Tencent Reportedly Leases 100,000 AI Chips From Oracle

China’s Tencent has signed a reported five-year agreement with US cloud computing company Oracle to access about 100,000 advanced Artificial Intelligence chips, highlighting the company’s accelerating investment in AI infrastructure.

The Financial Times reported that the deal is Tencent’s largest overseas lease agreement and is estimated to be worth about $7 billion. The arrangement would give Tencent access to advanced AI Chips housed across multiple Oracle data centers in Southeast Asia, according to people familiar with the matter. Reuters said it could not independently verify the report, while Oracle and Tencent did not immediately respond to requests for comment. 0

Tencent’s Reported $7 Billion Oracle Deal

Under the reported agreement, Tencent would lease computing capacity from Oracle for five years across data centers in Southeast Asia. The arrangement is designed to provide access to approximately 100,000 advanced AI chips that are not available to Tencent in China.

The reported value of the agreement is about $7 billion, with an upfront payment of approximately 30%, according to the Financial Times. The scale of the transaction reflects the enormous amount of computing power technology companies now require to train and deploy increasingly sophisticated AI models. 1

The deal also illustrates how major Chinese technology companies are looking beyond domestic infrastructure as they compete to expand their AI capabilities.

Why Tencent Needs More AI Computing Capacity

Tencent has been increasing its investment in artificial intelligence as it develops new large language models, AI agents and consumer-facing products.

The company is integrating AI into several parts of its technology ecosystem, including its widely used WeChat platform. It is also developing AI tools for consumers, creators and businesses, creating demand for significant computing resources.

The Financial Times reported that Tencent’s second-quarter capital expenditure increased 176% year over year to 53 billion yuan, or about $7.9 billion. The spending included AI infrastructure and prepayments for computing resources. 2

The reported Oracle agreement therefore forms part of a much broader expansion in Tencent’s AI infrastructure spending.

Advanced Chips Are Difficult to Access in China

One of the main factors behind the reported overseas arrangement is the limited availability of some advanced AI processors in China.

US export restrictions have limited Chinese companies’ access to certain advanced semiconductor technologies. At the same time, Beijing has encouraged technology companies to develop and use domestic alternatives.

The result has been a race among Chinese technology companies and AI developers to secure sufficient computing capacity for training and deploying AI systems.

The reported Oracle arrangement would give Tencent access to advanced processors through data centers located outside China, allowing the company to obtain computing resources that are more difficult to access domestically. The Financial Times reported that such leasing arrangements can provide access to advanced processors that are not available in China. 3

Southeast Asia Becomes an Important AI Infrastructure Hub

The reported agreement also highlights the growing importance of Southeast Asia as a location for data centers serving the expanding AI industry.

Oracle operates data center infrastructure in the region, and Chinese technology companies have increasingly looked to overseas facilities to obtain additional computing capacity.

According to the Financial Times, ByteDance and Alibaba remain among the largest data-center customers in Southeast Asia, while Tencent is increasing its presence through the reported Oracle arrangement. 4

The trend reflects the geographic expansion of AI infrastructure as companies seek access to electricity, data centers and advanced computing hardware.

Tencent’s AI Push Extends Beyond Large Language Models

Tencent’s AI strategy involves more than simply developing large language models. The company is also working on AI agents capable of performing tasks within its broader digital ecosystem.

The company has been developing its Hunyuan family of AI models and has been incorporating AI capabilities into products used by consumers and businesses.

The Financial Times reported that Tencent is rolling out Xiaowei, an AI agent integrated into WeChat that can execute tasks across the platform’s mini-program ecosystem using natural-language instructions. 5

Tencent has also developed WorkBuddy, an AI-powered productivity and office assistant aimed at the enterprise and workplace market.

Tencent Releases New AI Image-Generation Technology

The company’s push into AI also includes generative tools for professional creators. Tencent recently released a preview version of a new AI image-generation model offering text-to-image and image-to-image capabilities.

Such applications require significant computing resources during both model development and deployment. As Tencent expands the number of AI products it offers, demand for high-performance processors is expected to remain an important part of its infrastructure strategy.

AI Infrastructure Spending Is Surging

The reported Tencent-Oracle agreement comes during a period of exceptionally rapid investment in AI infrastructure worldwide.

Technology companies are committing billions of dollars to data centers, processors, networking equipment and electricity capacity as demand for AI services increases.

The Financial Times reported that Tencent’s capital spending jumped sharply in the second quarter as the company invested in AI infrastructure and computing resources. The spending increase demonstrates the scale of investment required to compete in the current AI market. 6

Other major technology companies have also announced large infrastructure commitments, making access to computing capacity a central issue in the global AI industry.

