AI Trends

Guangdong Issues '15th Five-Year Plan' to Advance Greater Bay Area Computing Hub Construction, Boosting A-Share Computing Rental Sector

Sina TechUpdated Nov 18 18:27

AI summary

On November 18, the Guangdong Provincial People's Government issued the "Digital Guangdong Construction 15th Five-Year Plan," proposing to comprehensively promote the construction of the Greater Bay Area computing hub and optimize facility layout. Driven by this positive policy development, the A-share computing rental concept sector rose on the day, with multiple component stocks such as Guangmai Technology and Xuanji Information hitting their daily limit up.

Body · AI translation

New Crowdsourced Testing Promotion Copy New Crowdsourced Testing Promotion Copy end Add App-exclusive Traffic Diversion

November 18, 2026 17:53:47

On November 18, the A-share computing power leasing concept rose. Stocks such as Guangmai Technology (920924), Xuanji Information (300324.SZ), Hongjing Technology (301396.SZ), Qingyun Technology-U (688316.SH), Huasheng Shares (600156.SH), and Lianhua Holdings (600186.SH) among others hit the daily limit up.

In terms of news, the People's Government of Guangdong Province issued the "Digital Guangdong Construction '15th Five-Year Plan'" (Shiziwu Guihua), which points out strengthening the high-quality computing power supply system. It will coordinate the construction of the Greater Bay Area computing hub, optimize the layout of computing facilities, and develop computing power in line with actual conditions.

Precise Implementation at the Policy Level

The "Digital Guangdong Construction '15th Five-Year Plan'" issued by Guangdong Province brings clear regional policy dividends to the computing power leasing industry and pushes the layout logic of national computing infrastructure construction to a deeper level. The plan explicitly proposes coordinating the construction of the Greater Bay Area computing hub, optimizing the layout of computing facilities, avoiding homogeneous low-level repetitive construction, and differentially laying out general computing, intelligent computing, and supercomputing resources based on the actual needs of industries within the region such as advanced manufacturing, artificial intelligence, and cross-border finance.

More imaginatively, the plan proposes expanding broadband capacity between networks in the Guangdong-Hong Kong-Macao Greater Bay Area, interconnecting digital intelligence application service platforms, laying out cross-border computing corridors, deepening cooperation in the field of artificial intelligence among Guangdong, Hong Kong, and Macao, and jointly building digital industry clusters. This means that the Greater Bay Area will break down barriers to computing power circulation between the three regions, promote unified scheduling and sharing of computing resources within the region, and build a global computing service highland. For domestic computing power leasing enterprises, this not only means adding a large amount of computing demand from cross-border scenarios but also promoting the upgrade of the industry's service capabilities from single computing rental to comprehensive computing solutions across regions and systems.

Before this, the demand for domestic computing power leasing was mainly concentrated in the training and inference scenarios of internet giants and AI large model enterprises, while a large number of small and medium-sized enterprises still faced the pain points of "difficulty in obtaining computing power and high computing costs." The Greater Bay Area gathers massive numbers of manufacturing enterprises, cross-border e-commerce enterprises, and technology startups. With the continuous advancement of the "Artificial Intelligence+" action, the intelligent transformation of vertical industries will release a huge amount of fragmented computing demand. This part of the demand has not been fully covered previously, which happens to provide a new growth curve for computing power leasing enterprises.

Clear Logic of Target Stratification

The internal structure of the A-share computing power leasing sector is also undergoing profound changes. The market's pricing logic has completely shifted from past "theme speculation" to the core dimensions of "real computing scale, ability to land long-term orders, and cash flow quality."

Currently, targets within the industry are clearly divided into three tiers. The first tier consists of companies that have completed large-scale computing deployment and hold long-term orders from top clients. These companies have the strongest earnings certainty, and their profit elasticity has already been verified in financial reports. For example, Xiechuang Data is the fastest-expanding new giant in the computing track. In early 2026, it announced its intention to spend 11 billion yuan purchasing servers. It has currently deployed over 80,000 P of computing power. In October, it officially signed a single computing cloud service agreement worth 37.49 billion yuan, equivalent to more than 13 times its intelligent computing revenue for the full year of 2025.

The second tier includes companies deeply bound to regional policies and having laid out high-end infrastructure such as liquid-cooled intelligent computing centers in advance. These companies will directly benefit from the construction dividends of computing hubs such as the Greater Bay Area. For example, Pinggao Shares is a rare domestic service provider focusing on high-end computing services for government and enterprises. It has cultivated the three high-barrier tracks of government affairs, military industry, and finance. Its independently developed cloud platform architecture adapts to the special computing needs of government compliance, military high security, and financial high stability, with a customer renewal rate exceeding 95%. It has laid out government and enterprise computing services in the Guangdong region for many years, able to precisely connect with the customized computing needs of government departments and local key industries during the Digital Guangdong construction process, thereby avoiding homogeneous low-price competition in general computing.

The third tier consists of cross-boundary players in transition, who are still in the stage of investing in computing resources, and whose performance release still requires time for verification. For example, Lianhua Holdings relies on the stable cash flow of its original traditional main business to support its layout in the computing track. It is a typical representative of traditional enterprises transitioning into computing power leasing. In July, it newly signed an 815 million yuan computing power lease contract. It has currently deployed over 15,000 P of computing power, with customers mainly consisting of financial and regional internet enterprises. The gross margin of its computing business exceeds 50%, but its overall computing scale is still in the middle-to-lower range of the industry, and it needs to continue investing to expand capacity in the future. Statement: Sina News exclusive manuscript, unauthorized reprinting is strictly prohibited. tech_web_article_v2015_block_bottom Non-directed 300*250 button 17/09 wenjing begin

Non-directed 300*250 button end

Source

Sina Tech