DSX Ready qualification and data center demand
LG Energy Solution has joined DSX Ready, Nvidia's qualification program for AI factory power and cooling infrastructure, as a battery energy storage system (BESS) partner. Its integrated energy storage system, JF2 AC Link, met the relevant requirements. At the program's launch, BESS solutions from three companies worldwide had qualified. LG Energy Solution is the only South Korean battery company among them.
As electricity consumption at AI data centers increases rapidly, battery energy storage systems are emerging as a new growth market. When large numbers of GPUs start or stop computing simultaneously, electricity demand changes sharply over a short period. BESS charge and discharge to buffer these fluctuations.

According to BloombergNEF, US data center electricity consumption is expected to more than double from 180 TWh in 2025 to 391 TWh in 2030.
Nvidia's DSX Ready is a program that verifies whether partner products related to power and cooling for AI factories meet requirements. In the BESS category, its main focus is not on the battery cells themselves but on the power control performance required when the system is connected to the actual grid at its AC terminals, including the power conversion system (PCS).
Integrated BESS features
LG Energy Solution's JF2 AC Link is an integrated BESS combining LFP batteries, a PCS, a liquid-cooling system and an energy management system (EMS).
Each block stores 5.1 MWh of energy and can supply its rated power of 2.5 MW for about two hours. Its operating temperature ranges from minus 30 degrees Celsius to plus 50 degrees Celsius, and it uses its proprietary AEROS EMS to manage charging, discharging and system status.
Rapid response to power fluctuations required by AI data centers is another key feature. It supports AI load leveling, which reduces the burden on the grid by adjusting BESS charging and discharging when AI server loads rise or fall sharply.
It also incorporates voltage support to help maintain a stable power supply during sudden grid voltage changes, as well as grid-forming functionality that enables the inverter to establish its own voltage and frequency references.
The deployment approach is modular. Blocks rated at 2.5 MW and 5.1 MWh can be combined to increase capacity in stages according to the data center's size and the pace of GPU expansion. The battery, PCS, cooling and EMS are integrated in advance to reduce on-site installation work.
The safety design uses LFP cells and takes relevant UL 9540 standards into account. The company said the system had also undergone UL 9540A testing, which evaluates thermal runaway and fire propagation characteristics, and large-scale fire testing. Gas detection, active venting, and fire detection and alarm systems are also included.
North American production and system operations
The company is also expanding North American production capacity to support its efforts to serve the large-scale AI data center market.
LG Energy Solution is operating or developing five North American production sites: Lansing and Holland in Michigan; NextStar Energy in Ontario, Canada; Ultium Cells in Tennessee; and L-H Battery Company in Ohio. It plans to expand global ESS production capacity to more than 60 GWh by the end of 2026, with more than 50 GWh located in North America.
Its US subsidiary Vertech is responsible for system deployment and operations. Vertech supports ESS design, power control, monitoring, data analysis and maintenance, using AEROS EMS and the cloud-based AEROS Cloud to manage operating conditions.
LG Energy Solution plans to use its participation in Nvidia's DSX Ready program to expand its BESS business for AI data centers. As rising AI computing demand extends data center investment competition beyond GPUs and servers to power and cooling infrastructure, the company's strategy is to secure new demand through local North American production and system integration capabilities.
Original Korean article: autodaily.co.kr
English brief by Korea Wind Intelligence, koreawindintel.com

















