China Southern Power Grid and Huawei signed a comprehensive agreement aimed at strengthening their strategic partnership. The signing ceremony took place in Shenzhen on October 27, as reported by Huawei on October 29, according to IT Home.
The meeting was attended by Qian Chaoyan, Chairman of the power system operator, and Huawei's Deputy Chairman, along with rotating chairman Wang Tao, who held negotiations and witnessed the document signing.
The parties will focus on four key areas: coordination of energy, carbon emissions, and computing; application of artificial intelligence; the Power HarmonyOS system; and the development of the industry ecosystem. This joint effort aligns with the country's goal of creating a new energy system, a new power supply system, and new grid construction, as well as the coordinated development of six networks: water, new energy, computing, next-generation communications, underground urban pipelines, and logistics.
The partnership is based on an operating system launched in 2023. That year, China Southern Power Grid, with support from Huawei and the OpenAtom Foundation, introduced Dianhong—the first IoT operating system for energy in the country. This system is built on OpenHarmony and openEuler and allows one integrated solution to work with various types and brands of energy equipment, enabling connected devices and general data exchange.
Dianhong opened its open-source community in January 2026 and was included in the OpenAtom Foundation's incubation program in June.
According to an analysis published by TMTPost, the Dianhong community comprises over 500 industry partners and nearly 1900 adapted terminals where scenario testing is conducted in demonstration zones such as Qianhai in Shenzhen. The Shenzhen Supply Bureau has installed over 2600 Power HarmonyOS monitoring terminals across its main network, distribution network, and client sites. Dianhong adaptation certification is carried out by third-party organizations authorized by the open-source community, and analysis shows that certified products are becoming a prerequisite for large-scale network deployment.
Early examples have already been implemented in the 'energy-carbon-computing' track. For instance, a YOLOv5 object detection training task performed by a user in Suzhou was migrated to a large southern data center in Guiyang and ran for 108 hours, resulting in a 10 percent reduction in overall costs and consuming 113.4 kWh of green electricity. Similarly, a training task from Guangzhou moved to Guiyang reduced the total electricity cost by 16 percent. Analysts note that the presented samples remain small for now.
Earlier this month, Huawei and the Shenzhen Supply Bureau also implemented what they termed the first quantum-secure fiber optic network in the energy industry, capable of transmitting two types of signals over a single fiber.
