Computing Power Rental Industry Enters Deep Consolidation Phase, Who Will Have the Last Laugh?

The rapid advancement of AI large models is pushing the computing power leasing industry into a capital-driven spotlight.

On one hand, “computing and electricity coordination” has been included in the national new infrastructure plan, with the East Data West Computing project upgraded; on the other hand, cloud providers and specialized leasing companies are competing on the same stage, with cross-industry players continuously entering, creating a “hundred schools of thought” scene.

Data shows that by 2026, China’s computing power leasing market is expected to reach 260 billion yuan, with an annual growth rate of over 20%. The high-end computing power gap exceeds 35%, the total market value of the sector reaches 3.68 trillion yuan, and top companies have an over 90% listing rate.

However, behind the boom, problems are emerging: regional mismatches between supply and demand, lack of industry standards, monopolies on high-end hardware, rising entry barriers for small and medium-sized enterprises, and some companies facing issues like insufficient green energy qualifications and weak technology.

An industry reshuffle centered on resources, technology, and capital is accelerating.

Industry Landscape Divergence

On March 18, Alibaba Cloud announced a maximum increase of 34% in AI computing power products, with the leasing index rising 3.94% that day; just one trading day later, the sector index fell back by 4.58%.

This sharp fluctuation triggered by a price adjustment reveals the underlying contradictions in the computing power leasing industry: price increases have not yet guaranteed profitability, demand surges and supply are diverging, and the sector is shifting from a concept frenzy into a deep reshuffle.

As the leading domestic cloud computing provider, Alibaba Cloud’s price adjustment has a benchmark effect on the industry. Baidu Smart Cloud and Tencent Cloud have followed suit, raising their computing service prices, pushing the industry price center up by 10%-20%.

This change stems from sustained growth in AI inference demand, rising hardware procurement costs, and cloud providers shifting resources toward high-value businesses. According to China Research and Planning, by 2026, China’s computing power leasing market could reach 260 billion yuan, with an annual growth rate exceeding 20%, and demand for intelligent computing power growing at 43%.

Behind the prosperity, structural issues are becoming more prominent: limited supply of high-end GPUs, small and medium-sized vendors struggling to secure stable chip resources; intense homogenous competition in low-end computing power, with some companies lowering prices to increase listings, squeezing profit margins; rapid growth in demand for domestic computing power in regulated scenarios like government and finance, but industry-wide challenges in adaptation and service capabilities.

Price hikes have not resolved fundamental industry contradictions but have instead accelerated differentiation among companies. Those with stable resources, clients, and technological barriers will gain more market share, while firms lacking core competitiveness are gradually pushed out.

Industry analysts observe three main trends: first, increasing market concentration, with leading companies leveraging scale, resources, and customer base to capture market share, while small and medium players exit gradually; second, a dual-track supply structure, with high-end overseas GPUs and domestic computing power developing in parallel, and domestic penetration in compliant scenarios steadily increasing; third, upgrading profit models from simple hardware leasing to a combination of “computing + services + models,” significantly increasing added value and stability.

Different companies adopt tailored strategies. Leading firms focus on expanding scale and securing long-term contracts with cloud providers to ensure occupancy and profitability; domestic computing power vendors deepen their presence in regulated sectors like government and finance, advancing chip adaptation and ecosystem development to benefit from policies and orders; small and medium enterprises target regional markets and edge scenarios, seeking survival through low costs and rapid response.

In terms of actual results, companies with stable long-term contracts, domestic chip adaptation capabilities, and cost control tend to realize profits faster, while those relying solely on concepts without substantial orders or technological barriers face ongoing operational pressure.

Pingshi Information’s Differentiated Positioning

Pingshi Information entered the computing power leasing sector through digital reading services, leveraging domestic computing power and partnerships with telecom operators to establish a differentiated position amid industry reshuffling.

The company controls Tianxin Electronics to develop servers independently, effectively reducing hardware procurement costs, and has completed adaptation with domestic chip manufacturers like Huawei Ascend and Cambrian, building delivery capabilities for domestic computing power integrated machines. As of March 2026, the company has accumulated computing power orders exceeding 2.4 billion yuan.

In the domestic computing power sector, Pingshi’s advantages focus on three points: first, deep integration with Huawei Ascend’s ecosystem to prioritize domestic chip resources and avoid risks from overseas chip bans; second, leveraging telecom operator channels for strong order certainty and benefiting from local computing power subsidies; third, adopting a “staged delivery and rolling production” model to control heavy asset investments and reduce financial risks.

In the short term, Pingshi is positioned in the second tier of the industry. Long-term prospects depend on continuously expanding domestic computing power capacity, optimizing customer structure, and enhancing service value, which could secure stable market share in regulated scenarios and regional markets.

Core Competitiveness Comparison of Mainstream Players in the Computing Power Leasing Industry

The industry remains in a high-growth cycle, with demand growth and domestic substitution supporting long-term development. However, investment logic has shifted from “speculating on concepts” to “verifying performance.” Companies with substantial orders, realized profits, and technological barriers will enjoy valuation premiums.

For Pingshi, in the short term, it needs to accelerate order fulfillment and capacity release, improve listing rates and gross margins, and turn losses into profits; mid-term, it should expand domestic computing power capacity, develop nationwide markets and high-end clients, and reduce regional and customer concentration risks; long-term, it must build an integrated “computing + services + scenarios” capability, transforming from a pure leasing provider to a comprehensive computing power service provider.

Industry reshuffling is not the end but a starting point for high-quality companies to break through. Under the trend of domestic substitution and digital economy growth, Pingshi’s core challenge is how to convert its domestic computing power advantage into sustainable profitability and growth, establishing a stable position in the infrastructure of the computing era.