来自 Maksym Misichenko · Nasdaq ·
AI智能体对这条新闻的看法
The panel consensus is bearish on OKLO and NuScale, citing substantial risks including heavy cash burn, long regulatory timelines, competition from other energy sources, and the challenge of securing project financing and offtake agreements. The 'AI power demand' thesis is considered speculative, with a potential inflection point not expected until the 2030s or later.
风险: Heavy cash burn and long regulatory timelines
机会: Potential long-term growth driven by AI-driven power demand
本分析由 StockScreener 管道生成——四个领先的 LLM（Claude、GPT、Gemini、Grok）接收相同的提示，并内置反幻觉防护。 阅读方法论 →
核能股票位于清洁能源和 AI 发展交叉点。
Oklo 正在设计可部署的现场发电反应堆。
NuScale 拥有 NRC 批准的 SMR 设计，可帮助公用事业公司提升电力容量。
当你想到公用事业公司时，可能不会想到核能。但随着能源格局的变化，总有一天你会把核能放在首位。
美国已经是全球最大的核电生产国，约 30% 的全球发电量在其境内。特朗普时期的白宫计划到 2050 年将该产能翻四倍，先进核技术，如小型模块化反应堆（SMR），预计将成为新核时代的主力。
AI 会造就世界首位万亿美元富豪吗？ 我们团队刚发布了一份关于一家鲜为人知公司的报告，称其为“不可或缺的垄断”，提供 Nvidia 和 Intel 都需要的关键技术。继续 »
谁知道呢：到 2030 年，我们甚至可能看到首个核反应堆被部署到太空。
无论是否发生，这都表明核电将在地球上扩大足迹。以下核能股票是未来数十年核能扩张的必备之选。
Oklo（NYSE: OKLO）是“新核能”的典型代表。这家先进核能公司正在开发一种名为 Aurora 的小型快堆，使用液态金属冷却剂，燃料既可以是新鲜铀，也可以是回收的核废料。
如果这些核能术语让你摸不着头脑，核心意思是：Oklo 想建造一种小型反应堆，能够提供稳定电力，而无需传统核电站那巨大的占地面积和成本。
这里的重点是“想”。过去几年，Oklo 的大想法一直卡在以慢著称的监管流程中。核监管委员会（NRC）尚未授予 Oklo 的 Aurora 设计商业运营许可证，也没有何时会获批的时间表。因此，Oklo 正在烧钱，第一季度净亏损约 3300 万美元，靠其 16 亿美元的现金及等价物维持运营。
但事情似乎在推进。公司最近在爱达荷国家实验室启动了首座发电站，并宣布与 Switch、Meta 等科技巨头潜在合作。其近期收购 Atomic Alchemy 可能打开数十亿美元的医用同位素市场，甚至在反应堆上线前就产生收入。
然而此时，投资者正为一部上映日期仍是大问号的电影付出高价。相信核能是未来的激进投资者，可能更愿意承担这家早期核能公司带来的风险。
NuScale Power（NYSE: SMR）常被拿来与 Oklo 对比，视为潜在竞争对手。两者有相似之处：和 Oklo 一样，NuScale 也想建造小型核反应堆，尤其是为满足人工智能（AI）及其数据中心的电力需求而填补空白。
但 NuScale 在若干关键方面与其竞争对手不同。首先，它已经获得 NRC 批准，不止一个，而是两个小型模块化反应堆（SMR）设计：一个 50 兆瓦（MW）设计和一个升级版 77 MW 设计。NuScale 仍是唯一一家在美国获得 SMR 监管批准的核能公司。
第二个区别在于两家公司针对的客户。虽然可能有一定重叠，但 NuScale 的反应堆更适合公用事业公司和大型基础设施项目，因而更适合大规模部署。NuScale 与田纳西河流域管理局（TVA）潜在的 6 GW 项目凸显了其反应堆可能达到的规模。
相对而言，Oklo 更适合规模较小或一次性项目，需求的反应堆数量或功率较低，例如数据中心使用的项目。
尽管 NuScale 在监管上抢先 Oklo，但它尚未产生显著收入，也未实现首笔确定性销售。这并不难理解：SMR 项目虽比传统电站小，却仍是多年巨额投入，公用事业公司并不会像你我买新手机那样轻易采购反应堆。
总的来说，Oklo 拥有（大客户），NuScale 缺乏；而 NuScale 拥有（NRC 批准的 SMR 设计），Oklo 缺乏。因此，两者近期都将面临极度波动的局面，因为它们各自业务中存在的问题。长期来看，两只股票都有巨大的价值爆发潜力，尤其是当高功耗数据中心继续建设时。
在买入 Oklo 股票前，请考虑以下因素：
Motley Fool Stock Advisor 分析团队刚刚确定了他们认为的10 只最佳股票，供投资者现在买入…而 Oklo 并未入选。入选的 10 只股票在未来几年可能带来巨额回报。
想想 Netflix 在 2004 年 12 月 17 日上榜时…如果你在我们推荐时投入 1,000 美元，你将拥有 471,072 美元！ 或 Nvidia 在 2005 年 4 月 15 日上榜时…如果你在我们推荐时投入 1,000 美元，你将拥有 1,303,352 美元！
值得注意的是，Stock Advisor 的平均总回报率为 983%，远超标普 500 指数的 210%。不要错过最新的前 10 名名单，使用 Stock Advisor 获取，并加入由个人投资者为个人投资者打造的投资社区。
