According to a research report, the construction of AIDC is experiencing high demand, with gas turbine and gas generator manufacturers seeing full order books. The insufficient supply of traditional power generation equipment is creating opportunities for new technologies. Consequently, the large-scale application of SOFC is expected to accelerate, driving order growth for domestic SOFC manufacturers, as well as component suppliers for fuel cell membranes, metal interconnects, temperature controllers, and raw material providers like metal chromium, presenting investment opportunities.
DFZQ outlines its key viewpoints as follows: SOFC has broad application prospects in AIDC power generation, and related domestic equipment and component manufacturers are poised to benefit.
Solid Oxide Fuel Cells (SOFC) can directly convert the chemical energy of fuel into electricity, meeting the needs of data centers for distributed power generation. They feature rapid deployment, high power generation efficiency, and environmental friendliness. Some market participants believe the high initial investment of SOFC compared to gas turbines makes large-scale adoption in AIDC construction difficult.
However, the report argues that this disadvantage of high initial cost is not critical. As AIDC power shortages intensify, gas turbine and gas generator manufacturers are operating at full capacity. The supply shortage of traditional power equipment is opening a window for new technologies. Therefore, the mass adoption of SOFC is likely to accelerate, benefiting related domestic equipment and component manufacturers.
Initial SOFC Investment Higher Than Gas Turbines, but Limited Impact on Data Center Scale-Up
From an equipment investment perspective, price increases for gas turbines are narrowing the initial investment gap. For instance, GE Vernova indicated in Q1 2026 that prices for new orders compared to backlog orders from Q4 2025 have increased by 10%-20%, with strong price growth anticipated for Q2 2026. From a comprehensive return standpoint, electricity costs constitute an extremely low proportion of data center project expenses.
Industry data suggests that, calculated over a 4-year depreciation period, GPU depreciation accounts for approximately 45% of the cost for a 1GW data center, while electricity costs are only about 5%. Thus, the report contends that given the still attractive returns on data center investments, the core demand from data center owners is rapid deployment, making the high initial investment of SOFC less critical.
Data Center Power Shortages Exacerbate Gas Turbine Supply Constraints, Highlighting SOFC's Rapid Deployment Advantage
The surge in AI computing demand continues to intensify data center power shortages. On July 8, 2026, Meta announced a $10 billion investment to build its first data center in Canada, with a capacity of 1 GW, expected to take 2-3 years to complete. The report views this as partially alleviating prior market concerns about computing power oversupply, indicating robust demand for AI computing resources.
Regarding the latest marginal changes in AIDC power generation equipment, the tight supply situation for gas turbines is worsening. Siemens recently revised its annual demand forecast for the gas turbine industry upwards to approximately 110-120 GW for the coming years during a conference call, reflecting an optimistic outlook for future gas turbine demand.
The report believes that against the backdrop of worsening AIDC power shortages, the core need for data center owners is rapid power deployment. Referencing the case of Bloom Energy deploying SOFC for an Oracle data center, fuel cells can be deployed within six months, whereas traditional power plant construction and grid expansion typically take years. This highlights SOFC's advantage in rapid deployment, which is expected to drive accelerated growth in SOFC installations.
Risk Factors
Potential risks include macroeconomic fluctuations leading to investment shortfalls, slower-than-expected expansion of AI applications, delays in data center construction progress, technological development setbacks, and changes in assumptions affecting calculation results.