Strategic Convergence: "Compute-Power Synergy" Elevates National Infrastructure Agenda

A Deep Dive into the Resonance of Green Power, Grid Equipment, and IDCs in the AI Energy Era

Date: March 6, 2026
Sector: Power Equipment & New Energy Infrastructure
Rating: OVERWEIGHT / RECOMMEND (Maintained)
Analysts: Fu Honghao (S1050521120004), Zang Tianlv (S1050522120001)
Source: Huaxin Securities Research Institute

Executive Summary

The Chinese power equipment sector is undergoing a structural paradigm shift, driven by the formal elevation of "Compute-Power Synergy" (算电协同) to a national strategic priority. On March 5, 2026, the Government Work Report explicitly mandated the implementation of "ultra-large-scale intelligent computing clusters" and "compute-power synergy" as part of new infrastructure engineering. This policy milestone marks the transition of energy-compute integration from an industrial optimization tactic to a core pillar of national digital and energy security strategy.

In the context of the explosive growth of Artificial Intelligence (AI) large models and the digital economy, the surging demand for computational power is creating an unprecedented strain on energy resources. Our analysis indicates that energy supply has become the primary bottleneck constraining the scalability of the AI industry. Consequently, the investment logic for the power equipment sector is no longer solely defined by traditional renewable energy installation rates but is now deeply intertwined with the capital expenditure cycles of data centers (IDCs) and the modernization of the electrical grid to support high-density loads.

We identify three converging investment主线 (main lines) that will define the sector’s alpha generation potential over the next 3-5 years:
1. Green Power & New Energy Operations: Benefiting from the spatial matching of compute loads with renewable-rich regions (e.g., "East Data, West Computing") and the rising premium for green electricity trading.
2. Grid & Power Equipment Upgrades: Driven by the need for high-voltage direct current (HVDC) transmission, solid-state transformers (SST), and robust distribution networks to handle the volatility and intensity of AI-driven loads.
3. IDC & Compute Infrastructure: Entering a new capex super-cycle characterized by high-power density racks, liquid cooling adoption, and integrated power management systems.

While the photovoltaic (PV) supply chain continues to navigate a period of price consolidation and inventory digestion, the overarching narrative of "AI-Energy Convergence" provides a robust defensive and offensive thesis for the broader power equipment sector. We maintain a "Recommend" (Overweight) rating on the sector. Key beneficiaries include Jinpan Technology and Sifang Shares in grid transformation; Kehua Data in IDC infrastructure; and Sungrow Power Supply in energy storage solutions.

Key Takeaways

1. Policy Catalyst: "Compute-Power Synergy" Enters the National Lexicon

For the first time, the term "Compute-Power Synergy" was included in the Government Work Report on March 5, 2026. The report calls for:
* Implementation of ultra-large-scale intelligent computing clusters.
* Strengthening national integrated monitoring and scheduling of computing power.
* Supporting the development of public cloud infrastructure.

This signifies a top-down directive to align the expansion of digital infrastructure (compute) with the constraints and opportunities of the energy system (power). The policy framework aims to resolve the temporal and spatial mismatch between energy generation (often intermittent and location-specific) and energy consumption (increasingly concentrated in data hubs). By leveraging mechanisms such as green power trading, source-grid-load-storage integration, and joint scheduling of compute and power, the state aims to enhance energy utilization efficiency while alleviating grid pressure.

2. The AI Energy Crunch: Quantifying the Demand Shock

The proliferation of Generative AI and Large Language Models (LLMs) has fundamentally altered the power profile of data centers. Unlike traditional web-serving workloads, AI training and inference require sustained, high-intensity computational power, leading to a exponential rise in Energy Usage Effectiveness (EUE) demands.

