Tencent adds fuel to domestic computing power

Author | Huang Yu

Editor | Wang Xiaojun

In the past year, the application of AI large models has been accelerating. The surging wave of AI is exponentially driving global computing power demand. In this global AI arms race, those who can provide usable computing power will become one of the winners at the current table.

At Tencent's annual most significant To B industry event—the Tencent Global Digital Ecosystem Conference—Tencent's deployment in computing power has also attracted widespread attention.

Tencent Cloud announced that it has fully adapted to mainstream domestic chips. At the same time, the collaborative full-stack optimization of software and hardware is Tencent Cloud's long-term strategic investment, integrating different types of chips through a heterogeneous computing platform to provide high-cost performance AI computing power externally.

In other words, although focusing on software, Tencent also aims to build a solid AI infrastructure. Meanwhile, against the backdrop of the continuous strengthening of the logic of domestic chip substitution, Tencent's comprehensive adaptation undoubtedly adds fuel to the domestic computing power.

Of course, the AI large model competition is a marathon. Tencent's more core advantage in AI lies in its extensive applications. Therefore, for a long time after the wave of AI large models surged, Tencent is not in a hurry to seize the AI-related market but prioritizes ensuring the self-research of large models and empowering its massive business.

After more than two years of efforts, AI technology has become the core growth engine for Tencent and Tencent Cloud, and the positive cycle of AI investment and high-quality growth continues to strengthen. Against this backdrop, since the beginning of this year, Tencent has also shifted from defense to offense in its AI strategy, revealing a more proactive and open Tencent.

AI is also bringing value reassessment to Tencent. Tencent's stock price recently hit a nearly four-year high, and returning to the peak of a 7 trillion market value seems not far off. However, this will not be the end; the market expects AI to recreate Tencent, bringing it greater imaginative space.

Computing Power Revenue Growth

Three years ago, Tencent Cloud proposed, "I would rather have 150 pounds of muscle than 200 pounds of fat." After three years of transformation, focusing on self-research and promoting integration, Tencent Cloud is now making a name for itself within the entire Tencent Group.

Tang Daosheng, Senior Executive Vice President of Tencent Group and CEO of the Cloud and Smart Industry Group, revealed that by the second quarter of 2025, Tencent's To B revenue achieved double-digit growth reaching 55.5 billion yuan; the scale of Tencent Cloud's overseas customers doubled year-on-year in the past year.

"I can proudly say that after these years of continuous focus on core products and technologies, reducing fat and increasing muscle, Tencent Cloud has become a highly competitive 'robust contender' in the cloud market."

Undoubtedly, AI has become a strong driving force for Tencent Cloud's growth. In the second quarter of this year, Tencent Cloud's revenue growth rate accelerated compared to the previous few quarters, mainly due to the increased demand from enterprise customers for AI-related services, including GPU leasing and API token usage, as well as the growth of merchant technical service fees Computing power is indeed a good business. Tang Daosheng stated to Wall Street News that the AI wave undoubtedly drives the demand for Tencent Cloud AI Infra and that the growth related to AI infrastructure will become more apparent in the revenue generated from AI-related businesses in the future.

Tencent's investment in computing power is also significant. According to Tencent's financial report, since the strategic acceleration in the fourth quarter of last year, Tencent's cumulative capital expenditure has reached 83.16 billion yuan by the second quarter of this year, mainly used to support the development of AI-related businesses.

At the performance meeting in August, Tencent's management also mentioned that Tencent has enough chips to train and continuously upgrade existing models. In addition, Tencent is enhancing inference efficiency through extensive software optimizations, allowing it to handle more workloads on the same number of chips.

Li Qiang, Vice President of Tencent Group and President of Enterprise Business, pointed out that as a cloud computing provider, it is essential to build a larger and more abundant computing power support under the backdrop of AI+, which is the infrastructure that will benefit various industries. In the second half of this year, Tencent will continue to invest, including building a relatively large AI Infra globally.

Fully adapting to domestic chips is undoubtedly an important step for Tencent Cloud to enhance its AI Infra.

Tang Daosheng explained that Tencent is collaborating with multiple chip manufacturers for adaptation. In fact, there are many types of models, both large and small. Different scenarios require different chip configurations, so Tencent will continue to work with various manufacturers.