How US Export Controls Affect Chinese AI Development

US restrictions on advanced semiconductor exports have become an important factor in the development of China’s AI industry.

Chinese technology companies are attempting to expand domestic chip production while also finding ways to secure computing resources through overseas infrastructure. The Financial Times reported that tightening US export controls and shortages of domestic supply have encouraged Chinese technology groups to seek additional computing capacity outside China. 7

These restrictions have also contributed to a broader push by Chinese companies to develop alternatives to US semiconductor technology.

The situation has created a complex environment in which Chinese companies are simultaneously investing in domestic technology and using overseas infrastructure where permitted.

Oracle Gains From Growing Demand for AI Computing

For Oracle, the reported agreement demonstrates the growing demand for its cloud infrastructure and data center capacity.

Oracle has increasingly positioned its cloud business as an important provider of AI computing resources. The company’s data centers can provide customers with access to large quantities of specialized computing capacity without requiring them to build and operate all the infrastructure themselves.

Long-term leasing arrangements can also give cloud providers greater visibility into demand for computing infrastructure while allowing customers to secure capacity for major AI projects.

However, the rapid expansion of AI infrastructure is also creating substantial financial and operational commitments across the technology industry.

Tencent’s AI Investment Comes With Higher Costs

The scale of Tencent’s infrastructure spending has financial implications. The Financial Times reported that the Oracle agreement contributed to Tencent recording negative quarterly free cash flow of 13.8 billion yuan, or about $2 billion, in the second quarter. It was reportedly the company’s first negative quarterly free-cash-flow figure in more than a decade. 8

This illustrates the significant upfront cost associated with securing AI computing capacity.

Technology companies are spending heavily before the full financial returns from AI products are established. The long-term success of these investments will depend partly on whether AI services generate enough revenue to justify the infrastructure costs.

Competition Among Chinese Technology Companies

Tencent is competing with several major Chinese technology companies and AI developers as the country’s AI industry expands.

Alibaba, ByteDance and other major technology groups are also investing heavily in computing infrastructure, AI models and applications.

The Financial Times reported that Tencent’s Hunyuan models have improved significantly and are narrowing the gap with other leading domestic AI models. The company is also expanding AI agents and productivity tools as it seeks to turn its AI investments into products used by consumers and businesses. 9

Access to computing capacity is therefore becoming an important competitive factor alongside model quality, software development, talent and access to data.

What the Reported Deal Means for Tencent’s AI Strategy

If the reported agreement proceeds as described, access to approximately 100,000 advanced chips would provide Tencent with a substantial increase in available computing resources.

The five-year structure would also give the company longer-term access to infrastructure rather than requiring it to build all of the necessary computing capacity itself.

For Tencent, the arrangement could support continued development of large AI models, AI agents and other generative AI applications. At the same time, the size of the reported commitment underscores the financial resources required to compete at the leading edge of the AI industry.

Key Points

• Tencent reportedly signed a five-year Oracle lease covering about 100,000 advanced AI chips.
• The Financial Times estimated the reported deal at approximately $7 billion, with about 30% paid upfront.
• The chips are housed in Oracle data centers across Southeast Asia and are reportedly not available to Tencent in China.
• Tencent is sharply increasing AI infrastructure spending as it develops models, AI agents and generative AI products.

Frequently Asked Questions

How many AI chips is Tencent reportedly leasing from Oracle?

The Financial Times reported that Tencent is gaining access to approximately 100,000 advanced AI chips through the agreement.

How much is the reported Tencent-Oracle deal worth?

The reported five-year lease is estimated to be worth about $7 billion, with an upfront payment of approximately 30%.

Where are the leased chips located?

The computing capacity is reportedly located across multiple Oracle data centers in Southeast Asia.

Why is Tencent using overseas AI infrastructure?

Access to some advanced AI processors is restricted in China, while domestic supply remains limited. Overseas data centers can provide additional computing capacity where permitted.

What is Tencent investing in AI for?

Tencent is developing large language models, AI agents, image-generation tools and productivity applications while integrating AI into its broader digital ecosystem.

What is Tencent’s Hunyuan AI?

Hunyuan is Tencent’s family of AI models that the company is developing for applications including generative AI and other intelligent services.

How much did Tencent’s capital expenditure increase?

The Financial Times reported that Tencent’s second-quarter capital expenditure increased 176% year over year to 53 billion yuan, including spending on AI infrastructure and computing resources.

Did Reuters independently verify the reported deal?

No. Reuters said it could not immediately verify the Financial Times report. Oracle and Tencent also did not immediately respond to Reuters’ requests for comment.

“`10