Overview of ZTE's New AIDC Solution

ZTE Corporation has unveiled a groundbreaking high-efficiency AI Data Center (AIDC) solution designed to meet the escalating demands of modern artificial intelligence workloads. The newly developed architecture focuses on achieving ultimate energy efficiency by seamlessly integrating an 800V High Voltage Direct Current (HVDC) power supply architecture with full-stack liquid cooling technology. As AI models grow in complexity, the thermal and power requirements of data centers have reached extreme levels. ZTE’s comprehensive approach addresses these challenges head-on, aiming to maximize inference efficiency while maintaining a sustainable and highly optimized operational footprint.

Technical Details: 800V HVDC Power Architecture

At the core of ZTE’s new AIDC solution is the implementation of an 800V HVDC power supply architecture. Traditional data center power distributions often rely on lower voltage alternating current systems, which incur significant conversion losses as power is stepped down through multiple stages to reach the server racks. By elevating the distribution voltage to 800V DC, ZTE’s architecture drastically reduces the number of power conversion stages. This high-voltage direct current approach minimizes transmission losses, improves overall power utilization efficiency, and ensures that a greater percentage of the drawn power is delivered directly to the compute hardware. For AI workloads, which are notoriously power-hungry, this reduction in electrical overhead is critical for maintaining operational viability.

Thermal Management: Full-Stack Liquid Cooling

Complementing the advanced power architecture is ZTE’s full-stack liquid cooling technology. Modern AI accelerators and high-density server configurations generate extreme thermal loads that traditional air-cooling systems can no longer manage effectively. ZTE’s full-stack liquid cooling approach encompasses the entire thermal management chain, from the chip level to the facility level. By utilizing liquid as the primary heat transfer medium, the system achieves vastly superior thermal conductivity compared to air. This allows for higher compute densities without the risk of thermal throttling, ensuring that the AI hardware operates within optimal temperature ranges to sustain peak performance.

Optimizing Inference Efficiency and Token Operations

The hardware innovations in power and cooling are explicitly designed to support ZTE’s broader strategic vision for AI, which was recently highlighted at MWC Shanghai 2026. During the event, ZTE showcased its full-stack AI capabilities, emphasizing a shift in industry competition from merely scaling computing power to optimizing token operations. The new AIDC solution directly supports this TCO-optimal AI factory concept. By resolving the physical bottlenecks of power delivery and heat dissipation, the solution ensures that AI accelerators can operate at maximum capacity for inference tasks, unlocking new momentum in token efficiency.

Market Context and Industry Impact

ZTE Corporation, a major provider of integrated information and communication technology solutions, is positioning itself at the forefront of the AI infrastructure boom. The industry is currently grappling with the physical limits of data center expansion; power grids are strained, and cooling footprints are expanding. By offering a solution that intrinsically addresses both power density and thermal limits, ZTE is providing a highly competitive alternative to conventional data center designs. This holistic approach not only reduces the physical footprint of AI factories but also aligns with global sustainability goals by maximizing energy efficiency across the board.

Implications for Future AI Data Centers

The integration of 800V HVDC and full-stack liquid cooling sets a new benchmark for future AI data center designs. As AI agents become more widespread, the baseline requirements for data centers will continue to escalate. ZTE’s solution demonstrates that achieving higher performance requires a fundamental rethinking of the facility's physical infrastructure. Key implications for future AIDCs include:

  • A shift from scaling raw compute to optimizing physical infrastructure for token efficiency.

  • The necessity of extreme synergy across computing, network, storage, energy, and software.

  • Adoption of high-voltage direct current and liquid cooling as standard paradigms for TCO-optimal AI factories.

Future Outlook

Looking ahead, ZTE’s high-efficiency AIDC solution represents a critical step in the evolution of AI infrastructure. As the company continues to leverage its end-to-end, full-stack AI capabilities, the focus will remain on system-level architecture innovations that drive token efficiency. The successful deployment of the 800V HVDC and full-stack liquid cooling technologies will provide ZTE with a distinct advantage in the rapidly expanding AI hardware market. Ultimately, by solving the extreme thermal and power requirements of modern AI workloads, ZTE is empowering the new era of token operations, ensuring that the next generation of AI applications can be deployed efficiently, sustainably, and at scale.