*Stock Advisor 截至 2026 年 5 月 28 日的回报。
Steven Porrello 持有 NuScale Power 和 Oklo 的仓位。The Motley Fool 持有并推荐 Meta Platforms。The Motley Fool 推荐 NuScale Power。The Motley Fool 有披露政策。
本文所表达的观点和意见仅代表作者本人，并不一定代表 Nasdaq, Inc. 的立场。
四大领先AI模型讨论这篇文章
"Execution and regulatory delays will likely keep both OKLO and SMR cash-flow negative well into the 2030s despite the nuclear-AI narrative."
The article frames OKLO and SMR as core plays on AI-driven power demand and a Trump-backed nuclear buildout, yet both remain pre-revenue with heavy cash burn—OKLO posted a $33M Q1 loss against $1.6B cash, while SMR holds NRC approval but zero firm sales. Procurement cycles for utilities stretch years, and NRC timelines remain opaque. The piece underplays execution risk: even approved designs face multi-year site licensing, supply-chain bottlenecks, and competition from gas or renewables-plus-storage. AI hyperscalers may prefer proven or behind-the-meter solutions over waiting for first-of-a-kind SMRs.
A single large direct PPA with a data-center operator could validate both business models overnight and compress the perceived regulatory timeline, turning current cash burn into growth capital.
"Regulatory approval is not a business model; neither company has a single paying customer or operational reactor, and the article's conflation of policy tailwinds with near-term revenue is the classic pre-revenue stock trap."
The article conflates regulatory approval with commercial viability—a critical error. NuScale has NRC approval for designs, but zero deployed units and zero firm orders. OKLO has $1.6B cash but burns $33M quarterly; at that rate, runway is ~4 years before needing capital raises that will dilute shareholders. The 'AI power demand' thesis is real, but it assumes: (1) utilities will actually build SMRs (capex-intensive, long timelines), (2) data centers won't solve this via efficiency or renewables + storage, (3) regulatory timelines accelerate (historically false). Both companies are pre-revenue bets on a 2030+ inflection that may never arrive at scale.
If Trump's 4x nuclear capacity target becomes law with expedited permitting, and if even 2-3 major data center operators commit to SMR power-purchase agreements in the next 18 months, both stocks could re-rate 3-5x on visibility alone—regardless of actual generation.
"The market is dangerously overvaluing pre-revenue SMR developers by ignoring the immense capital expenditure and regulatory hurdles inherent in nuclear infrastructure projects."
