Group 1 - The A-share market is supported by continuous policy benefits and accelerated capital inflow, leading to a strong performance of the index [1][2] - Major public fund institutions believe that the A-share market has a stable operational foundation and reasonable valuation, providing sufficient confidence for medium to long-term growth [1][2][3] - The total market value of A-shares has surpassed 100 trillion yuan, indicating a significant rise in quality enterprises and active market participation [2] Group 2 - The A-share index has reached a near 10-year high, with structural opportunities becoming more prominent, particularly in the technology sector [2][3] - The market is expected to maintain upward support due to overall economic stabilization and structural reforms, with investors likely to increase equity asset allocation in a low-interest-rate environment [3] - Consumption and technology stocks, including sectors like innovative pharmaceuticals and semiconductors, are highlighted as key investment focuses [4] Group 3 - The AI and computing power sectors are seen as significant growth areas, with expectations for clear trends and investment opportunities emerging in the fourth quarter [5] - Investment strategies are focusing on new productivity-related fields, overseas expansion, and cost-effective consumption as the domestic economy improves [4][5]

Investment Rating - Industry rating: Recommended [5] Core Viewpoints - The overseas AI-related heat continues, with OpenAI partnering with AMD to deploy up to 6GW of AMD Instinct GPUs over the next few years, with the first batch of 1GW expected to be operational by 2026 [2][7] - The capital market shows strong enthusiasm for AI-related transactions, as evidenced by AMD's average increase of 30% [3][8] - Domestic AI applications are expected to accelerate breakthroughs, with a focus on the domestic computing power industry chain [3][9] Summary by Sections 1. Overseas AI Developments - OpenAI has launched the Sora 2 model, which features intelligent synchronization of audio and visuals, and has announced a significant increase in active users and API usage [2][8] - Samsung and SK Group have entered into a partnership with OpenAI to supply DRAM and HBM wafers for AI data centers [3][9] 2. Investment Recommendations - Focus on the domestic computing power industry chain, with notable growth from Alibaba Cloud and Huawei's super node cluster products [3][9] - Key beneficiaries include companies like Cambricon, Moore Threads, and SMIC, as well as those involved in advanced process manufacturing and chip architecture upgrades [3][9][11] - Recommendations extend to companies in the switch, optical device, PCB, and other related sectors, such as ZTE, Unisoc, and Longfly Fiber Optics [3][11] 3. Communication Sector Insights - Emphasis on two main areas: Beidou and satellite applications, and AI computing power [12] - Recommended companies include major telecom operators and equipment manufacturers, as well as third-party computing power leasing firms [12][14]

Core Insights - The recent release of multimodal models, particularly Sora2, is considered a "revolutionary" milestone in the industry, enhancing user engagement and willingness to pay for AI-generated content [1][2] Group 1: International Developments - OpenAI launched the Sora2/Pro App on October 1, supporting up to 15 seconds of text-to-video generation, achieving the top position in the US App Store within three days [1] - The developer conference on October 7 announced that ChatGPT can now directly access third-party applications, marking a shift from a single dialogue tool to an AI application and social platform [1] - xAI introduced the "Imagine" visual generation module on October 6, enhancing its capabilities in creating high-quality images and videos from text [1] - Anthropic released the Claude Sonnet 4.5 programming model on September 30, emphasizing its ability to build "production-ready" AI agents [1] Group 2: Domestic Developments - Kuaishou's Ling2.5Turbo topped the global video generation model rankings on October 2, showcasing its international leadership in video generation and content quality [2] - ByteDance partnered with UCLA on October 2 to launch Self-Forcing++ video generation technology, significantly improving visual stability [2] - Tencent released and open-sourced the mixed Yuan Image 3.0 on September 28, quickly rising to the top of the Hugging Face leaderboard [2] Group 3: Domestic Computing Power Investment Logic - The rise of domestic computing power is driven by demand from AI applications, marking a shift from supply-side to demand-side dynamics [3] - DeepSeek's release of DeepSeek-V3.2-Exp on September 30 demonstrated lower inference costs and compatibility with domestic chip ecosystems [3] - Alibaba's open-source Qwen3-VL series multimodal model, released on October 4, achieved zero-day adaptation with domestic chips, accelerating the local hardware ecosystem [3] Group 4: Investment Recommendations - Recommendations for cloud computing power include companies like Cambrian, Haiguang Information, and Chipone [4] - For edge computing power, companies such as Amlogic and Rockchip are recommended [4]

2025.10.04 本文字数：2393，阅读时长大约4分钟 作者 |第一财经 彭海斌 国庆之际，中国算力企业收到了大礼包。 中国联通和工商银行近日公布了服务器招标信息，两家企业的项目合计金额超过了100亿人民币。这些服务器的通用芯片以国产为主，工商银行30亿元海 光芯片服务器项目被浪潮信息（000977.SZ）一家企业拿下。 "主要还是国产芯片这几年技术起来了，外加政策又利好。"一位业内人士评价说。中国本土芯片企业正以肉眼可见的速度成长，它们也在市场竞争中逐步 站稳脚跟。 百亿"红包" 这次集采标包之一的中标候选人包括紫光华山、浪潮信息和超聚变三家企业；标包之二的中标候选人包括新华三、武汉长江计算、联想开天、河南昆仑技 术、平治信息等八家公司。 以金额计算，中国联通这次招标中，以国产算力为基础的服务器提供商拿走了绝大部分果实。 10月4日，浪潮信息方面向第一财经确认，该公司独家获得了工商银行的30亿服务器招标项目。 工商银行为这次投标设置了相对苛刻的条件。这些条件之一，是投标人应该为所投服务器产品的原厂商。工行同时要求，投标人所投海光芯片服务器同品 牌产品，在2023年1月1日至2025年6月30日，必须已 ...

