ESS and power infrastructure have become the key energy foundation for AI compute expansion

Overview

This report is the J.P. Morgan Global China Summit meeting note on 'ESS and Power Infrastructure's Role in the AI Race.' The core discussion covered AI data center expansion, improving energy-storage business models, grid infrastructure upgrades, and renewed attention on nuclear power amid APAC energy-security concerns. Rather than a single-company deep dive, the report presents thematic views and a focused OW stock basket across China's utilities and renewable-energy chain.

Core views

The core views include: first, China's ESS is shifting from policy-driven to market-based and value-driven economics, with power-price mechanisms and capacity fees improving independent storage returns; second, AIDC's high power density creates huge peak-load and grid-connection challenges, making storage an important configuration for peak shaving, load smoothing, and lowering electricity costs; third, aging grids, AIDCs, electrification, and renewable-energy integration are driving global grid-upgrade investment; fourth, energy-security concerns are bringing APAC nuclear power back into focus; fifth, the meeting observations suggest rising demand for distributed-generation ESS in Southeast Asia and improving policy clarity for China's offshore wind, while some utilities and renewables segments still face risks from cost pass-through, market-driven power pricing, and the pace of industry consolidation.

Analysis framework

The report uses a panel-remark and thematic-investing format, combining Global China Summit expert views with macro themes including developed-market data center deployment, China's power-market reform, the energy transition, energy security, AI compute-load growth, and underlying industry-cycle factors, and then screens for key OW stocks in China's utilities and renewable-energy chain.

Methodology notes

• Theme investingAI compute energy infrastructure framework

  Maps AIDC's high power-density demand to ESS, grid upgrades, and power-equipment investment opportunities.

  AIDC rack power density is far above that of traditional data centers, raising pressure around peak load, grid connection, and electricity-cost optimization; energy storage and smart grids are becoming key infrastructure supporting AI expansion.

• Policy and market-based mechanismsESS return-improvement framework

  Electricity-price liberalization and capacity fees are pushing storage projects from policy-driven to value-driven economics.

  After allowing intraday peak-valley electricity prices and introducing ESS capacity fees, independent storage project IRRs may improve from 4-5% to above 10%, and prime-location projects may even reach 20%.

• Energy securityAPAC nuclear re-evaluation framework

  Energy security, geopolitics, and supply-chain capability together influence nuclear investment decisions.

  Middle East conflicts have increased attention on energy resilience, and some APAC countries are re-evaluating nuclear power because of limited oil and gas resources; however, nuclear decisions are still shaped by geopolitics, safety agreements, long-term fuel supply, and technology pathways.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• ESS

  A core beneficiary direction driven by AI data centers, power-price reform, and renewable-energy integration.

  Strengths

  China's installed capacity is growing strongly, the business model is shifting from policy-driven to value-driven, and project IRRs still have room to improve.

  Weaknesses

  Some utility-scale ESS orders have long lead times from order to delivery, and cost increases take time to pass through.

  Comparison

  Compared with traditional data centers, AIDC's higher power density makes storage demand more inelastic.

  Risks

  Policy changes, returns falling short of expectations, delayed cost pass-through, and grid-connection constraints.

• AIDC

  An important growth engine for new power demand and ESS demand.

  Strengths

  High power density creates demand for peak shaving, load smoothing, and lower electricity costs, and related ESS TAM may grow 15-20x by 2030.

  Weaknesses

  Current adoption is still early, the base is low, and deployment pace may be uneven.

  Comparison

  AIDC rack power density is 60-200+kW+, far above the 5-8kW of traditional data centers.

  Risks

  Data center buildout pace, grid-connection approvals, electricity-price policy, and capex-cycle volatility.

• Power Grid Infrastructure

  An infrastructure-investment direction absorbing pressure from AIDCs, electrification, and renewable-energy integration.

  Strengths

  Global aging-grid upgrade demand is huge; China is the largest single market, and Southeast Asia also has significant investment space.

  Weaknesses

  Project cycles are long and are affected by regulation, approvals, and the pace of grid investment.

  Comparison

  Smart grids, demand-side management, and renewable-energy absorption capacity are key themes.