• Global Load Growth: According to Vertiv, the total IT load of newly added intelligent computing centers globally is projected to exceed 100 GW between 2023 and 2028, representing a Compound Annual Growth Rate (CAGR) of approximately 9.5%.
• Consumption Surge: IDC data reveals that IT energy consumption for AI data centers (including servers, storage, and networking) reached 55.1 TWh in 2024. This figure is expected to jump to 77.7 TWh in 2025—doubling the 2023 baseline—and further escalate to 146.2 TWh by 2027.
• CAGR Implication: The five-year CAGR (2022-2027) for AI data center energy consumption stands at 44.8%, implying a six-fold increase in energy demand within a half-decade.

This trajectory positions AI servers as the single largest contributor to new electricity demand growth. The inability of legacy grid infrastructure to accommodate this rapid, localized surge in load without significant upgrades creates a compelling investment case for power equipment manufacturers.

3. Industrial Logic: Three Pillars of "Compute-Power Synergy"

The strategic alignment of compute and power manifests through three distinct industrial pathways, each offering specific investment opportunities:

4. Sector Performance & Market Sentiment

Despite broader market volatility, the Power Equipment sector demonstrated resilience in the week ending March 6, 2026:
* Relative Outperformance: The Shenwan Power Equipment Index rose by 0.55%, outperforming the Shanghai Composite Index by 1.48 percentage points and the CSI 300 Index by 1.62 percentage points. It ranked 4th among all first-level sub-industries.
* Stock Movers: Significant gains were observed in companies exposed to grid automation and smart metering, such as Shunna Shares (+27.40%) and Jianan Intelligent (+25.68%), reflecting investor appetite for grid-side upgrades. Conversely, PV material suppliers like Autowell (-19.37%) faced pressure due to ongoing price wars in the solar supply chain.

5. Photovoltaic Supply Chain: Stabilization Amidst Weakness

While the long-term outlook for renewables remains positive under the "Compute-Power" framework, the immediate PV supply chain is experiencing a corrective phase:
* Polysilicon: Prices have softened, with transaction expectations dropping below RMB 50/kg for dense material, and buyers pushing towards RMB 40/kg. Inventory pressure remains high.
* Wafers: Prices declined across major sizes (183N, 210RN). The average price for 183N wafers settled at RMB 1.05/piece. The anticipated pre-tax-refund stocking rush in April has had a limited impact on price stabilization.
* Cells & Modules: N-type cell prices dropped to ~RMB 0.42/W. Module prices saw a slight nominal increase (TOPCon distributed avg. +RMB 0.03 to RMB 0.79/W), but this is attributed to rising silver paste costs rather than robust demand. Margins remain compressed.
* Outlook: The sector is in a "bottoming-out" phase. While prices are stabilizing, a true recovery in profitability awaits demand acceleration or supply-side consolidation.

6. Energy Storage: Price Recovery and Volume Growth

The energy storage segment shows signs of healthy demand and pricing power recovery:
* Volume: In February 2026, domestic storage orders totaled 52.7 GWh. Grid-side projects in Ningxia, Gansu, and Xinjiang led the procurement activity.
* Pricing: Average bidding prices for integrated storage systems (excluding rate differentiation) rebounded to above RMB 0.50/Wh. Major tenders from CGN, Beijing Jianglai Energy, and Huadian reported averages of RMB 0.503/Wh, RMB 0.523/Wh, and RMB 0.54/Wh respectively, representing a 10-18% increase compared to December 2025 levels.
* Implication: The price stabilization suggests that the brutal price war of 2024-2025 may be easing, allowing leading manufacturers with cost advantages to restore margins.

Detailed Industry Analysis

I. The Macro-Policy Framework: From "East Data, West Computing" to "Compute-Power Synergy"

The evolution of China’s digital infrastructure policy has moved from geographical redistribution to systemic integration. The initial "East Data, West Computing" (东数西算) project focused on relocating data centers to western provinces with abundant renewable energy. However, this created challenges in latency for real-time applications and required massive transmission infrastructure investments.

