23H1 of the three major telecom operators: All in Cloud Computing + AI

Stable market structure, substantial dividends, and the familiar telecom operators.

However, they have made some changes this year.

They have transformed from simple communication service providers into digital technology companies, with a shift in focus from the consumer (C-end) to the business (B-end). This transformation will bring new growth opportunities for telecom operators in the next decade.

In the mid-year report of the 23rd year, the three major telecom operators have something new.

Key points of this article: Digital transformation and AI will usher in the next 10-year cycle for telecom operators.

Profitable first half of the year, generous dividends of 63 billion

Telecom operators have the longest cycle among all TMT industries.

The electronics industry may have a technology cycle of four years, applications may have a cycle of hard three years, soft three years, and business models may have a cycle of another three years. However, telecom operators have a cycle of ten years.

And now, we are at the starting point of a ten-year cycle for telecom operators.

From the financial reports: The mid-year reports of the three major telecom operators continue the stable revenue of the first quarter. In addition to announcing huge dividends, digitalization remains a key focus.

China Mobile: Achieved operating revenue of 530.7 billion yuan, a year-on-year increase of 6.8%; net profit attributable to the parent company's shareholders was 76.2 billion yuan, a year-on-year increase of 8.4%.

China Unicom: Achieved operating revenue of 191.83 billion yuan, a year-on-year increase of 8.8%; net profit attributable to the parent company was 5.44 billion yuan, a year-on-year increase of 13.7%.

China Telecom: Achieved operating revenue of 258.7 billion yuan, a year-on-year increase of 7.7%, maintaining growth for 10 consecutive years; net profit attributable to shareholders of the listed company was 20.2 billion yuan, a year-on-year increase of 10.2%.

Cloud services have become the core driving force for digital revenue in the industry, specifically:

Mobile Cloud: Revenue reached 42.2 billion yuan, a year-on-year increase of 80.5%, with a top 3 market share in IaaS revenue, and a growth rate of over 100% in self-owned IaaS+PaaS revenue.

Unicom Cloud: Achieved revenue of 25.5 billion yuan, a year-on-year increase of 36%. The company stated that by 2023, Unicom Cloud aims to achieve revenue exceeding 50 billion yuan.

Tianyi Cloud: Revenue reached 45.9 billion yuan, a year-on-year increase of 63.4%. The amount of signed projects in the first half of the year was 26.4 billion yuan, with over 3.5 million industry customers, including over 600,000 government customers, accounting for 27.8% of the revenue.

It is worth noting that the three major telecom operators have announced generous dividends for the mid-term: China Mobile, China Telecom, and China Unicom will distribute dividends of 47.551 billion yuan, 13.104 billion yuan, and 2.491 billion yuan, respectively. By this, it can be seen that the underlying strength of the operators remains robust. Although the overall revenue growth is not fast, it is very stable and continuous, with considerable dividend returns.

Currently, the operators' cloud resource pool and edge nodes have begun to take shape. As the builders and operators of communication infrastructure, the abundant cloud and network resources can better serve the digital transformation of enterprises, which will become a window for the future growth of operators.

Operators - The Most Powerful "Transit Station" for Computing Power

The computing power market is full of vitality.

In the first half of this year, the vitality of the technology and communication industries mainly came from AI, large models, and data elements markets. And all of this is inseparable from cloud computing. Therefore, the demand in the cloud computing market has been ignited.

This has also prompted the three major operators to transform their capital expenditures: in 2023, the three major operators will all tilt their capital expenditures towards the field of computing power networks.

China Telecom: In the first half of the year, it completed a cumulative capital expenditure of 41.6 billion yuan, and its planned capital expenditure for the whole year is 99 billion yuan. China Telecom has particularly strengthened its investment in computing power. The addition of intelligent computing power reached 1.8 EFLOPS, reaching 4.7 EFLOPS, an increase of 62%; the expansion of general-purpose computing power added 0.6 EFLOPS, reaching 3.7 EFLOPS, an increase of 19%. The number of IDC external racks reached 534,000, with more than 70% of the new racks deployed in hotspots.

China Unicom: In the first half of the year, it completed a cumulative capital expenditure of 27.59 billion yuan. In 2023, it will further expand its investment, and the level of capital expenditure will reach 76.9 billion yuan, with the proportion of investment in computing power networks exceeding 19%. The scale of IDC racks exceeds 380,000, and the average delay of the backbone network remains at the industry-leading level.

China Mobile: In the first half of the year, it completed a cumulative capital expenditure of 81.4 billion yuan, a decrease compared to the first half of 2022. The planned capital expenditure for the whole year is 183.2 billion yuan. The number of C racks reached 478,000, with a net increase of 11,000 racks. The cumulative production of computing power servers exceeded 804,000 units, with a net increase of over 91,000 units. The scale of computing power reached 9.4 EFLOPS.

All three major operators have publicly stated that their capital expenditures on computing power network construction will continue to increase, but China Unicom's investment progress is significantly lagging behind.

Wu Chao, the director of CITIC Securities Research Institute, once publicly stated, "Every drop of water flowing into large models eventually flows into computing power."

Operators are important participants in the deployment of computing power networks and the data element market ecosystem, and they will fully benefit from the rapid development of the industry. This will become the most clear direction for business growth in the next decade for operators.

Previously, we discussed in detail in an article The "Data Elements" Imagination of China Mobile.

Layout of Large Models, Unwilling to Fall Behind

This year, the large model track is the most attractive, and operators are inevitably unwilling to fall behind.

In the industry, computing power is seen more like someone selling shovels, with higher certainty of prosperity but less flexibility compared to large models.

Regarding this, Yang Jie from China Mobile stated, "For the application of AI, the future direction is mainly focused on B2B, developing specialized large models for applications and industries."

China Mobile: Currently, it has the Jiutian·HaiSuan government affairs large model, Jiutian·Customer Service large model, and the "Jiutian" artificial intelligence platform, forming an industrial-level intelligent service capability from algorithms, platforms, capabilities to scaled applications.

China Unicom: It has language large models and Honghu graphic and text large models. Honghu graphic and text large models currently have two versions, with 800 million training parameters and 2 billion training parameters, respectively. This means that the large model has powerful computing and learning capabilities, enabling various complex graphic and text processing functions.

China Telecom: It has the Xinghe general vision large model 2.0 and the TeleChat generative semantic large model. In the future, the focus will be on providing AI services across the entire website, relying on Tianyi Cloud to carry artificial intelligence products.

Telecom operators themselves are demanders of large AI models and highly complex AI capabilities.

The three major telecom operators have chosen to bypass the general large model field and combine their own data advantages and user bases, using the technological transformation brought by large models to strategically layout in the field of AI communication, making existing services more intelligent.

This initiative not only enhances the operational capabilities and network efficiency of the operators themselves but also represents a wise strategic choice.