KRES POWER SYSTEMS ONTARIO CORPORATION

Abstract

In June 2026, the Energy Bureau of Guangdong Province officially approved the energy-saving report for Phase I of China Mobile Guangdong-Hong Kong-Macao Greater Bay Area (Huizhou) Data Center, which passed the review successfully. With a total investment of 905.7 million yuan, the project is scheduled to be put into full operation in December 2030. Upon completion, it will deliver an intelligent computing capacity of 8,496 PFLOPS. Boasting an ultra-low PUE within 1.249 and 100% coverage of green power consumption, it will be built into a benchmark green intelligent computing hub featuring high standards and low energy consumption across the Guangdong-Hong Kong-Macao Greater Bay Area.

1. Project Overview: Huizhou Adds Bay-Level Computing Infrastructure

The newly approved computing project is located in the Guangdong-Hong Kong-Macao Greater Bay Area (Huizhou) Data Industrial Park in Baihua Town, Huidong County, Huizhou City. It will be reconstructed and built based on four existing buildings within the park. By revitalizing stock site resources, the construction cycle can be effectively shortened and overall energy consumption during construction reduced. This key priority project enables China Mobile to improve its computing power layout in the eastern part of the Greater Bay Area and consolidate regional digital infrastructure.

The project boasts a total investment of 905.7 million yuan and is planned to complete construction and launch operation in December 2030. A total of 4,572 physical racks will be deployed, equivalent to 16,576 standard 2.5kW racks by industry metrics, with an overall intelligent computing scale of 8,496 PFLOPS. In the future, it will mainly undertake computing demands of high-end digital industries including large AI model training, cloud computing services, digital twins and autonomous driving, providing core computing support for the development of emerging industries in the Greater Bay Area.

2. Full-Gradient Computing Layout Adapted to Diverse Business Scenarios

To precisely match computing load demands of various scenarios, the project adopts a gradient and differentiated rack configuration scheme. All installed equipment meets national Class I energy efficiency standards, enabling refined energy-saving control while delivering robust computing output. The total IT power of the project reaches 41,440 kW.

Specifically, 272 ultra-high-density liquid-cooled racks of 42kW are configured exclusively for high-computing scenarios such as large AI model training and ultra-high-load intelligent computing. Alongside them are 52 medium-high density 15kW racks to support mid-to-high-end cloud computing and high-frequency real-time computing businesses. Meanwhile, 3,748 general-purpose 7kW racks serve as the main computing units to fully cover mainstream commercial scenarios including government and enterprise cloud services and AI inference. Complementary to these are 500 lightweight 6kW racks catering to basic businesses such as mass data storage, edge computing deployment and lightweight hosting, forming a complete full-coverage computing system with high and low-density collocation.

3. Leading Energy-Saving Technologies to Build Ultra-Low PUE Green Data Center

The project integrates multiple cutting-edge industry green energy-saving technologies to comprehensively optimize three major dimensions: data center air distribution, cooling system architecture and power supply guarantee, drastically cutting operational energy consumption and capping the overall PUE at below 1.249, a leading level among domestic data centers.

In terms of data center layout, racks are arranged in a standardized back-to-back layout fitted with closable sliding doors at the head and tail of each row to form independent hot and cold aisles. Combined with a scientific air distribution mode of horizontal air supply and rear-end air return, mixing of hot and cold air that causes energy loss is completely eliminated, greatly improving the overall heat exchange efficiency of the data center.

A scenario-customized cooling system is deployed. Cold plate liquid cooling paired with chilled water cooling is adopted for the 42kW ultra-high-density high-load racks, while water-cooled in-row air conditioners are uniformly equipped for all other conventional racks. All cooling equipment features N+1 redundant backup and an intelligent group control management system that dynamically adjusts operating power based on real-time data center loads, balancing energy conservation and stable operation.

For power supply, a highly reliable dual-redundancy power distribution system is constructed with independent dual A/B circuits and dual bus power distribution architecture. Paired with large-capacity uninterruptible power supplies and emergency diesel generators, it delivers all-round guarantee for 24/7 stable power supply of the data center, fully meeting the high stability operation requirements of core businesses such as artificial intelligence, finance and government affairs.

4. Strict Carbon Consumption Control to Achieve 100% Green Power Coverage

The project strictly aligns with national and Guangdong provincial standards for 'dual carbon' development, with all energy consumption, carbon emission and energy efficiency indicators under rigorous control, highlighting prominent green and low-carbon attributes. Upon completion, the annual comprehensive energy consumption shall not exceed 47,219.86 tons of standard coal, annual electricity consumption shall be controlled within 383.6094 million kWh, and total annual carbon emissions shall not surpass 169,676.18 tons. Its carbon utilization efficiency and carbon emissions per unit industrial added value both outperform industrial benchmarks.

In terms of renewable energy application, the project commits to 100% green power consumption across the board via purchasing green electricity certificates to achieve full green power coverage. It will also continuously optimize its operations in accordance with the latest national, provincial and municipal policies on green power and green certificates, and further develop low-carbon operation models for green computing power.

5. Project Value: Consolidating the Foundation for Digital Economic Development in the Greater Bay Area

As a core pivotal node in China Mobile’s computing power network layout across the Guangdong-Hong Kong-Macao Greater Bay Area, the construction of the Huizhou Intelligent Computing Data Center effectively addresses the shortage of high-end intelligent computing resources in the eastern Greater Bay Area and further improves the coordinated scheduling system for computing power covering the entire bay region.

Once fully operational, it will continuously provide high-computing, low-energy-consumption and highly reliable infrastructure support for emerging digital industries including artificial intelligence, big data, industrial internet and autonomous driving. It will help Huizhou develop into a key computing power hub in the Greater Bay Area, and continuously empower large-scale, low-carbon and high-quality development of the regional digital economy.

Jiayida (Suzhou) Fluid Power Technology Co., Ltd. focuses on providing one-stop liquid cooling fluid system solutions and services for data center scenarios. The product portfolio covers OCP-standard UQD, UQDB and BMQC quick disconnect couplings, water manifold assemblies, Continental EPDM hoses and hose assemblies, FF9 full-flow double interlock quick disconnect ball valves as an alternative to FD83 series products, API heat exchangers, Gates data center-dedicated EPDM hoses, as well as WIKA pressure gauges and sensors from Germany.

Phase I of China Mobile Greater Bay Area (Huizhou) Intelligent Computing Data Center Approved with Total Investment of 905.7 Million Yuan