Major manufacturers are stepping up to address one of AI's biggest infrastructure challenges: keeping data centers powered and cool as GPU energy consumption skyrockets. LG Electronics and LG Energy Solution have joined Nvidia's newly launched DSX Ready program, bringing battery storage and advanced cooling solutions to a market desperately seeking reliable power infrastructure for AI factories.

What Is the DSX Ready Program and Why Does It Matter?

Nvidia introduced DSX Ready as a qualification program for partner products that meet the requirements of its DSX AI factory reference designs. The program focuses specifically on power and cooling infrastructure, two areas becoming increasingly critical as AI data centers grow larger and more power-intensive. Think of it as a certification seal that tells data center operators which products are compatible with Nvidia's standardized approach to building AI facilities.

The initial lineup includes some heavyweight names. LG Energy Solution's JF2 AC Link became the only Korean battery maker in the program's first wave, joining Tesla and Hitachi Energy as qualified battery storage providers. Hitachi Energy's battery energy storage system also qualified as DSX Ready BESS, meeting applicable reference design requirements for AI factory power infrastructure.

How Are These Companies Addressing AI's Power Problem?

LG Energy Solution's approach focuses on intelligent power management. The JF2 AC Link uses lithium iron phosphate batteries and features what the company calls "AI load smoothing," which charges and discharges in response to changes in server power consumption to reduce sudden fluctuations in electricity demand. The system also includes a voltage ride-through function designed to maintain stable power supply during abrupt drops or spikes in grid voltage.

The modular design is particularly important for data center operators. It reduces construction time and allows capacity to be added as data centers expand, meaning operators don't need to build out their entire power infrastructure at once. LG Energy Solution is backing this up with real infrastructure investment, planning to secure more than 50 gigawatt-hours of energy storage system production capacity in North America by the end of 2026, targeting growing demand from AI data centers in the region.

LG Electronics, meanwhile, is tackling the cooling side of the equation. The company has qualified its 2.5-megawatt coolant distribution unit (CDU) as DSX Ready, designed for in-row cooling configurations that group multiple server racks together and cool them through a high-capacity system. This setup is suited to high-density AI data centers, where increasingly powerful chips drive up power consumption and heat generation.

The 2.5 MW model represents an expansion of LG Electronics' cooling portfolio. The company previously received Nvidia specification approvals for 600-kilowatt and 1 MW CDU models in July and September, respectively. CDUs are essential components of liquid-cooling systems for AI servers, and as power density per rack rises from hundreds of kilowatts toward the megawatt range, data center operators are shifting from cooling individual racks to larger systems capable of handling multiple racks at once.

What Makes These Solutions Stand Out in the Market?

Both companies have built comprehensive portfolios addressing different aspects of data center infrastructure. LG Electronics has developed what it calls a "chip to chiller" portfolio covering the broader cooling chain for AI data centers. This includes:

• Coolant Distribution Units: High-capacity systems that cool multiple server racks simultaneously in in-row configurations.
• Chillers: Equipment that removes heat from the coolant circulating through the data center.
• Computer Room Air Handlers: Systems that manage overall air temperature and humidity in the facility.
• Cold Plates: Components that draw heat directly from AI chips before it spreads throughout the system.

Hitachi Energy's solution emphasizes flexibility and grid integration. The company's qualified BESS combines energy storage, power conversion, automation, and digital technologies in an integrated architecture designed to support the dynamic energy requirements of AI-driven operations. By enabling precise power balancing, grid support, and energy optimization, the solution helps align battery storage deployments with the performance, efficiency, and reliability objectives of Nvidia's DSX Platform.

"As AI factories continue to scale, customers need power systems that can be deployed, integrated, and expanded quickly while maintaining reliability and resilience. As one of the first qualified BESS solutions in the NVIDIA DSX Ready program, this recognition reinforces Hitachi Energy's position at the forefront of AI factory electrification combining advanced technologies, power quality solutions and systems expertise to support end-to-end power infrastructure for the next generation of AI applications," stated Marco Berardi, Head of Grid and Power Quality Solutions and Service at Hitachi Energy.

Marco Berardi, Head of Grid and Power Quality Solutions and Service at Hitachi Energy

Why Is This Timing Critical for the AI Industry?

The qualification of these solutions comes at a crucial moment. AI data centers are consuming electricity at unprecedented rates, and the infrastructure to support them hasn't kept pace. Traditional power grids and cooling systems designed for conventional data centers simply cannot handle the sustained, high-intensity power demands of modern AI workloads.

Battery energy storage systems serve a dual purpose in this context. Beyond storing energy, they act as flexible energy platforms supporting power continuity, power quality, and operational flexibility while helping facilities respond to changing load conditions and applicable grid requirements. For data center operators, this means they can better integrate renewable power sources, support grid code compliance, and operate more effectively within available grid capacity.

LG Electronics plans to combine its cooling products into customized solutions for AI data center operators as it expands its global market presence. This modular, customizable approach reflects a broader industry shift toward standardized yet flexible infrastructure that can adapt to different facility sizes and configurations.

The DSX Ready program itself represents a significant step toward standardization in AI infrastructure. By establishing clear reference designs and qualification criteria, Nvidia is helping the industry move faster. Data center operators can now identify compatible products more easily, reducing deployment planning complexity and supporting reliable operations across interconnected systems including power, cooling, monitoring, and facilities management.

As AI continues to reshape industries and drive electricity demand, the infrastructure supporting these facilities will become as important as the computing power itself. Companies like LG and Hitachi are positioning themselves at the center of this critical infrastructure challenge, offering solutions designed specifically for the unique demands of AI factories.