As the scale of data center construction evolves alongside advancements in technologies such as cloud computing and artificial intelligence, the number of servers and data centers is rising rapidly. Coupled with the imperatives of energy conservation and the need to manage power consumption per unit of computing power, requirements for power supply systems have become increasingly stringent. Consequently, new solutions are being prioritized, imposing more rigorous operational and technical demands on the capacitors used within these power systems.

2025: Global AI server market size of $158.7 billion (IDC data); broad AI infrastructure spending of approximately $319 billion (estimated).

2026: AI server market size projected at $240–$250 billion; broad AI infrastructure spending forecast at $497 billion (a year-over-year increase of nearly 56%).

Growth drivers: Surging capital expenditure by hyperscale cloud providers, explosive demand for inference, and rising average selling prices for high-value GPU servers.

Data centers and server power supplies generate massive demand for electrolytic capacitors and impose strict requirements on their design specifications.

Demand:

Scale in 2025: approximately 64.55 billion yuan (a year-on-year increase of 5.1%). The benchmark data comes from the China Electronic Components Industry Association.

‌2026 forecast‌: approximately ‌67.8 billion-68.5 billion yuan‌ (calculated based on a growth rate of 5.0%-5.2%),

Driven by the demand for new energy and AI servers, the data center and server power supply fields are one of the important factors driving its growth.

Regarding individual server power supply units, the quantity and specifications of electrolytic capacitors vary according to the power rating; for instance, a 550W unit employs a 475V 390μF high-voltage filtering electrolytic capacitor, an 1100W unit uses a 500V 680μF capacitor, and a 2200W unit utilizes multiple 475V 820μF capacitors connected in parallel.

Design parameter requirements:

High capacitance and high voltage rating: Data center server power supplies must maintain stable capacitance values under high-voltage conditions and effectively withstand system surges and voltage fluctuations, thereby ensuring efficient operation and high reliability.

Low Equivalent Series Resistance (ESR): Electrolytic capacitors with low ESR effectively reduce energy loss within the power system and enhance power conversion efficiency. In high-frequency operating environments, low ESR characteristics are critical to the stability and reliability of data center server power supplies.

Regarding charge-discharge durability, the capacitor is required to maintain stable, high-performance operation during high-energy load switching. Under 100 Vpp test conditions, it must withstand frequent load fluctuations associated with server operations without sustaining damage, necessitating long-term testing involving hundreds of millions of charge-discharge cycles.

Long Service Life: Data center servers require continuous, 24/7 operation, necessitating electrolytic capacitors with a long service life. For instance, SANCON’s KP series electrolytic capacitors offer a lifespan of 3,000 hours at 105°C, while the higher-performance BA series boasts an even longer lifespan of 5,000 hours at 125°C, effectively meeting the requirements for the long-term, stable operation of data center servers.

Compact Size and High Power Density: Advancements in server power supply technology have driven a demand for greater miniaturization and higher power density in electrolytic capacitors. For instance, SANCON’s BP series 500V 390μF electrolytic capacitors feature a compact footprint and high power density, enabling higher capacitance values within limited space.

Excellent temperature characteristics: The operating temperature range of the power supply environment significantly impacts the performance of electrolytic capacitors; therefore, the capacitor's rated operating temperature range must encompass the power supply's actual operating environment. SANCON capacitors ensure stable, long-term operation across a wide temperature range—from a minimum of -60°C (without hindering system startup or operation) to a maximum of +125°C—guaranteeing reliable performance under varying temperature conditions.

The figure above illustrates the low-temperature capacity characteristics of Sanxin’s ultra-low-temperature products; at peak performance, capacity degradation remains under 25% at -60°C. These products impose fewer constraints on operating conditions, adapting to a wider temperature range and outdoor environments, and are better suited for low-temperature startup without the need for additional heating or thermal insulation.

The state-of-the-art, large-scale capacitor manufacturing workshop is immaculate, and equipment is regularly upgraded to keep pace with customer demands, boost efficiency, enhance production speed, and improve the reliability and consistency of the manufacturing process—ultimately delivering greater value to customers.

A precision winding process enhances product consistency, enabling the fulfillment of increasingly stringent customer requirements and ensuring the reliability and uniformity of the final product.

The fully computer-controlled intelligent burn-in oven precisely records the burn-in process for every single capacitor, enabling the early detection of latent defects and the screening out of flawed units, thereby further enhancing product stability and reliability.

Comprehensive 5S lean management and end-to-end process control have elevated product quality to a new level, while a commitment to excellence ensures the highest standards in manufacturing processes.

We treat every single capacitor with meticulous care, ready to meet our customers' new challenges with excellence. We produce the finest capacitors to satisfy even the most demanding requirements and product specifications.