Group 1 - The core point of the article highlights the performance of the Sci-Tech Chip ETF Index (588920), which opened at a decline of 0.46% at 1.520 yuan on September 24 [1] - Major stocks within the Sci-Tech Chip ETF Index showed mixed performance, with SMIC down 1.27%, Haiguang Information up 0.52%, Cambrian down 1.51%, and others like Zhongwei Company down 2.81% [1] - The performance benchmark for the Sci-Tech Chip ETF Index is the Shanghai Stock Exchange Sci-Tech Board Chip Index, managed by Penghua Fund Management Co., Ltd., with a return of 52.79% since its establishment on July 16, 2025, and a return of 16.96% over the past month [1]

Group 1 - The development of OCS technology is accelerating, with Google deploying thousands of OCS optical switches for AI training, benefiting companies in the supply chain such as Tengjing Technology and Dekeli [1] - CITIC Electronics emphasizes that the demand for etching equipment is driven by upgrades in storage and logic chip technologies, with an expected increase in domestic production rates, involving semiconductor equipment manufacturers like Zhongwei Company [1] - OpenAI plans to increase investments in AI chips and data centers, with Luxshare Precision receiving hardware cooperation orders and expanding its AI server support, including optical modules and connector products [1] Group 2 - Guosheng Securities identifies three key logics in the AI storage sector, covering storage modules (Shannon Chip Creation, Jiangbolong, etc.), storage chips (Zhaoyi Innovation, Lanke Technology, etc.), semiconductor equipment (Zhongwei Company, Beifang Huachuang, etc.), and materials (Yake Technology, Dinglong Co., etc.) [2] - Zheshang Securities highlights that the electronics industry will benefit from breakthroughs in AI cloud and edge computing by 2025, focusing on the premium expectations for HBM4 in the storage sector and the expansion of customized storage application scenarios [2] - Both institutions focus on the technology upgrades and domestic substitution trends in the semiconductor industry chain, particularly in storage, manufacturing, and packaging related to AI [2] Group 3 - Related products include the Sci-Tech Chip ETF Index (588920), Semiconductor ETF (159813), Big Data ETF (159739) [3] - Related stocks include Haiguang Information (688041), Cambricon-U (688256), SMIC (688981), Lanke Technology (688008), Zhongwei Company (688012), Chip Origin (688521), Hushi Silicon Industry (688126), Dongxin Co. (688110), Huahong Company (688347), and Huahai Qingke (688120) [3]

Group 1 - Cambricon's Q2 revenue increased by 4425% year-on-year to 1.769 billion yuan, with a record high net profit margin of 38.57%, indicating growth potential driven by the large-scale application of the Siyuan 590 chip in model training and other fields [1] - The penetration rate of liquid cooling technology is expected to rise from 5% to 35-50%, with cold plate technology dominating the market, as Chinese manufacturers leverage price advantages to expand secondary pipeline supply, driven by high-performance chip demand [1] - South Asia's CCL series has increased prices by 8%, with AI driving demand for high-speed copper-clad laminates, leading to overseas breakthroughs for companies like Shengyi Technology [1] Group 2 - Huachuang Securities highlights that the surge in AI computing power demand is accelerating the iteration of advanced packaging technology, with TSMC's CoWoS platform achieving ultra-high-density interconnection through 2.5D packaging, becoming a mainstream configuration for AI chips from companies like NVIDIA and AMD [2] - Huajin Securities focuses on the synergy between the storage industry chain and advanced packaging, noting that companies with dual capabilities are strengthening their market competitiveness in embedded storage and enterprise-level storage through main control chip design and wafer-level packaging technologies in the AI era [2] - Related products include the Sci-Tech Chip ETF Index (588920), Semiconductor ETF (159813), Big Data ETF (159739), Intelligent Connected Vehicle ETF (159872), and Robotics ETF (159278) [2]

Group 1 - DeepSeek launched MKFP8+UE8M0 technology, enhancing computing efficiency through optimized FP8 format, benefiting domestic chip manufacturers like Cambricon and Haiguang [1] - Huawei's next-generation chip 910x is expected to support FP8 precision, with Cambricon 690 and other domestic chips already adapting to this technology, highlighting Huawei's software ecosystem advantages [1] - CICC analysis indicates that DeepSeek V3.1's UE8M0 FP8 format can dynamically optimize computing resources, alleviating the computing power bottleneck of domestic AI chips, with positive expectations for the domestic computing chain in the second half of the year [1] Group 2 - Aijian Securities notes that the computing chip market is undergoing a restructuring opportunity, with ASIC, FPGA, CPU, and GPU forming a multi-layered computing system based on performance and application differentiation [2] - ASICs are gaining an energy cost advantage in AI edge inference scenarios due to low power consumption and high customization, while FPGAs offer flexibility for small to medium-scale applications [2] - Northeast Securities focuses on storage chips, emphasizing that Lianyun Technology is driving storage controller technology iteration through PCIe Gen4 mass production and AIoT chip extension, forming a core for edge AI hardware solutions [2]

Group 1 - The current market is driven by incremental capital, with limited impact from fund redemption pressure on growth sectors like electronics, suggesting a potential shift towards public offerings and institutional dominance, which supports technology sectors such as semiconductor chips [1] - The expectation for domestic computing power has been reiterated, alongside discussions on emerging applications like 3D printing and liquid cooling technology, which may boost demand expectations in the chip industry [1] - The results of the 301 investigation concerning China's mature chips are expected to be announced by September 23, with potential market volatility due to technology controls and tariff policies, although short-term pressure is manageable as the Trump administration plans to suspend a 24% equivalent tariff until November 10 [1] Group 2 - The AI chip ecosystem at the edge consists of edge SoC, storage chips, sensor chips, and smart modules, with storage chips facing challenges in capacity, speed, and power consumption; Lianyun Technology has developed a product line from SATA to PCIe 4.0, targeting lightweight laptops and pre-installed markets [2] - The emphasis on semiconductor self-sufficiency continues, with ongoing iterations of domestic AI computing power chips and enhanced supply chain capabilities; the analog IC sector is experiencing volume growth with stable prices, and automotive-grade products are entering a replacement window, indicating direct support for the semiconductor sector [2] Group 3 - Related products include various ETFs such as the Sci-Tech Chip ETF Index (588920), New Energy ETF (159261), Semiconductor ETF (159813), and others [3] - Key associated stocks include Cambrian (688256), Haiguang Information (688041), and others within the semiconductor industry [3]