Tencent is reportedly making a major move to secure the computing power required for its growing artificial intelligence ambitions. This comes amid reports of a Tencent Oracle AI chip deal designed to strengthen its tech infrastructure.
The Chinese technology company has signed a reported five-year agreement with Oracle worth about $7 billion, giving it access to around 100,000 advanced AI chips located in Oracle data centres across Southeast Asia, according to the Financial Times. If confirmed, the arrangement would represent Tencent’s largest overseas chip-leasing deal with a US cloud provider.
The reported agreement comes as Tencent expands its AI models and adds artificial intelligence features across its consumer and enterprise products. At the same time, restrictions on advanced semiconductor exports to China have made access to high-end AI computing increasingly complicated.
Tencent Reportedly Signs $7 Billion Oracle Agreement
The reported deal would give Tencent access to approximately 100,000 AI chips through Oracle infrastructure in Southeast Asia. Rather than importing the processors into China, Tencent would use computing capacity hosted outside the country, according to the Financial Times.
The agreement reportedly runs for five years and carries a total value of about $7 billion. Around 30% of the reported value would be paid upfront. Neither Tencent nor Oracle had publicly confirmed the arrangement when Reuters reported on the Financial Times story, and Reuters said it could not independently verify the report.
If the agreement goes ahead as reported, it would give Tencent a large and relatively predictable source of overseas AI computing capacity.
Why Tencent Needs More AI Compute
Tencent’s AI ambitions extend well beyond developing another chatbot. The company is incorporating artificial intelligence across its enormous digital ecosystem while continuing development of its Hunyuan family of models.
Training sophisticated models requires substantial computing resources. Serving those models to users creates another demand altogether, particularly when AI features become embedded inside widely used consumer and business services.
That makes access to processors a strategic issue for Tencent. More computing capacity gives its engineering teams greater room to train models, run experiments and support AI applications at scale.
The reported Oracle arrangement would provide that capacity without requiring Tencent to physically bring the advanced processors into mainland China.
Overseas Data Centres Take on Greater Importance
The location of the reported chips is one of the most significant parts of the agreement. US export controls have restricted Chinese companies’ ability to obtain some advanced AI processors, making overseas computing infrastructure increasingly relevant to companies building large AI systems.
Under the reported Oracle arrangement, the processors would remain in data centres outside China while Tencent accesses their computing power. This differs from directly purchasing advanced chips and importing them into the Chinese market.
Southeast Asia has become an increasingly important destination for data-centre investment, supported by growing cloud, digital-services and AI demand. A major Tencent commitment would add another layer to the region’s role in the global AI infrastructure market.
Tencent Pushes Hunyuan and AI Products
Tencent has been investing heavily in its Hunyuan AI model family as Chinese technology companies compete to develop increasingly capable systems. The company has also been integrating AI into products rather than keeping its development efforts confined to research environments.
Its AI strategy includes applications for consumers, businesses and creators. Tencent has also been developing AI agents and other tools that could operate across parts of its broader product ecosystem.
That expansion creates a simple infrastructure problem: the more AI products Tencent launches, the more computing power it needs to operate them.
The reported Oracle agreement would give the company a significant pool of overseas computing resources to support that expansion.
Southeast Asia Becomes an AI Infrastructure Battleground
The reported Tencent deal also highlights the growing importance of Southeast Asia in the AI infrastructure economy. The region is attracting data-centre investment from global cloud companies while Chinese technology groups seek computing capacity outside mainland China.
Other major Chinese technology companies, including Alibaba and ByteDance, have also been linked to substantial overseas computing requirements.
For cloud providers, Chinese technology companies represent significant potential demand. For those technology companies, overseas data centres can provide access to computing resources that may be harder to secure domestically.
The result is a more geographically distributed AI infrastructure market, with data centres becoming strategic assets in their own right.
A Five-Year Commitment to AI Computing
The reported five-year term is important because it points to a longer infrastructure strategy rather than a temporary response to a chip shortage. AI development requires continuing investment in both model training and inference, meaning companies need computing capacity long after a model reaches production.
Tencent’s reported commitment of around $7 billion would therefore represent a substantial infrastructure expense spread across several years.
The structure also reflects how AI companies are changing the way they acquire computing power. Instead of simply purchasing processors, companies can lease access to entire pools of hardware through cloud infrastructure providers.
For Tencent, that approach could provide flexibility while avoiding some of the logistical and regulatory complications associated with moving advanced processors directly into China.
The AI Race Is Becoming a Compute Race
The Tencent-Oracle report shows how the AI competition is increasingly tied to infrastructure. Model architecture and software remain critical, but access to high-performance processors can determine how quickly companies train models, launch products and serve users.
A reported 100,000-chip commitment would put Tencent among the technology companies making very large infrastructure bets as the demand for AI computing continues to rise.
The deal also illustrates the changing geography of artificial intelligence. The companies developing AI may be based in one country, their customers may be somewhere else, and the processors running their models may sit in data centres thousands of kilometres away.
That is becoming a defining feature of the AI economy.
Sources
- Financial Times — China’s Tencent leases 100,000 chips from Oracle to accelerate AI push
Read the Financial Times report - Reuters — China’s Tencent leases 100,000 chips from Oracle to accelerate AI push, FT reports
Read the Reuters report - The Straits Times — Tencent leases 100,000 chips from US tech firm Oracle to accelerate AI push
Read The Straits Times report