The new "Compute-Power Synergy" strategy addresses these limitations by emphasizing dynamic coupling:
1. Temporal Matching: Using energy storage and flexible load management in data centers to absorb excess renewable generation during peak production hours (e.g., midday solar).
2. Spatial Optimization: Continuing the trend of locating hyperscale training clusters in renewable-rich zones while keeping inference nodes closer to user bases, supported by a strengthened ultra-high-voltage (UHV) grid.
3. Market Mechanisms: Enhancing green power trading platforms to allow data center operators to procure certified renewable energy, thereby meeting ESG commitments and potentially lowering long-term energy costs.

This policy shift effectively de-risks the renewable energy sector by providing a large, growing, and relatively stable base load customer (data centers), while simultaneously ensuring that the AI industry has access to the vast amounts of power it requires without exacerbating carbon emissions.

II. Global Context: The US Tech Giants’ Power Dilemma

The Chinese strategy mirrors a global trend. In the United States, the energy constraints of AI are becoming a critical boardroom issue for Big Tech.
* White House Commitment: On March 5, 2026, representatives from Microsoft, Google, OpenAI, Amazon, Meta, xAI, and Oracle signed a commitment to self-powered data center initiatives. This move was endorsed by President Trump, aiming to alleviate public concern over rising electricity bills and grid strain.
* Infrastructure Bottlenecks: The International Energy Agency (IEA) forecasts global data center electricity demand will more than double to 945 TWh by 2030. In the US, this expansion is currently hindered by turbine shortages, slow grid interconnection queues, and transformer supply chain constraints.
* Investment Parallel: Just as US tech giants are investing directly in power generation (nuclear, gas, renewables) and grid infrastructure, Chinese counterparts are likely to follow suit, driving demand for domestic power equipment manufacturers who can deliver integrated solutions.

III. AI Industry Developments: Fueling the Power Demand

The underlying driver of this energy transition is the relentless expansion of AI capabilities, evidenced by recent mega-deals:
* Meta-AMD Partnership: Meta signed a $100 billion+ multi-year agreement with AMD to secure 6 GW of AI computing capacity. This deal highlights the shift towards diversified GPU sourcing (beyond Nvidia) and underscores the sheer scale of power required (6 GW is equivalent to the output of several large nuclear reactors). The inclusion of equity warrants for AMD further cements the strategic alliance.
* OpenAI Funding Round: OpenAI raised $110 billion in a new financing round led by SoftBank, Nvidia, and Amazon, achieving a pre-money valuation of $730 billion. A significant portion of this capital is earmarked for infrastructure expansion, including next-generation inference compute capabilities in partnership with Nvidia and Amazon Web Services (AWS).

These capital injections confirm that the AI capex cycle is far from peaking. Instead, we are in the early stages of a decade-long infrastructure build-out, where every dollar spent on chips necessitates multiple dollars spent on power and cooling infrastructure.

IV. Photovoltaic Supply Chain Tracking: A Sector in Transition

While the long-term demand for solar is secured by the AI-energy nexus, the short-term dynamics of the PV supply chain remain challenging. The following analysis details the price movements and market sentiment as of early March 2026.

1. Polysilicon: Downward Pressure Persists
* Market Status: No new large-scale contracts were executed this week; transactions were limited to the fulfillment of pre-holiday orders. Negotiations for March deliveries are underway, but buyer sentiment is weak.
* Price Trends: Sellers, facing inventory accumulation, are accepting lower prices. Transaction prices for dense polysilicon have fallen to the RMB 44-53/kg range, with granular silicon at RMB 42-45/kg. Buyers are aggressively targeting the RMB 40/kg level.
* International Markets: Overseas polysilicon prices remain around $18/kg, with Malaysian production stable. Oman-based production is still in negotiation phases. US domestic prices remain elevated due to trade protections.
* Outlook: With production schedules for March expected to rise, and cost support weakening, polysilicon prices are likely to remain under pressure until inventory levels normalize.

