Tencent hopes that AI can help extend the growth runway of its advertising business. The chip inventory is sufficient to support the training of future generations of models.

At the quarterly performance conference call on the evening of May 14, 2025, Tencent President Liu Chiping said that Tencent is increasing its investment in AI in the form of capital expenditures and operating expenses. Some GPU and AI investments have already generated revenue for Tencent. AI optimizes targeted advertising, driving advertising revenue growth. Tencent will seize this opportunity and believes that the value space in this area will be even greater in the future. Tencent hopes to extend the growth runway of its advertising business. Liu Chiping responded that Tencent had previously purchased a considerable number of chips, which was very useful for executing its AI strategy. Based on the allocation of these chips, they will obviously be prioritized for applications that can bring immediate returns to Tencent, such as advertising and content recommendation products. Secondly, they will be used for large-scale model training, which indeed requires higher-end chips. "But in the past few months, we have realized that we must abandon the so-called 'law of scale' by American tech companies. Now we find that even with a smaller cluster, we can achieve good training results. Moreover, in the optimization phase after training, we have a lot of potential to explore, and these processes do not necessarily require a very large cluster. Our existing inventory of high-end chips is sufficient to support the training of future generations of models."