[E-Daily Reporter JAEMIN SONG ] The competitive landscape is expanding for the three major domestic battery manufacturers targeting the artificial intelligence (AI) data center market. The data center battery market, which had previously centered on energy storage systems (ESS) designed to store large amounts of electricity, is now diversifying to include power supply equipment located close to servers, such as uninterruptible power supplies (UPS) and battery backup units (BBU). Although the specific areas of focus vary by company, the role that batteries play is expanding as power demand in AI data centers surges.

According to the battery industry on the 30th, the three major battery manufacturers have recently been successively expanding their involvement in AI data center power infrastructure. While LG Energy Solution and SK On are targeting the AI data center market primarily with large-scale ESS systems, SAMSUNG SDI CO.,LTD. has expanded its product lineup to include UPS and BBU systems. This reflects a trend where battery demand—which began with large-scale power storage outside data centers or linked to the power grid—is now expanding into the realm of stable power supply within data centers.

SAMSUNG SDI CO.,LTD. is a prime example of a company that has broadened its scope to include power supply devices located close to servers. At “Data Center World Asia (DCWA) 2026,” held in Singapore on the 29th and 30th, SAMSUNG SDI CO.,LTD. unveiled its 21700 cylindrical lithium iron phosphate (LFP) battery for BBUs for the first time. The product features enhanced output, capacity, and safety thanks to the company’s proprietary structural technology. BBUs are installed in server racks to immediately supply power in the event of a power outage or power abnormality. Along with this, SAMSUNG SDI CO.,LTD. showcased a product lineup covering the entire data center, including UPS systems and the SBB (Samsung Battery Box) 2.0 for large-scale energy storage systems (ESS).

The expanding role of batteries within data centers is driven by the unique power consumption patterns of AI servers. As the number of high-performance graphics processing units (GPUs) installed increases, server power density has risen, and load fluctuations—where power consumption spikes and drops rapidly within short time frames—have become more pronounced. Consequently, there is a growing need to stabilize power at multiple levels—not only through ESS systems that store large amounts of power but also through facility-level UPS systems and server-rack-level BBUs.

LG Energy Solution and SK On are expanding their presence in the AI data center power market, with a focus on large-scale ESS. LG Energy Solution recently joined NVIDIA’s “DSX Ready” partner program for its next-generation AI factory platform, “DSX.” The certified BESS is a modular product with capacities of 2.5 megawatts (MW) and 5.1 megawatt-hours (MWh), designed to mitigate sudden load fluctuations in AI servers and support a stable power supply.

SK On is also developing next-generation products under its own ESS brand, “GRIDON,” targeting data center operators. In collaboration with SK Group affiliates, the company is analyzing data center power demand patterns to develop a multifunctional ESS for AI data centers that combines emergency power supply with load-following capabilities.

The ESS market for AI data centers, which forms the foundation of battery demand, is also growing rapidly. Market research firm TrendForce forecasts that the global installed capacity of ESS for AI data centers will grow from 15.7 gigawatt-hours (GWh) in 2024 to 216.8 GWh in 2030, representing an average annual growth rate of 46.1%. TrendForce analyzed that the power infrastructure of AI data centers is evolving toward a structure that utilizes large-scale BESS, rack-level BBUs, and supercapacitors on a floor-by-floor basis.

An industry official stated, “In AI data centers, it is becoming increasingly important not only to secure large amounts of power but also to respond stably to sudden changes in load,” adding, “As the power density of AI servers continues to rise, we expect demand to grow not only for ESS but also for batteries that support power stability within the data centers themselves.”