2. Wafers: Price Corrections Accelerate
* Market Status: Wafer prices have seen a marked decline, with manufacturers adjusting quotes frequently (every 1-2 days).
* Price Trends:
* 183N Wafers: Mainstream price dropped to RMB 1.05/piece. Previous transactions at RMB 1.10 have ceased.
* 210RN Wafers: Trading range narrowed to RMB 1.13-1.18/piece, with RMB 1.15 emerging as the mainstream volume price.
* 210N Wafers: Held steady at RMB 1.40/piece, but downward pressure is building as negotiations continue.
* Drivers: The anticipated surge in demand ahead of the April export tax rebate deadline has been muted. Increased wafer output in March is helping to digest battery production schedules but has not created a supply shortage. The weakening polysilicon cost base further removes support for wafer prices.

3. Cells: Margins Compressed
* Price Trends:
* N-Type (183N/210N): Average price fell to RMB 0.42/W.
* 210RN: Average price declined to RMB 0.43/W.
* Demand Dynamics: Demand for 183N cells, previously supported by overseas orders, has slowed. 210N cells face weak domestic utility-scale demand. 210RN commands a slight premium due to its prevalence in module designs for distributed and terminal applications.
* Cost Pass-Through: As silicon and silver paste prices fall, downstream module makers are demanding lower cell prices. Despite a sequential increase in March production schedules, the lack of robust demand means cell prices may decline further.
* Export Markets:
* P-Type: 182P cells stable at $0.05/W (global avg) and $0.085/W (SE Asia to US).
* N-Type: 183N export avg dropped to $0.057/W. SE Asia-produced cells using non-Chinese silicon for the US market remain at $0.115/W.
* Trade Barriers: The imposition of anti-subsidy duties on Indonesian and Laotian cells has effectively cut off these routes to the US market, redirecting supply to India and leaving only minimal legacy orders for the US.

4. Modules: Cost-Driven Price Adjustments
* Domestic Prices:
* TOPCon Distributed: Avg price rose by RMB 0.03 to RMB 0.79/W.
* TOPCon Utility: Avg price rose by RMB 0.01 to RMB 0.70/W.
* BC (Back Contact) Distributed: Avg price adjusted to RMB 0.86/W.
* BC Utility: Avg price adjusted to RMB 0.81/W.
* Reality Check: While quoted prices for TOPCon modules are in the RMB 0.85-0.90/W range, actual transaction prices for utility projects are RMB 0.68-0.70/W, and distributed projects are RMB 0.76-0.83/W.
* Overseas Prices: Average TOPCon module price is $0.096/W. Due to export tax rebate changes, overseas orders are being renegotiated, with distribution and project quotes rising to $0.10-0.13/W.
* Analysis: The recent price uptick is primarily cost-push (driven by higher silver paste and auxiliary material costs) rather than demand-pull. Terminal demand remains tepid, and the lack of substantial volume growth means that manufacturer margins are not significantly improving. The market is in a state of cautious observation, with procurement delayed pending clearer demand signals.

5. Auxiliary Materials: Mixed Signals
* EVA Particles: Prices remained flat this week, but expectations for next week are bullish due to rising crude oil prices and potential supply disruptions from Middle East tensions. Petrochemical producers face rising costs and potential production cuts.
* Backsheet PET: Prices rose 2.2% due to crude oil volatility and concerns over the Strait of Hormuz, which pushed up PTA costs.
* Aluminum Frames: Prices rose 1.9%. Supply is stable, but social inventory is rising. Downstream demand is cautious, though some "buy-on-rise" sentiment exists.
* Copper (Cables): Prices rose 1.4%. High inventory levels are offset by tight spot liquidity and holder resistance to price drops.
* Steel (Mounts): Hot-rolled coil prices fell 0.2% due to high inventory and slow demand recovery.
* Photovoltaic Glass: Prices fell 2.9%-4.8%. Supply remains stable, and while some restocking is occurring, inventory levels remain high. Prices are expected to stabilize in the near term.

