ever agreement designed to deliver a mass-manufactured nuclear microreactor to a U.S. military base. |
| SU003 | Radiant Nuclear | Air Force selects Radiant to deliver microreactors to Buckley Space Force Base | The ANPI program seeks to have at least one advanced nuclear reactor operating on at least one DAF installation by 2030 or sooner. Next steps include siting and environmental analyses as part of the National Environmental Policy Act process. |
| SU004 | Radiant Nuclear | Radiant Announces Equinix Preorder & Deposits for 20 Kaleidos Microreactors | Equinix ... has signed a preorder agreement and submitted deposits for the purchase of 20 Kaleidos microreactors. |
| SU005 | Equinix | Equinix Collaborates with Leading Alternative Energy Providers to Power AI-Ready Data Center Growth | Today Equinix announced a preorder agreement for the purchase of 20 of Radiant's Kaleidos microreactors. |
| SU006 | World Nuclear News | Equinix signs further agreements with SMR developers | |
| SU007 | Data Center Dynamics | Equinix signs three major deals with advanced nuclear firms for more than 750MW of power | |
| SU008 | Nuclear Engineering International | Equinix signs nuclear data centre deals - Nuclear Engineering International | |
| SU009 | Marketplace | Inside a company making nuclear reactors the size of a shipping container | |
| SU010 | Department of the Air Force | DAF announces next steps in Advanced Nuclear Power for Installations initiative | The ANPI initiative seeks to have at least one advanced nuclear reactor operating on at least one DAF installation by 2030 or sooner. Next steps include siting and environmental analyses as part of the National Environmental Policy Act process. |
| SU011 | U.S. Space Force | DAF announces next steps in Advanced Nuclear Power for Installations initiative | |
| SU012 | Defense Innovation Unit | DoD Names Eligible Companies for Advanced Nuclear Power for Installations | The companies are now eligible to receive Other Transaction awards to provide commercially available dual use microreactor technology at various DOD installations. |
| SU013 | American Nuclear Society | Radiant signs contract on microreactors for the military | |
| SU014 | American Nuclear Society | Air Force selects three microreactor developers for ANPI | |
| SU015 | Breaking Defense | Department of Air Force picks bidders for nuclear microreactors, assigns locations - Breaking Defense | Critics, meanwhile, have argued that ANPI's benefits are illusory, and come with too much cost and risk. |
| SU016 | World Nuclear News | US Air Force announces selections for microreactor deployments | |
| SU017 | POWER Magazine | Air Force ANPI Picks Put Radiant, Antares, Westinghouse on Track for First On‑Base Microreactors by 2028 | |
| SU018 | Partnership for Global Security | Department of Defense Advanced Reactor Programs - Partnership for Global Security | |
| SU019 | FedTech | Insight | Microreactors, Macro Strategy: How the DoD's SMR Push Could Reshape Energy Innovation and Infrastructure | |
| SU020 | Goldman Sachs Research | Is nuclear energy the answer to AI data centers’ power consumption? | Nuclear is the preferred option for baseload power, but the difficulty of building new nuclear plants means that natural gas and renewables are more realistic short-term solutions. |
| SU021 | Data Center Knowledge | Equinix Bets on Nuclear Power to Fuel Data Center Energy Demands | Many of these next-generation nuclear technologies are still several years away from deployment, so it's a little early to be thinking about site selection. |
| SU022 | Data Center Frontier | Equinix Bets on Nuclear and Fuel Cells to Meet Exploding Data Center Energy Demand | |
| SU023 | Axios | Neighbors react to nuclear plan at Buckley Space Force base | Neighbors in the shadows of Buckley Space Force Base say the April announcement naming the base a preferred site for a nuclear project came without warning. |
| SU024 | Equinix | Equinix Collaborates with Leading Alternative Energy Providers to Power AI-Ready Data Center Growth | |
| SU025 | U.S. Department of Energy | DOE Releases New Report Evaluating Increase in Electricity Demand from Data Centers | |
| SR001 | Nuclear Regulatory Commission | Regulatory Activities | |
| SR002 | Pillsbury Winthrop Shaw Pittman LLP | NRC Proposes Part 57 Framework for Microreactor Licensing | |
| SR003 | Utility Dive | Nuclear regulators lighten microreactor restrictions | |
| SR004 | American Nuclear Society | NRC to review Radiant R-50 Part 70 license application | |
| SR005 | U.S. Space Force | DAF announces next steps in Advanced Nuclear Power for Installations initiative | |
| SR006 | Defense Innovation Unit | DoD selects eligible companies for the Advanced Nuclear Power for Installations Program | |
