Innovative Grid-Interactive AIDC Solution Redefines AI Infrastructure at HUAWEI CONNECT 2026

Huawei's Grid-Interactive AIDC Solution

On September 20, 2026, Huawei officially launched the HUAWEI CONNECT 2026 in Shanghai alongside the HUAWEI AIDC Facility Summit. The event showcased Huawei's innovative Grid-Interactive AIDC solution, setting a new benchmark in AI infrastructure and focusing on maximizing tokens per watt.

Architectural Innovations and Systematic Design

During the summit, Hou Jinlong, a Director on Huawei's Board and President of Huawei Digital Power, delivered a keynote address highlighting four key insights into the architectural innovations and systematic designs essential for addressing challenges related to AIDC (Artificial Intelligence Data Centers). These challenges include power supply, power quality, rack energy usage, and heat dissipation.

1. Innovative Power Supply Architecture: A state-of-the-art power supply strategy combined with liquid cooling mechanisms ensures a stable power provision while effectively managing heat dispersion, resulting in reliability and operational efficiency.
2. Multi-Layer Hybrid Energy Storage: This system compensates for local fluctuations in electricity supply, enhancing the energy quality and grid compatibility of AIDCs.
3. Integration of Power and Computing Synergy: By blurring the lines between energy consumption and computational requirements, Huawei’s approach transitions AIDCs from mere energy loads to vital components of a modern energy ecosystem.
4. Modular Prefabrication: Innovations in design and delivery significantly reduce lead times, facilitating quicker deployment of AIDC solutions.

Hou emphasized that achieving maximum tokens per watt while minimizing token costs will be vital to the competitiveness of data centers in the future.

Adapting to Evolving Energy Grids

Bob He, Vice President of Huawei Digital Power, elaborated on how the increasing share of renewable energy influences grid characteristics and the need for a responsive energy infrastructure. The fluctuations inherent in AI workloads necessitate a seamless integration of computation and energy infrastructures.

By leveraging "3+1" innovations, Huawei’s Grid-Interactive AIDC solution emerges as a reliable, energy-efficient, and rapidly deployable framework that focuses on maximizing token production per watt.

Huawei proposes an innovative future-oriented MIMO (Multiple Input Multiple Output) power supply architecture that adapts to industry challenges. Their latest AIDC 1.0 solution integrates grid-friendly UPS systems, smart lithium batteries, and energy storage systems capable of grid formation, allowing AIDCs to support the energy grid while accommodating AI workload fluctuations.

Advanced Cooling with AI

For cooling efficiency, Huawei introduced an AI-powered liquid cooling system that anticipates the states of operational fluids and facilitates predictive maintenance, keeping data centers at optimal temperatures without compromising performance.

Future of AI Infrastructure

According to Xia Qin, Huawei’s Chief Strategy Planning Expert, the AIDC landscape is evolving towards a more integrated architecture combining computing power, networking, and storage, with energy being a pivotal element in this equation. A highly reliable, energy-efficient, and rapidly deployable energy system is crucial for the dependable operation of hyperscale AI computing clusters, thus laying a robust foundation for extensive training and inference processes in the age of agent-based systems.

Traditional data centers have limited connections to the power grid. However, AI workloads are susceptible to frequent changes that can disrupt network operations, leading to outages and instability. Josh Chen, founder and chairman of the VNET Group, discussed redefining energy sources and campus layouts to foster mutual support between energy sources and buildings.

Conclusion

HUAWEI CONNECT 2026 marked a significant step in AI infrastructure evolution, with Huawei unveiling its Grid-Interactive AIDC solution and comprehensive AI inference approach, promising a remarkably dependable, energy-efficient, and swiftly deployable computing infrastructure for AI clusters. These advancements underscore Huawei's endeavor to strengthen its position as a leader in the field, effectively addressing the increasing energy demands of AI technologies.