V. Energy Storage Market: Signs of Stabilization and Growth

The energy storage sector is exhibiting stronger fundamentals than the pure-play PV manufacturing sector, driven by mandatory storage policies for renewable projects and the increasing need for grid flexibility.

1. Order Volume Analysis (February 2026)
* Total Orders: 52.7 GWh completed domestically.
* Breakdown:
* Storage Systems & EPC (with equipment): 13.4 GW / 36.9 GWh.
* DC Side Orders: 8.8 GWh.
* Battery Cell Procurement: 7 GWh.
* Regional Leaders: Ningxia, Gansu, and Xinjiang topped the procurement charts. Ningxia, in particular, showed high activity, releasing a list of 40.6 GWh grid-side projects and finalizing 3.9 GWh in orders.

2. Pricing Trends: The Rebound
After a prolonged period of price erosion, storage system bids in February showed a notable recovery.
* Major Tenders:
* CGN (7.2 GWh): Avg price RMB 0.503/Wh.
* Beijing Jianglai Energy (6 GWh): Avg price RMB 0.523/Wh.
* Huadian (12 GWh): Avg price RMB 0.54/Wh.
* Comparison: These averages represent a 10-18% increase compared to the December 2025 Huaneng Clean Energy Research Institute tender (which averaged ~RMB 0.45-0.48/Wh depending on configuration).
* Cell Prices: SPIC’s 7 GWh cell tender saw bids ranging from RMB 0.325-0.384/Wh, with an average of RMB 0.344/Wh, a 10.3% increase from the December 2025 average of RMB 0.312/Wh.

3. System & EPC Pricing by Duration
* 2-Hour Systems: Avg bid RMB 0.579/Wh (Range: 0.488-0.670), down 1.9% MoM.
* 4-Hour Systems: Avg bid RMB 0.538/Wh (Range: 0.452-0.602), down 9.1% MoM.
* 2-Hour EPC: Avg bid RMB 1.088/Wh, up 4.6% MoM.
* 4-Hour EPC: Avg bid RMB 0.951/Wh, up 5.2% MoM.

Interpretation: The divergence between system prices (slightly down or mixed) and EPC prices (up) suggests that while battery cell costs may have stabilized or slightly increased, the value-added services, integration complexity, and balance-of-system components are commanding higher premiums. The overall rise in average tender prices indicates that the "race to the bottom" is pausing, allowing healthier margins for tier-1 integrators.

VI. Emerging Themes: Robotics and Circular Economy

1. Robot Recycling Regulations
Effective March 1, 2026, the revised "Technical Specifications for Pollution Control in the Treatment of Waste Electrical and Electronic Products" includes new categories such as smart robots, service terminals, and wearable devices. This regulatory framework establishes standards for storage, dismantling, and material recovery, creating a nascent but structured market for e-waste management in the robotics sector. This is particularly relevant as the humanoid robot market scales up.

2. Humanoid Robot Leasing Market
The commercialization of humanoid robots is accelerating via the leasing model. During the 2026 Spring Festival, rental orders surged 70% week-over-week.
* Pricing: Daily rental rates have dropped from nearly RMB 10,000 last year to RMB 1,000-3,000 today.
* Applications: Entertainment (34%), Commercial Marketing (31%), Education/Tourism (19%), and Lifestyle/Emotional Consumption (16%).
* Implication: The drastic reduction in rental costs lowers the barrier to entry for commercial users, accelerating data collection and real-world testing, which in turn drives iterative improvements in hardware and software. This trend indirectly supports the power equipment sector through increased demand for charging infrastructure and high-performance computing for robot fleet management.

Investment Strategy & Stock Recommendations

Given the macro backdrop of "Compute-Power Synergy," we recommend a barbell strategy: overweighting companies with direct exposure to AI-driven grid upgrades and IDC infrastructure, while maintaining selective exposure to leading renewable energy operators who benefit from long-term green power demand.