| SR007 | Breaking Defense | Department of Air Force picks bidders for nuclear microreactors, assigns locations | |
| SR008 | KUNC | The Pentagon wants to operate a nuclear microreactor in Colorado. Here’s what that might look like | |
| SR009 | Axios | Neighbors react to nuclear plan at Buckley Space Force Base | |
| SR010 | U.S. Department of Energy | HALEU Availability Program | |
| SR011 | U.S. Department of Energy | U.S. Department of Energy HALEU Allocation Process | |
| SR012 | U.S. Department of Energy | U.S. Department of Energy to Distribute First Amounts of HALEU to U.S. Advanced Reactor Developers | |
| SR013 | World Nuclear News | DOE selects first recipients of HALEU | |
| SR014 | World Nuclear News | DOE delivers HALEU feedstock for advanced reactor fuel | |
| SR015 | Centrus Energy | Centrus Energy secures contract extension from Department of Energy to continue HALEU production | |
| SR016 | BWX Technologies | BWXT delivers full core of TRISO nuclear fuel for Project Pele microreactor | |
| SR017 | BWX Technologies | Kairos Power and BWXT to collaborate on commercial TRISO manufacturing | |
| SR018 | The National Law Review | Department of Energy Establishes National Environmental Policy Act Categorical Exclusion for Advanced Reactors | |
| SR019 | National Conference of State Legislatures | News Reactor | May 2026 | |
| SR020 | POWER Magazine | NRC unveils Part 57: A streamlined path for high-volume microreactor licensing | |
| SR021 | Idaho National Laboratory | New study examines US markets for microreactors | |
| SR022 | American Bar Association | The pressure to succeed: small modular nuclear reactors | |
| SR023 | Union of Concerned Scientists | Analysis published in Science finds HALEU fuel can be used to make nuclear weapons | |
| SR024 | Beyond Nuclear | Radiant booted from Wyoming, heads to TN | |
| SR025 | U.S. Department of Energy | 3 Microreactor Experiments to Watch Starting in 2026 | |
| SR026 | Marketplace | Inside a portable nuclear reactor company | |
| SR027 | Data Center Frontier | Fuel cells and next-gen nuclear give Equinix the capacity to power tomorrow’s digital infrastructure | |
| SR028 | Data Center Knowledge | Equinix bets on nuclear power to fuel data center energy demands | |
| SR029 | Bloom Energy | Data centers plan to reduce reliance on grid finds Bloom Energy’s 2026 power report | |
| SR030 | Business Wire | Data Centers Plan to Reduce Reliance on Grid Finds Bloom Energy’s 2026 Power Report | |
| SR031 | DCVC | Radiant raises $300 million Series D toward portable nuclear energy at scale | |
| SR032 | Data Center Dynamics | Equinix-backed microreactor firm Radiant raises $300m in latest funding round | |
| SR033 | Tennessee Department of Economic and Community Development | Radiant selects Tennessee to build world’s first mass-produced nuclear generator factory | |
| SV001 | Radiant Nuclear | Radiant | Customers can purchase Kaleidos either through direct sales or through a power purchase agreement. |
| SV002 | Radiant Nuclear | Radiant closes $165 Million Series C with additional funding from Giant Ventures, StepStone, ARK Venture Fund, and others | With this raise, total venture funding raised by the company now stands at $225 million. |
| SV003 | Radiant Nuclear | Radiant raises over $300 million in new funding to mass-produce portable nuclear reactors | Radiant has raised over $300 million in new funding to support commercialization of its portable Kaleidos microreactors. |
| SV004 | Radiant Nuclear | DOME selection | Radiant will become the first reactor design to test at DOME in Spring 2026. |
| SV005 | Radiant Nuclear | DOE PDSA approval | Radiant is the first reactor company to secure approval of a Preliminary Documented Safety Analysis. |
| SV006 | Radiant Nuclear | Buckley Space Force | The Department of the Air Force selected Radiant to work toward delivering a reactor to Buckley Space Force Base. |
| SV007 | Radiant Nuclear | Factory site license | Radiant is on track to deliver its first factory-built microreactors in 2028 and scale to 50 reactors per year within a few years. |
| SV008 | Newswire | Radiant announces Equinix preorder deposits for 20 Kaleidos | Equinix ... has signed a preorder agreement and submitted deposits for the purchase of 20 Kaleidos microreactors. |
| SV009 | World Nuclear News | Equinix signs further agreements with SMR developers | Equinix has signed agreements with Oklo, Radiant, Rolls-Royce SMR and Stellaria. |
| SV010 | Data Center Dynamics | Equinix signs three major deals with advanced nuclear firms for more than 750MW of power | Equinix signed agreements with Oklo, Radiant, and Deep Fission-linked developers as part of a broader power strategy. |
