In Guangdong, the electricity consumption of computing power increased by 21% in six months, and the annual increase in electricity consumption of computing power nationwide will exceed one billion degrees.

With the acceleration of AI large model iteration, the domestic electricity demand for computing power infrastructure continues to explode. On July 21, a reporter from the Economic Daily learned from Guangdong Power Grid under the Southern Power Grid that in the first half of 2026, the total electricity consumption of 309 data centers in Guangdong Province reached 59.74 billion kilowatt-hours, a substantial increase of 21.15% compared to the same period last year. Among them, the electricity consumption of data centers in Shaoguan City surged by as much as 409.10%, leading the province. Behind the continuous growth of the electricity consumption for computing power is the accelerating development trend of the AI large models, big data, and industrial internet industries landing and expanding in Guangdong Province. Therefore, "electricity-computing synergy" has become a key factor supporting the two-way development of the digital industry and the new power system. Guangdong Power Grid stated that it will continue to build a two-way development model of "electricity supporting computing, computing optimizing electricity" to promote Guangdong's construction as a leading industrial highland for computing power nationwide. On July 14, Zhang Lin, Director of Planning and Development Department of China Electric Power Industry Association, announced at a related conference that it is estimated that during the "14th Five-Year Plan" period, the annual average increase in electricity consumption for computing power nationwide will exceed 100 billion kilowatt-hours, reaching an estimated 800 billion kilowatt-hours by 2030, accounting for around 6% of the total social electricity consumption. The electricity consumption for computing power will become a major increment of future power consumption.