Core Investment Lines

1. IDC & Power Infrastructure (High Conviction)

   • Logic: Direct beneficiaries of the AIDC construction boom. Focus on companies providing HVDC power supplies, uninterruptible power supplies (UPS), and integrated data center solutions.
   • Top Pick: Kehua Data (002335.SZ). As domestic AIDC tenders ramp up, Kehua’s integrated expertise in UPS, PV inverters, and IDC operations positions it well. Note: Unrated in our table, but closely monitored.

2. Grid Transformation & HVDC/SST (High Growth)

   • Logic: The integration of high-power AI loads requires advanced grid technologies. High-Voltage Direct Current (HVDC) and Solid-State Transformers (SST) are critical for managing power quality and efficiency. These segments have high barriers to entry and expanding total addressable markets (TAM).
   • Top Picks:
     • Jinpan Technology (688676.SH): Leader in dry-type transformers and digitalized grid solutions. Strong exposure to overseas markets and domestic grid upgrades. Rating: BUY.
     • Sifang Shares (601126.SH): Core provider of grid protection automation and control systems. Essential for the digitalization and stability of the new power system. Rating: BUY.

3. Critical Components & Power Electronics

   • Logic: Specific components within the power distribution chain are seeing value uplift.
   • Top Picks:
     • Liangxin Shares (002706.SZ): High-end low-voltage electrical appliances. Its high-voltage circuit breakers are core to distribution transformation. Rating: BUY.
     • Megmeet (002851.SZ): Leading player in power electronics for AI servers and data centers. Rating: BUY.
     • Zhongheng Electric (002364.SZ) & Oulu Tong (300870.SZ): Key suppliers of power modules and server power supplies for AIDC. Note: Unrated/Monitor.

4. Energy Storage & Thermal Management

   • Logic: Storage is essential for smoothing renewable input to data centers. Thermal management is critical for high-density AI chips.
   • Top Picks:
     • Sungrow Power Supply (300274.SZ): Global leader in PV inverters and energy storage systems. Benefits from both the green power trend and the storage price stabilization. Rating: BUY.
     • Tongfei Shares (300990.SZ): Specialized in precision temperature control (liquid cooling) for data centers and energy storage. Direct beneficiary of the shift to liquid-cooled AI racks. Rating: BUY.

Valuation & Financial Forecasts

The following table summarizes our earnings estimates and valuations for key covered companies. Note that for unrated companies, consensus estimates from Wind are used.

Source: Wind, Huaxin Securities Research. Note: PE ratios are based on closing prices as of March 6, 2026. EPS forecasts for unrated companies are Wind consensus estimates.

Valuation Commentary:
* Jinpan Technology and Sifang Shares trade at forward PEs of ~46x and ~52x for 2025, respectively. While these multiples appear elevated compared to traditional utilities, they are justified by the high growth visibility associated with grid modernization and AI-related capex. The PEG ratio remains attractive given the expected earnings CAGR.
* Sungrow Power Supply offers a more value-oriented entry point with a 2025E PE of ~20.75x, reflecting its mature market position and strong cash flow generation, making it a stable core holding.
* Tongfei Shares commands a premium valuation due to its niche leadership in liquid cooling, a technology critical for next-gen AI chips. The rapid EPS growth (from 0.91 to 2.65 between 2024-2026) supports the multiple.

Risks / Headwinds

Investors should be aware of the following risks that could impact the sector’s performance:

1. Technological Development Risks:

   • AI Efficiency Gains: If AI chip efficiency improves faster than expected (e.g., through breakthroughs in neuromorphic computing or extreme low-power architectures), the projected growth in energy demand could be overstated, reducing the urgency for grid upgrades.
   • Storage Technology Disruption: Rapid advancements in alternative storage technologies (e.g., solid-state batteries, flow batteries) could disrupt the current lithium-ion dominated supply chain, affecting incumbent manufacturers.