| SV011 | Data Center Dynamics | Equinix-backed microreactor firm Radiant raises $300m in latest funding round | Radiant has raised over $300 million in funding as it prepares to build its Tennessee factory. |
| SV012 | DCVC | Radiant raises $300 million+ Series D toward portable nuclear energy at scale | Radiant's $300M+ Series D ... values the company at over $1.8 billion. |
| SV013 | Tennessee Department of Economic and Community Development | Radiant selects Tennessee to build world's first mass-produced nuclear generator factory | Radiant officials announced the company has selected Tennessee for a $280 million investment. |
| SV014 | Roane County, Tennessee | Radiant Selects Roane County to Build World's First Mass-Produced Nuclear Generator Factory | Radiant will invest $280 million to build its factory in Roane County, where it will create 175 high-quality jobs. |
| SV015 | Marketplace | Inside a company making nuclear reactors the size of a shipping container | Bernauer said he expects Radiant’s first microreactors will be commercially available by 2028 at the latest. |
| SV016 | U.S. Department of Energy | Energy Department Announces First Microreactor Experiments in DOME Test Bed | The testing campaigns are self-funded by the applicants ... based on technology readiness, fuel availability, and a regulatory approval plan. |
| SV017 | U.S. Department of Energy | U.S. Department of Energy to Distribute First Amounts of HALEU to U.S. Advanced Reactor Developers | HALEU is not currently available from domestic suppliers. |
| SV018 | U.S. Nuclear Regulatory Commission | Kaleidos | Nuclear Regulatory Commission | Radiant R-50 Microreactor Facility Greater than Critical Mass Application ... Accepted. |
| SV019 | U.S. Nuclear Regulatory Commission | Radiant - R-50 Special Nuclear Materials License Application | Radiant is applying to the U.S. Nuclear Regulatory Commission to obtain a greater than critical mass license under 10 CFR Part 70. |
| SV020 | Union of Concerned Scientists | Analysis Published in Science Finds High Assay Low-Enriched Uranium Fuel to be Produced for Small Nuclear Power Reactors Poses a Greater Proliferation Threat than Previously Acknowledged | HALEU above about 12% uranium-235 could be used to make practical weapons. |
| SV021 | Beyond Nuclear | Radiant booted from WY Heads to TN | Radiant's vague and indefinite plan ... hauled back across the country ... is not convincing. |
| SV022 | American Bar Association | Pressure to succeed: Small modular (nuclear) reactor approvals on the horizon? | There is currently only one commercial North American SMR under construction—in Canada. |
| SV023 | TechCrunch | Radiant Nuclear raises $300M for its semi-sized 1 MW reactor | Given the string of investments, it seems reasonable to ask whether the nuclear world is in a bubble. |
| SV024 | Stock Analysis | Oklo Inc. (OKLO) Statistics & Valuation | Oklo Inc. has a market cap of $11.33 billion. Revenue n/a. |
| SV025 | U.S. Securities and Exchange Commission | Oklo Quarterly Report on Form 10-Q | Forward-looking statements include statements concerning ... future revenues ... capital requirements ... and the adequacy of our funding. |
| SV026 | Stock Analysis | NuScale Power (SMR) Statistics & Valuation | NuScale Power has a market cap of $4.14 billion and revenue of $18.67 million. |
| SV027 | U.S. Securities and Exchange Commission | NuScale Power Quarterly Report on Form 10-Q | For the three months ended March 31, 2026 ... Power Plant and NPM related services $487 [thousand]; Energy Exploration Centers $78 [thousand]. |
| SV028 | Stock Analysis | NANO Nuclear Energy (NNE) Statistics & Valuation | NNE has a market cap of $1.31 billion. Revenue n/a. |
| SV029 | U.S. Securities and Exchange Commission | NANO Nuclear Energy Annual Report on Form 10-K | As of the date of this Report, we have not generated any revenues. |
| SV030 | X-energy | X-energy closes oversubscribed $700 million Series D financing round | X-energy ... announced it has closed an oversubscribed Series D financing round of approximately $700 million. |
| SV031 | Helion | Helion announces $425M Series F investment | This latest round ... values the company at $5.425 billion post-money. |
| SV032 | Business Wire | Last Energy Announces Oversubscribed $100 Million Series C | Last Energy ... announced it has closed an oversubscribed Series C of more than $100 million. |
| SV033 | POWER Magazine | Advanced nuclear developers raise new capital as 2025 investment hits record levels and demonstrations near | By the beginning of the third quarter of 2025, nuclear fission companies had already raised $1.3 billion in equity funding—the sector’s highest annual total on record. |