Chinese tech giants reveal how they're dealing with U.S. chip curbs to stay in the AI race

Core Viewpoint - Tencent and Baidu are adapting their strategies to maintain competitiveness in the global AI landscape despite U.S. semiconductor export restrictions, focusing on stockpiling chips, optimizing AI models, and leveraging domestic semiconductor capabilities [1][2][12]. Tencent's Approach - Tencent has a strong stockpile of graphics processing units (GPUs), which are essential for training AI models, and believes it can achieve effective training results with fewer chips than American companies suggest [3][4]. - The company is enhancing efficiency through software optimization, allowing it to utilize existing GPUs more effectively for AI tasks, and is exploring smaller models that require less computing power [5][6]. Baidu's Strategy - Baidu emphasizes its "full-stack" capabilities, integrating cloud computing, AI models, and applications, which allows it to deliver value even without the most advanced chips [7]. - The company is also focusing on software optimization to reduce operational costs and improve GPU utilization, which it considers a competitive advantage [8]. - Baidu highlights the progress of domestic Chinese technology firms in developing AI semiconductors, which could mitigate the impact of U.S. chip restrictions [10]. Domestic Semiconductor Development - China is increasing its focus on developing a domestic semiconductor ecosystem, although it still lags behind the U.S. in GPU and AI chip technology [11]. - Analysts note that stockpiling chips is one strategy Chinese companies are employing to navigate export restrictions, and there have been advancements in semiconductor technology within China [12]. - The development of self-sufficient chips and efficient software stacks is seen as a foundation for long-term innovation in China's AI ecosystem [10].