2. Competitive Intensification:

   • PV Supply Chain: The photovoltaic sector remains highly fragmented. Continued price wars and overcapacity could lead to prolonged margin compression, even for leading firms.
   • Grid Equipment: As the market for AI-related power equipment heats up, new entrants may erode market share and pricing power for established players like Jinpan and Sifang.

3. Macroeconomic & Systemic Risks:

   • Interest Rates: Higher-for-longer interest rates could increase the cost of capital for large-scale infrastructure projects (IDCs, grid upgrades), potentially delaying capex decisions.
   • Market Volatility: Broad market corrections or systemic financial risks could lead to de-rating of growth stocks in the power equipment sector, regardless of fundamental performance.

4. Company-Specific Execution Risks:

   • Order Delays: Government infrastructure projects often face bureaucratic delays. Slower-than-expected rollout of "Compute-Power" pilots could impact near-term revenue recognition.
   • Earnings Misses: Recommended companies may fail to meet the aggressive earnings forecasts embedded in their current valuations, particularly if raw material costs (copper, silver) rise unexpectedly.

5. Policy & Regulatory Risks:

   • Trade Barriers: Escalating trade tensions (e.g., tariffs on Chinese power equipment or EVs/batteries in the US/EU) could limit export opportunities for companies like Sungrow and Jinpan.
   • Electricity Pricing: Changes in regulated electricity prices or green power certification rules could alter the economic viability of certain "Compute-Power" business models.

Rating / Sector Outlook

Sector Rating: RECOMMEND (OVERWEIGHT)

We maintain a positive outlook on the Power Equipment sector, driven by the structural tailwinds of AI-induced energy demand and national strategic support for "Compute-Power Synergy." While the traditional renewable installation cycle is maturing, the quality of demand is shifting towards high-value, technology-intensive segments such as grid automation, high-voltage power electronics, and integrated data center energy solutions.

Short-Term View (1-3 Months):
Expect continued volatility in PV manufacturing stocks due to ongoing price adjustments. However, grid equipment and IDC infrastructure stocks should outperform as investors rotate into themes with clearer near-term catalysts (government tenders, AI capex announcements).

Medium-to-Long Term View (1-3 Years):
The convergence of AI and Energy is a multi-year secular trend. Companies that successfully position themselves at the intersection of digitalization and electrification will command premium valuations. We anticipate a re-rating of the sector as earnings from AI-related power projects begin to materialize on income statements in 2026-2027.

Investment View

The narrative of "Compute-Power Synergy" is not merely a policy slogan; it is a fundamental restructuring of the relationship between the digital and physical economies. For institutional investors, this presents a unique opportunity to invest in the "picks and shovels" of the AI revolution—not just the chips, but the energy infrastructure that powers them.

Strategic Allocation Recommendations:

1. Overweight Grid Modernization: Prioritize companies with strong moats in high-voltage equipment and grid automation (Jinpan Technology, Sifang Shares). These firms benefit from both domestic grid upgrades and potential international expansion as global grids struggle with similar AI-driven loads.
2. Selective IDC Exposure: Focus on integrated solution providers rather than pure real estate plays. Kehua Data and Megmeet offer exposure to the high-margin power supply and management layers of the IDC stack.
3. Thermal & Storage Plays: As power densities rise, cooling becomes as critical as power. Tongfei Shares is a pure play on this trend. Meanwhile, Sungrow offers a balanced exposure to the necessary storage buffer required for renewable-powered compute.
4. Caution on Pure PV Manufacturing: While long-term demand is secure, the near-term margin outlook for polysilicon, wafer, and cell manufacturers remains challenging. Prefer vertically integrated leaders or those with significant overseas exposure where pricing power is stronger.

Conclusion:
The era of "AI Energy Consumption" has arrived. The stock market is beginning to price in the reality that power is the new currency of AI. By aligning portfolios with the "Compute-Power Synergy" strategy, investors can capture value from both the digital explosion and the necessary energy transition. We advise accumulating positions in high-quality grid and IDC infrastructure names on any market dips, as the fundamental demand drivers are only strengthening.

