TENCENT(HK:00700) Mei Ri Jing Ji Xin Wen·2025-05-14 23:28Core Viewpoint - Tencent's significant investment in AI is beginning to show results in its financial performance, with a focus on enhancing advertising and gaming businesses through AI capabilities [1][3]. Financial Performance - In Q1 2025, Tencent reported revenue of approximately 180 billion yuan, a year-on-year increase of 13% - Gross profit reached about 100.5 billion yuan, up 20% year-on-year - Operating profit (Non-IFRS) was 69.3 billion yuan, reflecting an 18% year-on-year growth [1]. AI Investment and Impact - Tencent is increasing its investment in AI through capital expenditures and operational costs, with some GPU and AI investments already generating revenue [1][3]. - AI is enhancing advertising targeting, leading to a 20% year-on-year growth in marketing services revenue, amounting to approximately 31.9 billion yuan [3]. - The company aims to leverage AI to improve user engagement on platforms like video accounts, potentially increasing user time spent and driving revenue growth without necessarily increasing ad load [3]. WeChat Ecosystem and E-commerce - Tencent is restructuring its WeChat business group, establishing an independent e-commerce product department to enhance operational efficiency [2]. - The company is focused on building a robust e-commerce ecosystem within WeChat, improving consumer shopping experiences, and attracting more merchants [5]. - WeChat's transaction GMV is showing positive growth, with plans to enhance user connections to products through various modules [5]. GPU and AI Strategy - Tencent has procured a significant number of chips to support its AI strategy, prioritizing their use in applications that yield immediate returns, such as advertising and content recommendation [6]. - The company has recognized the need to move away from the "scale law" of large training clusters, finding that smaller clusters can also achieve effective training results [6]. - There is a growing demand for GPU on the inference side, with potential strategies to optimize usage and meet increasing needs without solely relying on GPU procurement [6].