scmp.com via Reddit

China's 36-Reactor Buildout Targets 110GW by 2030

china ai ai infrastructure ai-infrastructure energy

Key insights

  • China had 36 reactors under construction as of April, comprising nearly half of all reactors being built globally.
  • Beijing targets 110 gigawatts of nuclear capacity by 2030, driven partly by surging AI data center electricity demand.
  • Analyst Damien Ma of Gavekal Technologies projects China will dominate the global nuclear industry 'by a wide margin' through 2035.

Why this matters

Nuclear baseload capacity is becoming a structural constraint on AI infrastructure at scale, and China's 36-reactor construction pipeline gives it a reliability and cost advantage that takes years to close. The US fleet has added only three reactors since 1996, meaning the output gap will widen precisely as AI compute-driven electricity demand accelerates through the rest of the decade. Beijing's 110GW target by 2030, backed by the world's most active construction program and two concurrent demand drivers -- AI growth and Strait of Hormuz disruptions -- positions China to offer operators cheaper, more reliable zero-carbon power at a moment when data center energy procurement has become a front-line competitive issue.

Summary

China is on track to overtake the US as the world's largest nuclear power producer within five years, driven by AI data center electricity demand and disruptions from the prolonged Strait of Hormuz closure. As of April, China had 36 reactors under construction -- nearly half the global total -- and is targeting 110 gigawatts of installed capacity by 2030 under its latest five-year plan. The US still holds the world's largest active nuclear fleet but has added only three new reactors since 1996. Essentially: (China, US) are moving in opposite directions on nuclear output at the moment reliable baseload is becoming a core AI infrastructure asset. - China's 36 active construction projects make up close to half of all reactors being built worldwide. - Beijing's 110GW target is driven by both AI sector demand and Strait of Hormuz energy market disruptions. - Gavekal Technologies analyst Damien Ma: China leads the global nuclear industry "by a wide margin" through 2035. The US retains an edge in nuclear fusion research, but in operational fission capacity -- the kind powering data centers today -- China's trajectory has no near-term peer.

Potential risks and opportunities

Risks

  • The US's near-zero reactor construction pace -- three new units since 1996 -- leaves American AI data center operators structurally dependent on fossil and intermittent renewable power, compounding a cost-competitiveness gap as training and inference workloads scale.
  • If any of China's 36 active reactor construction projects hit safety, regulatory, or supply-chain delays, the 2030 110GW target slips, threatening grid reliability for AI infrastructure built around nuclear baseload assumptions.
  • A worsening Strait of Hormuz situation that is partly driving China's nuclear push could disrupt global supply chains for reactor components, introducing delivery risk into the very build timeline the capacity projections depend on.

Opportunities

  • Chinese state nuclear engineering enterprises, now the world's most prolific reactor builders by active construction volume, are positioned to expand international nuclear export deals as other countries seek to replicate the model.
  • The visible US-China nuclear capacity gap gives American policymakers a concrete AI competitiveness framing to accelerate reactor permitting reform, a policy lever that has resisted action for decades.
  • AI data center developers globally will increasingly weight nuclear proximity and committed baseload capacity as a primary site selection criterion, advantaging regions with active construction programs over those relying on intermittent renewables.

What we don't know yet

  • Whether China's 110GW target counts only new capacity to be added by 2030 or includes existing operational reactors in the total.
  • How quantitatively significant the Strait of Hormuz closure has been relative to AI demand in accelerating China's nuclear build -- the article cites both drivers but does not break down their relative weight.
  • Whether the US edge in nuclear fusion research translates into any operational timeline that would change the fission-based capacity comparison before 2035.