Appendix: Detailed Data Tables & Charts Reference

(Note: The following data points are derived from the original research report and are included for reference.)

Chart 1: Key Company Valuations
(See Table in Investment Strategy Section)

Chart 2: Photovoltaic Supply Chain Prices (Week of Mar 2-6, 2026)

Chart 3: Energy Storage Tender Results (February 2026)

Chart 4: Sector Performance (Week Ending Mar 6, 2026)

Top Gainers in Power Equipment:
1. Shunna Shares: +27.40%
2. Jianan Intelligent: +25.68%
3. Ankao Zhidian: +24.24%
4. Zhonglai Shares: +20.26%
5. China XD Electric: +20.02%

Top Losers in Power Equipment:
1. Autowell: -19.37%
2. Haiyou New Material: -15.22%
3. Jingsheng Electromechanical: -14.96%
4. *ST Weil: -14.73%
5. Mingguan New Material: -13.98%

Analyst Certification & Disclosures

Analyst Certification:
The analysts named in this report, Fu Honghao and Zang Tianlv, certify that they have the requisite professional qualifications registered with the Securities Association of China. They declare that the views expressed in this report accurately reflect their personal, independent, and objective research opinions. They confirm that they have not received, nor will they receive, any direct or indirect compensation for the specific recommendations or views contained herein.

Investment Rating Definitions:

• Stock Ratings:

  • BUY: Expected return > 20% relative to the benchmark index.
  • ACCUMULATE: Expected return 10%—20% relative to the benchmark index.
  • NEUTRAL: Expected return -10%—10% relative to the benchmark index.
  • SELL: Expected return < -10% relative to the benchmark index.

• Industry Ratings:

  • RECOMMEND (OVERWEIGHT): Expected industry index return > 10% relative to the benchmark index.
  • NEUTRAL: Expected industry index return -10%—10% relative to the benchmark index.
  • AVOID (UNDERWEIGHT): Expected industry index return < -10% relative to the benchmark index.

• Benchmark Indices:

  • A-Shares: CSI 300 Index.
  • New Third Board: NEEQ Index or NEEQ Market-Making Index.
  • Hong Kong: Hang Seng Index.
  • US: Dow Jones Industrial Average.

Disclaimer:
Huaxin Securities Co., Ltd. ("Huaxin Securities") holds the qualification for securities investment consulting business approved by the China Securities Regulatory Commission. This report is produced by Huaxin Securities for the exclusive use of its clients. Receipt of this report does not constitute a client relationship.

The information contained herein is derived from public sources believed to be reliable. Huaxin Securities and its researchers strive for accuracy and completeness but make no warranties, express or implied, regarding the accuracy or completeness of such information. The report is intended to be objective and fair, but the information and opinions expressed do not constitute an offer or solicitation to buy or sell any securities. They do not take into account the specific investment objectives, financial situation, or particular needs of any recipient. Investors should independently evaluate the information and opinions in this report and consult with professional advisors as necessary. Huaxin Securities and its affiliates shall not be liable for any consequences arising from the use of this report.

Huaxin Securities or its affiliates may hold positions in the securities mentioned in this report and may engage in trading activities. They may also provide investment banking, financial advisory, or other financial services to the companies mentioned. Such relationships are disclosed in accordance with applicable laws and regulations.

The data, opinions, and forecasts in this report reflect judgments as of the date of initial publication and are subject to change without notice. Huaxin Securities may issue other reports with inconsistent opinions or forecasts. There is no obligation to update this report for all recipients.

The copyright of this report belongs solely to Huaxin Securities. No institution or individual may reproduce, distribute, or publish any part of this report in any form without written authorization. If other institutions distribute this report to their clients, they are solely responsible for such distribution. Huaxin Securities assumes no liability for such actions. Unauthorized reproduction or forwarding of this report is strictly prohibited, and violators will be held legally accountable. Investors are advised to use only authorized versions of Huaxin Securities research reports.