In other words, when some tech giants personally venture into chip manufacturing, Tencent chooses to focus on the software part, the model part, and digital technology, collaborating with various chip manufacturers with an open mindset.

At this conference, Tencent Cloud's intelligent computing underwent a significant upgrade towards "AI Infra closer to Agent," establishing a complete solution that includes Agent Infra solutions + Cloud Mate intelligent agent services and full-link security capabilities, helping Agentic AI transition from "laboratory" to "production-level."

Li Li, Vice President of Tencent Cloud, stated that as the number of agents explodes, the demand for cloud computing infrastructure from enterprises will rapidly upgrade from "resource supply" to "business value." AI Infra infrastructure must possess faster inference efficiency, more flexible tool integration, more reliable system guarantees, and more automated service capabilities.

It is reported that the core advantage of Tencent Cloud's intelligent computing system lies in "same source and same structure," eliminating the need to build AI-exclusive storage and other facilities. Relying on the cloud-native capabilities honed in the past, it can meet intelligent computing demands and achieve deep collaboration between cloud and intelligent computing capabilities.

AI Landscape

Tencent Cloud is Tencent's window for exporting AI capabilities. At this global digital ecosystem conference, Tencent Cloud also announced that it will fully open its accumulated technological deposits and AI practices while connecting the technology and scenarios of both C-end and B-end.

This is undoubtedly a milestone moment for Tencent Cloud. Last year, Tencent Cloud clearly stated that AI, going global, and integrated innovation would be the main channels for seeking incremental growth in the future. At this conference, AI and going global became two core themes, with more achievements presented to everyone Regarding the highly anticipated AI, Tencent has currently built a full-link product matrix around large models, including underlying infrastructure, self-developed large models, model development platforms, intelligent agent development platforms, and diverse intelligent applications aimed at various scenarios, helping enterprise clients quickly implement large models in their scenarios.

At this conference, Tencent also launched the "Tencent Cloud Intelligent Agent Strategic Panorama," as well as the Tencent Cloud Intelligent Agent Development Platform ADP 3.0, the Agent Infra solution Agent Runtime, Tencent Cloud Expert Service Intelligent Agent Cloud Mate, and the Hongyuan 3D 3.0 base model.

It is worth mentioning that at the beginning of this year, Tencent classified many AI products and applications aimed at the C-end, such as Yuanbao, QQ Browser, and ima, into Tencent Cloud, which has also taken on the responsibility of creating AI products for both B-end and C-end.

As a pioneer of Tencent's C-end AI native applications, Yuanbao has rapidly upgraded its capabilities around search, document understanding, and image processing this year. At this conference, Yuanbao also took the lead alongside Tang Daosheng, showcasing its Q&A capabilities on site.

According to data provided by Tencent, the number of questions users ask Yuanbao daily has reached the total amount for a month at the beginning of this year, and Yuanbao has become one of the top three AI native applications in terms of daily active users in China.

In addition, Yuanbao continues to deeply integrate with Tencent's existing applications, now connecting with over 10 applications, including Tencent Meeting, Tencent Documents, WeChat, Tencent Video, and QQ Music.

Tang Daosheng pointed out that after the organizational restructuring at the beginning of this year, Tencent Cloud has had greater room for development and more capabilities, allowing it to connect the C-end and B-end, making products on both sides more competitive. "So, we are still very excited."

However, Tang Daosheng also told Wall Street News that ToC AI native applications are still in the investment phase, and Tencent Cloud is currently not deeply considering commercialization; the focus at this stage is still on refining product and service experiences. For example, Yuanbao and ima will focus more on enhancing user experience and expanding scenario coverage, allowing users to solve practical problems more naturally and efficiently during use.

With ongoing investments, the landscape of Tencent's AI applications is fully opening up from B-end to C-end. At the same time, AI has become Tencent's "new business gene," driving strong growth in core businesses such as gaming and advertising.

The capital market has also given a warm response to Tencent's AI strategic achievements: pushing its market value to briefly return to HKD 6 trillion on September 17, reaching a nearly four-year high.

Global tech giants have joined the ranks of reshaping their businesses with AI, and the "AI" factor has become an important indicator for the capital market to reassess a tech company. Although AI commercialization is still in its early stages, Tencent has clearly succeeded in positioning itself