Workflow
高盛:中国AI基础设施叙事重燃 企业级大模型Token消耗量暴涨 阿里(09988)是核心受益者
智通财经网·2025-09-17 03:12

Core Insights - The demand for enterprise-level applications is driving a resurgence in investment narratives for AI infrastructure in China, with Alibaba positioned favorably due to its market share and capabilities [1][5]. Group 1: AI Infrastructure Investment - Goldman Sachs reports that the daily token consumption for enterprise-level large models in China reached 10.2 trillion in the first half of 2025, a staggering increase of 363% compared to the second half of 2024 [2]. - The capital expenditure for Chinese cloud service providers (CSPs) is projected to grow by 39% year-on-year in Q3 2025, providing a hardware foundation for sustained AI cloud revenue growth [4]. - The diversification of chip supply strategies is reshaping the development prospects of China's AI cloud industry, reducing reliance on overseas chip supplies [4]. Group 2: Market Position and Growth Potential - Alibaba holds a 47% share of the Chinese public cloud market and has successfully penetrated most of the Fortune 500 companies deploying generative AI in China, making it the leading choice among model providers [3][5]. - Goldman Sachs raised Alibaba's target price from $163 to $179, maintaining a "buy" rating, and increased the growth expectations for Alibaba Cloud from 28%-30% to 30%-32% for Q2 to Q4 of FY2026 [1][4]. Group 3: Competitive Landscape and Innovations - Major players like Alibaba, ByteDance, and DeepSeek are the top choices for enterprises selecting general large models, with Alibaba leading in the market [3]. - Recent advancements in AI models include Alibaba's Qwen3-Next, which boasts ten times the performance of its predecessor at a fraction of the cost, and Baidu's ERNIE X1.1, which competes with global models like GPT-5 [6]. - AI is increasingly being integrated into mainstream applications, with companies like Meituan and Alibaba introducing AI assistants to enhance user experiences [6]. Group 4: Commercialization Challenges - Despite rapid advancements in AI models and applications, the commercialization of AI in China still faces significant challenges, with the annual recurring revenue (ARR) for Chinese AI applications at only $1.5 billion, representing 5% of the global total of $30 billion as of August 2025 [7].