SINOLINK SECURITIES·2025-08-24 11:56Investment Rating - The report suggests a positive outlook for the AI chip sector, particularly in the context of domestic developments and overseas demand for servers and optical modules [4]. Core Insights - Domestic AI development is driving growth in sectors such as servers and IDC, while overseas AI advancements are boosting demand for servers and optical modules [4]. - The launch of DeepSeek-V3.1 is expected to facilitate the large-scale deployment of domestic AI chips, enhancing the competitive landscape [1][2]. - The liquid cooling market is anticipated to expand significantly, with companies like Invec benefiting from strong demand in AI data centers [1][12]. Summary by Sections Communication Sector - Telecom business revenue reached 905.5 billion yuan in the first half of the year, showing a year-on-year growth of 1% [3][14]. - The number of fixed broadband users reached 684 million, with gigabit users accounting for 33% [20]. Server Sector - The server index increased by 8.30% this week and 17.28% this month, with Industrial Fulian reporting improved efficiency and yield for its GB200 and GB300 models [2][6]. - GB300 is expected to become a significant profit driver for the company's AI server business next year [6]. Optical Module Sector - The optical module index rose by 4.34% this week and 14.59% this month, driven by Google's adoption of OCS technology [2][9]. - Coherent and Lumentum reported their first revenue from OCS orders, indicating market expansion [9]. IDC Sector - The IDC index increased by 6.78% this week and 14.41% this month, with DeepSeek-V3.1's release expected to alleviate the shortage of domestic AI inference cards [2][12]. Liquid Cooling Sector - Invec reported a revenue of 2.573 billion yuan in the first half of 2025, a year-on-year increase of 50.3%, driven by strong demand for liquid cooling solutions [12][52]. - The establishment of the UQD alliance aims to enhance the reliability and interchangeability of liquid cooling connectors, reducing operational complexity [12].