虎嗅

After losing tens of billions, where have all the former kings of the new energy industry gone?

原文:亏掉几百亿之后,新能源“旧王”都去哪了

Summary in Plain Language

Over the past few years, the new energy sectors of photovoltaics (PV), energy storage, and charging stations have become highly competitive: the total production in the industry has more than doubled the global demand, and prices are sold below cost. The top five companies have lost a combined amount of 18 billion yuan in just half a year. Even in the power generation sector, there have been instances where electricity prices fall below 0.1 yuan per kilowatt-hour, resulting in losses. The traditional market's ability to generate revenue is almost non-existent. Meanwhile, the AI industry has experienced a surge, and intelligent computing centers equipped with thousands of expensive GPUs are in dire need of electricity. The expansion of the power grid cannot keep up with the energy demands of these advanced systems. This convergence has created a new, billion-dollar opportunity: PV and energy storage companies can repurpose their decades of experience in inverter technology and green energy solutions to provide uninterrupted power to AI centers. The capital market has already boosted the valuations of related companies, but only those that can secure orders, comply with regulations, and deliver reliably will turn this trend into a sustainable business.

---

Detailed Analysis

1. How bad is the new energy industry really?

The current situation in the PV industry goes beyond regular cyclical fluctuations; it's a result of excessive capacity expansion. The industry can produce 1100 GW of PV modules annually, but the global demand is only 600 GW, leading to a supply-to-demand ratio of over 2:1. This forces companies to cut prices, pushing module prices below the cost of raw materials, with losses of up to 0.1 yuan per watt sold. The industry's operational rate is only 30%-40%, and a quarter of the companies have been in the red for three consecutive years. The price war has spread from manufacturing to power generation: silicon material prices have dropped from over 50 yuan per kilogram to 32 yuan, and the electricity generated by many PV bases in the northwest is sold for less than 0.1 yuan per kilowatt-hour. Some power companies even have to pay to use the electricity they produce. There is no room for growth in the traditional market, so the industry must find a new niche that can absorb excess capacity and generate higher profits.

2. Why can PV and energy storage companies enter the AI power supply market?

Many think entering the AI market is just riding on the AI trend, but the underlying technologies are similar. For example, companies that have been in the trucking industry for decades can easily transition to AI power supply without retraining. The core requirement for AI centers is uninterrupted power supply, which includes AC/DC conversion, precise voltage control, and immediate backup power in case of outages—these are precisely the strengths of PV and energy storage companies. NVIDIA's latest AI cabinets use 54V DC power, and the future 800V DC architecture, as well as next-generation solid-state transformers, are all technologies that PV companies have mastered. For instance, Sungrow Power Supply has already started supplying solid-state transformers to data centers, and KSTAR has secured manufacturing orders from major North American cloud providers, immediately turning their technical expertise into revenue.

3. The new market has changed the rules: low prices are no longer the key.

In the past, PV companies competed on price, but now the focus is on reliability. The cost of hardware in an AI center can reach tens of millions or even hundreds of millions of yuan, and even a brief voltage instability can disrupt large-scale model training, resulting in significant losses. Data centers are willing to pay significantly more for power systems with 99.999% availability, reducing downtime to less than 5 minutes per year, which is much more profitable than traditional energy storage solutions. The current supply-demand imbalance overseas favors Chinese companies, as the U.S. power grid expansion is slow (128 weeks for large transformers), forcing many to build their own off-grid microgrids. This aligns with China's rapid PV and energy storage capabilities, making it a popular export opportunity.

4. Opportunities abound, but there are also pitfalls.

While the new market seems accessible, there are hidden barriers. Geopolitical and regulatory compliance requirements, such as increased tariffs on energy storage batteries in the U.S., pose significant challenges. Companies that have established local operations in the Middle East and Southeast Asia have a competitive advantage. The domestic market also requires access to cheap green energy from large PV bases and direct power lines to data centers. The current capital market valuations (30-40x PE) are much higher than those of traditional energy companies (15-22x), but many companies lack the necessary resources to fulfill these requirements, and their valuations will likely fall once the hype subsides.

5. The competition is shifting: those closer to AI have more control over pricing.

The new market focuses on providing stable power to AI centers. Companies that develop inverter and backup power solutions for data centers are in the best position, followed by those that supply green energy. The capital market's high valuations are a temporary advantage; only those that can deliver reliable services will succeed. In the second half of this competition, those closest to AI technology will have the greatest influence over pricing.

In summary, the new energy industry is undergoing a transformation, with the focus shifting from low-cost production to providing reliable, high-quality power solutions for AI. Only companies that can effectively leverage their technology and services will thrive in this new landscape.