SpaceX’s completed IPO has put fresh capital in focus, and the company’s own prospectus points to a clear priority: scaling artificial intelligence workloads. SpaceX argues that the largest addressable market it sees is overwhelmingly tied to AI—while warning that the binding constraint for AI growth is energy supply for data centers.
With AI demand increasing, the debate for investors is shifting toward power availability and the technology that can deliver large, reliable electricity faster. In that context, nuclear energy—particularly small modular reactors—has drawn renewed attention from major power users and AI-linked customers.
Key takeaways
- What moved: The post-IPO discussion around SpaceX centers on AI expansion, with energy supply identified as the key bottleneck for data centers.
- Catalyst: SpaceX’s IPO prospectus highlighted constrained global energy availability for AI compute.
- Key implication: Investors are likely to scrutinize power-generation options that can scale quickly, including small modular reactors.
- What’s uncertain: Small modular reactor commercialization timelines remain unproven, and economics face challenges from cost variability and intensifying competition from solar and wind.
What drove the focus on AI after SpaceX’s IPO
SpaceX’s IPO prospectus frames the company’s long-term growth around artificial intelligence. According to the prospectus, SpaceX estimates a total addressable market of $28.5 trillion, with more than 90% of that opportunity linked to AI.
The prospectus also lays out why AI scaling is not simply a software or hardware story. It warns that AI technology depends on data centers, and that the required infrastructure is energy-intensive. According to SpaceX’s disclosures, energy supply is constrained globally due to rising demand and limited availability of power to run AI compute.
The energy constraint—and the debate over “orbital” data centers
In the long run, SpaceX suggests orbital data centers as a potential solution, describing data centers placed in orbit around Earth that would harness solar energy. However, many industry experts remain skeptical about when—or whether—orbital data centers can be commercialized.
That skepticism matters for near- to medium-term investment planning. If orbital infrastructure takes longer than expected, AI expansion would still require terrestrial power sources to support data center build-outs. The prospectus discussion also implies a portfolio approach: renewables such as wind and solar could contribute, but the scale and reliability required for large AI loads may increase scrutiny of baseload generation options.
Why nuclear and small modular reactors are back in the spotlight
Big tech companies have shown renewed interest in nuclear energy as a way to supply steady baseload power while limiting additional carbon emissions. According to the article, Alphabet is involved in the construction and rehabilitation of several nuclear energy sites.
One of the barriers for conventional nuclear power plants is timing: large projects often take a decade or more from start to power delivery. Small modular reactors, or SMRs, are positioned as a potential workaround, with proponents citing less land needed, shorter construction periods, and enhanced safety features.
Bank of America’s report, cited in the article, argues that advancements in SMR technology could reshape nuclear power supply chains over the next decade. The bank’s view is framed around the potential supply benefits relative to conventional plants, in a world where energy demand tied to computing infrastructure continues to rise.
Two pure-play SMR bets: different customer strategies, shared execution risk
The article highlights two SMR-focused companies: NuScale Power and Oklo. It describes NuScale as pursuing grid-scale deployments mainly through partnerships with electric utilities. By contrast, Oklo is described as targeting data center operators more directly, a strategy consistent with the buyer profile of AI infrastructure demand.
Both companies, as presented in the article, have customer pipelines, but neither has yet commercialized an SMR system. That execution gap is central to investor risk: deals on paper may not translate into operating projects, and AI power buyers may hesitate to commit large resources to technologies that remain unproven at commercial scale.
Bank of America also cautions that SMR economics are complex. According to the article, proponents point to modular design advantages and potential cost benefits, while critics warn about current cost unpredictability. The report cited in the article notes that SMRs can face cost overruns and delays in demonstration projects, raising questions about economic viability—especially as renewable technologies such as solar and wind continue to improve their cost competitiveness.
Taken together, the energy-storage and power-supply bottleneck described in SpaceX’s prospectus is intersecting with a more demanding investment environment for new generation technologies. For investors, the question is not only whether SMRs can scale, but also whether their projected economics will hold up against faster-moving alternatives.
Bigger picture: what investors will track next
Next, investors are likely to focus on evidence that AI-driven power demand can be met with reliable supply—both in timing and cost. For nuclear and SMR-related plays, watch for progress from demonstration activity toward commercial operations, along with clearer timelines and contracting structures with utilities or data center customers.
More broadly, developments in AI compute expansion plans, grid capacity updates, and follow-on regulatory and financing milestones for power-generation projects could shape expectations for how quickly new electricity supply will come online.
