Delta Electronics Integrates Energy and Compute Infrastructure to Accelerate NVIDIA DSX-Based AI Factory Deployment

Delta Electronics has unveiled a system-level engineering approach designed to speed the deployment of AI factories built on the NVIDIA DSX AI Factory Platform, bridging the gap between available energy and productive compute. The NVIDIA DSX platform combines reference designs, open software, accelerated computing systems, facilities infrastructure and partner technologies into a codesigned framework for AI factory design, deployment and operations. Delta's contribution focuses on the physical infrastructure layer — connecting onsite energy, facility power, row and rack-level power and cooling, and chip-level solutions to help customers accelerate token production and generate higher AI yield from every available megawatt.

The initiative addresses a central challenge in AI factory economics: how much intelligence can be produced from each megawatt of available power. According to Austin Tseng, President of Delta Electronics Americas, the answer depends on more than facility power alone, as power quality, GPU load transients and heat must be engineered together all the way into the rack. By connecting energy, power, cooling and components across these layers, Delta and NVIDIA aim to help AI factory operators reach time-to-first-token sooner and optimize token output at scale.

NVIDIA's vice president of AI infrastructure, Vladimir Troy, noted that AI factories must be designed as complete systems with power, cooling and compute working together, adding that Delta's expertise in 800 VDC power delivery and liquid cooling will help customers deploy AI factories based on NVIDIA DSX faster and extract more AI performance from every megawatt.

At the facility level, Delta integrates energy storage systems, solid-state transformers and medium-voltage infrastructure to store, convert and distribute power required by high-density AI workloads. Across the IT space, the company combines 800 VDC power delivery with row-to-chip liquid cooling, aligning power flow and heat removal with the dynamic demands of high-density GPU computing. This system-level approach is intended to close integration gaps between facility and IT infrastructure for AI factories based on NVIDIA DSX.

Delta's 800 VDC power architecture reduces conversion stages between facility power and compute while delivering up to 98 percent power conversion efficiency and responding rapidly to GPU load transients. High-density in-row systems deliver up to 800 kilowatts in a single power rack, while tailored battery and capacitance backup technologies support transient stability. At the board level, Delta's 800 VDC DC-DC power distribution technology steps down directly to 50 VDC or 12 VDC with peak efficiency of up to 98.5 percent. Coordinating these technologies reduces electrical losses, heat generation, copper requirements and stranded capacity while reclaiming valuable white space for compute.

The same design continuity applies to thermal management. Delta's portfolio spans 2.4 megawatt and 3 megawatt liquid-to-liquid cooling distribution units, in-rack cooling and thermal solutions for next-generation AI racks. By coordinating electrical and thermal infrastructure at the facility, row, rack and component levels, Delta helps power and cooling systems meet the density and operating requirements of AI factories based on NVIDIA DSX.

To further accelerate deployment, Delta applies prefabrication and modular design to physical infrastructure. The company's Prefabricated AI Modular Data Center Solution integrates 800 VDC in-row power and 3 megawatts of liquid-cooling capacity into infrastructure blocks assembled and tested in the factory. Moving more integration and validation offsite reduces onsite complexity, streamlines commissioning and supports phased capacity expansion as demand grows.

Delta's system-level engineering and prefabricated delivery approach connect infrastructure design with execution, making constrained power more usable, accelerating the path to productive GPU capacity and supporting reliable token output at scale.