The market is currently pricing these SMR plays as if they are software-as-a-service companies, ignoring the brutal reality of capital-intensive, highly regulated infrastructure. While the AI data center narrative is compelling, the 'nuclear renaissance' faces a massive bottleneck: supply chain maturity and fuel availability. Oklo and NuScale are essentially pre-revenue R&D projects with significant execution risk. Unless these firms can demonstrate a viable path to project financing that doesn't involve massive shareholder dilution, they remain speculative lottery tickets. Investors are conflating 'total addressable market' with 'investable business model.' The real winners here won't be the SMR startups, but the legacy utility giants and uranium miners that provide the actual infrastructure.
If these SMR designs achieve standardized, factory-line production, they could drastically lower the levelized cost of energy (LCOE) and trigger a massive valuation re-rating that makes current prices look like a bargain.
"The risk-adjusted upside for early-stage SMR bets is lower than the hype suggests unless regulatory, financing, and policy catalysts materialize sooner than consensus."
Despite the rosy framing, the core risks for Oklo and NuScale are substantial and not fully addressed in the piece. Oklo’s Aurora license remains unsettled, and its cash burn persists while it builds pilot capacity. NuScale has NRC-approved 50 MW and 77 MW designs, but no firm utility agreements; the TVA 6 GW-scale potential illustrates scale rather than certainty. SMR economics demand huge upfront capex, long permitting, and dependable grid integration—any hiccup in financing or policy slows adoption or raises costs. A policy drag or cheaper competing technologies could erode the upside; a favorable tailwind is possible, but base-case odds favor protracted, lumpy returns.
Counterpoint: If policy stays supportive and utilities push SMR procurements, licensing and deals could accelerate, offering a sharper near-term upside that tilts risk-reward toward bulls.
"HALEU scarcity hits advanced SMRs and legacy nuclear builds alike, erasing the edge Gemini assigns to incumbents."
Gemini's claim that legacy utilities and uranium miners will capture the gains ignores the HALEU fuel bottleneck specific to advanced designs like OKLO's Aurora. Current DOE output remains under 1 metric ton annually versus the tens of tons needed for even a handful of deployments. This shared constraint could delay every new reactor equally, flattening the supposed advantage for incumbents and extending cash-burn timelines for all players.
"HALEU bottleneck delays all players equally, but policy-backed startups may navigate it faster than legacy utilities constrained by existing commitments."
Grok's HALEU constraint is real, but it cuts differently than claimed. Yes, it delays everyone—but OKLO and NuScale benefit from policy tailwinds (DOE expansion commitments, Biden-era fuel reserve builds). Legacy utilities face *existing* fleet pressures; they can't wait years for fuel supply to mature. Startups with patient capital and regulatory momentum may actually move *faster* through the bottleneck than incumbents hamstrung by legacy operations. Gemini's 'uranium miners win' thesis assumes incumbents have structural advantage; HALEU scarcity may invert that.
"HALEU scarcity favors utilities with deep balance sheets over pre-revenue startups, effectively pricing the latter out of the fuel market."
Claude, your optimism regarding 'patient capital' for startups during a HALEU shortage is misplaced. While legacy utilities face fleet constraints, they possess balance sheets capable of securing long-term fuel contracts that pre-revenue firms cannot touch. OKLO and NuScale are price-takers in a supply-constrained market. If HALEU remains scarce, these startups won't just be delayed; they will be outbid by major utilities or sovereign-backed entities, rendering their 'regulatory momentum' irrelevant when they lack the fuel to actually operate.
"HALEU supply constraints won't unlock SMR value without bankable PPAs and financing for regulated revenue."
Gemini suggests incumbents win post-HALEU. I disagree: even if HALEU scales, the central hurdle remains project finance and offtake risk. Utilities can defer or outbid on fuel; more determinative is whether Oklo/NuScale can secure bankable PPAs and grid interconnections at a reasonable LCOE. Without a credible, near-term path to regulated revenue, HALEU is not a rescue, it's just a gating item that buys time.
The panel consensus is bearish on OKLO and NuScale, citing substantial risks including heavy cash burn, long regulatory timelines, competition from other energy sources, and the challenge of securing project financing and offtake agreements. The 'AI power demand' thesis is considered speculative, with a potential inflection point not expected until the 2030s or later.
Potential long-term growth driven by AI-driven power demand
Heavy cash burn and long regulatory timelines