Tencent to Increase Capital Expenditure, Deploy More Domestic Chips in the Second Half of the Year

On May 13th, Tencent's management indicated in a conference call following the release of the first-quarter financial report that the company's capital expenditure will increase, with more domestically produced chips to be gradually put into use in the second half of this year. Tencent Cloud has previously lacked a sufficient number of GPU resources, which has been challenging to meet all external customer demands, affecting revenue generation and market share enhancement.

Tencent's management stated that with the iteration of the Hunyuan series models and their integration into internal core products and external customer scenarios, the demand for AI services is rapidly increasing. Following the release of the Hunyuan T3 preview version, Tencent reported improvements in the model's inference capabilities and intelligent entity capabilities. Currently, the Hunyuan model has been integrated into various business lines such as Yuanbao, WeChat ecosystem, advertising systems, and gaming, and has begun serving mini-program developers.

Tencent's management emphasized that AI investment includes both short-term projects and long-term investments; the company does not manage each product on a quarter-by-quarter basis but instead examines them from a portfolio and lifecycle perspective. In mature businesses such as advertising and gaming, AI investment has already brought clear revenue and profit contributions with a relatively short payback period. However, model training, fundamental capability accumulation, and free C-end services are more long-term focused and rely on continuous investment in the short term.