  Risks

  Capex delays, policy approvals, supply-chain bottlenecks, and uncertain return mechanisms.

• Nuclear Power

  Has regained attention in APAC amid energy security and power-resilience concerns.

  Strengths

  Can provide stable baseload power; South Korea has international certification and project-delivery advantages, while China has cost and supply-chain scale advantages.

  Weaknesses

  Large-scale reactors are currently more favored than SMRs, and technology and regulatory barriers are high.

  Comparison

  Large-scale reactors currently have the edge over SMRs because of their higher technology maturity.

  Risks

  Geopolitics, safety agreements, long-term fuel supply, technology-path selection, and public acceptance.

• China Renewables and Power Equipment OW Ideas

  The key stock directions selected after overlaying macro themes with industry-cycle considerations.

  Strengths

  Covers inverters, submarine cables, wind power, charging/power equipment, and metering, benefiting from ESS, grid upgrades, and renewable-energy investment.

  Weaknesses

  Different companies are affected differently by prices, orders, policy, and industry competition.

  Comparison

  The report lists Sungrow-A, Orient Cables-A, Deye-A, Goldwind-H, Qingdao TGOOD Electric, and Wasion Holdings Ltd-H as the main OW ideas.

  Risks

  Valuation volatility, policy coming in below expectations, slow industry de-capacity consolidation, and overseas demand and FX risks.

Key data

• China ESS cumulative installed capacity144GW by 2025; 450GW by 2030The report says this implies about 25% CAGR.
• Improvement in independent ESS project IRR4-5% to above 10%; prime locations can reach 20%The improvement comes from policies such as electricity-market liberalization and ESS capacity fees.
• AIDC rack power density60-200+kW+ per rackSignificantly higher than the 5-8kW per rack typical of traditional data centers.
• AIDC-related ESS TAM growth15-20x into 2030The report believes this scenario is still at a low base.
• Power grid upgrade demandTrillions of dollars globally; multibillion-dollar scale in Southeast AsiaDriven by AIDCs, electrification, renewable-energy integration, and aging grids.
• Key OW stocksDeye-A, Goldwind-H, Orient Cables-A, Qingdao TGOOD Electric, Sungrow-A, Wasion Holdings Ltd-HThe report disclosed related prices as of May 25, 2026 or May 22, 2026.

Impact & implications

If AI data center construction, power-market liberalization, and renewable-energy integration continue to advance, storage, grid equipment, smart grids, offshore wind-related equipment, and nuclear supply chains with cost advantages may attract greater attention. In investment terms, the report leans toward companies in equipment and utility chains that can benefit from ESS capacity expansion, AIDC power demand, grid upgrades, and China/Southeast Asia energy-infrastructure investment.

Risks

• Improved independent ESS returns depend on electricity-price mechanisms, capacity fees, and project location, so actual IRRs may fall short of the meeting scenarios.
• If AIDC deployment and grid-connection pace are slower than expected, demand for related storage and grid investments will be affected.
• Rising utility-scale ESS costs may be difficult to pass through in time because of long order-to-delivery cycles.
• Power-market liberalization may pressure the earnings of some wind-farm operators.
• Nuclear development is affected by geopolitics, safety agreements, fuel-supply relations, and technology-path choices, leaving high decision uncertainty.
• It will still take time for anti-involution policies in segments such as polysilicon to drive industry consolidation.

What to watch

• Implementation of China's ESS capacity fees, electricity-price liberalization, and peak-valley spread policies.
• The pace of AIDC buildout in China, developed markets, and Southeast Asia, along with rack power density and grid-connection demand.
• Capex for grid upgrades, smart-grid tenders, and demand-side management projects in China and Southeast Asia.
• Policy clarity for China's offshore wind sector and progress in approvals for new projects.
• APAC nuclear project approvals, technology-path selection, and the competitive landscape between Korean and Chinese supply chains.
• Order trends, gross margins, and overseas-demand changes for the key OW companies Sungrow-A, Orient Cables-A, Deye-A, Goldwind-H, Qingdao TGOOD Electric, and Wasion Holdings Ltd-H.