国庆之际，中国算力企业收到了大礼包。 中国联通和工商银行近日公布了服务器招标信息，两家企业的项目合计金额超过了100亿人民币。这些 服务器的通用芯片以国产为主，工商银行30亿元海光芯片服务器项目被浪潮信息（000977.SZ）一家企 业拿下。 "主要还是国产芯片这几年技术起来了，外加政策又利好。"一位业内人士评价说。中国本土芯片企业正 以肉眼可见的速度成长，它们也在市场竞争中逐步站稳脚跟。 百亿"红包" 10月4日，浪潮信息方面向第一财经确认，该公司独家获得了工商银行的30亿服务器招标项目。 工商银行为这次投标设置了相对苛刻的条件。这些条件之一，是投标人应该为所投服务器产品的原厂 商。工行同时要求，投标人所投海光芯片服务器同品牌产品，在2023年1月1日至2025年6月30日，必须 已经与六大国有商业银行、12家股份制商业银行、移动、电信、联通、广电网络（含分支机构）中至少 一家签订过不低于2000万元的合同。 服务器是能耗大户。服务器提供商正在向液冷模式转变，但风冷仍然占据最主要的市场份额。这次工商 银行的海光芯片服务器采购项目同样是风冷项目。 浪潮信息最终成为30亿项目的正选中标方，中兴和联想成为备选方案 ...

中国联通和工商银行近日公布了服务器招标信息，两家企业的项目合计金额超过了100亿人民币。这些 服务器的通用芯片以国产为主，工商银行30亿元海光芯片服务器项目被浪潮信息（000977.SZ）一家企 业拿下。 "主要还是国产芯片这几年技术起来了，外加政策又利好。"一位业内人士评价说。中国本土芯片企业正 以肉眼可见的速度成长，它们也在市场竞争中逐步站稳脚跟。 百亿"红包" 10月4日，浪潮信息方面向第一财经确认，该公司独家获得了工商银行的30亿服务器招标项目。 国庆之际，中国算力企业收到了大礼包。 工商银行为这次投标设置了相对苛刻的条件。这些条件之一，是投标人应该为所投服务器产品的原厂 商。工行同时要求，投标人所投海光芯片服务器同品牌产品，在2023年1月1日至2025年6月30日，必须 已经与六大国有商业银行、12家股份制商业银行、移动、电信、联通、广电网络（含分支机构）中至少 一家签订过不低于2000万元的合同。 服务器是能耗大户。服务器提供商正在向液冷模式转变，但风冷仍然占据最主要的市场份额。这次工商 银行的海光芯片服务器采购项目同样是风冷项目。 浪潮信息最终成为30亿项目的正选中标方，中兴和联想成为备选方案 ...

Core Viewpoint - The company, iFlytek, has revised the use of funds from its 4 billion yuan private placement, allocating 2.4 billion yuan for computing power platform construction, while the remaining 1.6 billion yuan will be evenly distributed for the development of the Spark Education large model and typical products, as well as for working capital [2] Group 1: Fund Allocation and Strategy - The initial plan on August 22 allocated 3.2 billion yuan for working capital and 800 million yuan for the development of the Spark Education large model and typical products [2] - iFlytek aims to expand its computing power scale through a computing power leasing model to support large model research and algorithm upgrades [2] - The company is currently researching a new version of the iFlytek Spark large model that requires a larger-scale domestic computing power cluster for support [2] Group 2: Domestic Computing Power Development - iFlytek claims to have independent and controllable computing power compared to other domestic large model manufacturers [2] - In October 2023, iFlytek launched the first fully domestic intelligent computing platform, "Feixing No. 1," based on Huawei's Ascend 910B chip, achieving over 90% performance of a similarly scaled NVIDIA A800 cluster [2] - The company plans to train China's first trillion-parameter large model based on domestic computing power by January 2024 [3] Group 3: Challenges and Collaborations - The industry still heavily relies on NVIDIA's ecosystem for larger parameter models and algorithm innovations, facing challenges in adapting and optimizing algorithms based on domestic computing power [3] - iFlytek collaborates with domestic computing power manufacturers like Huawei and Cambrian to address hidden issues in domestic computing hardware and software [3][4] - The company acknowledges that adapting to new algorithms and architectures on domestic computing platforms incurs additional costs and time, slowing down model release progress [4]

Group 1: Semiconductor Stocks Performance - Semiconductor stocks continue to rise strongly, with Huahong Semiconductor increasing over 15% and reaching a new historical high, while leading company SMIC rose by 2.88%, also hitting a historical high [1] - Various semiconductor ETFs, including the Fortune and Guotai ETFs, saw gains of over 2% [1] Group 2: ETF Performance Details - The Fortune Sci-Tech Chip ETF (588810) rose by 2.96% with a 5-day increase of 8.32% and an estimated scale of 577 million [2] - The Guotai Sci-Tech Chip ETF (589100) increased by 2.87% with a 5-day increase of 8.34% and an estimated scale of 641 million [2] - The top ten weighted stocks in the Sci-Tech Chip ETF include Cambricon, Haiguang Information, SMIC, and others, focusing on semiconductor materials, equipment, design, manufacturing, packaging, and testing [2] Group 3: AI Chip Industry Developments - DeepSeek announced a significant update to its services, reducing API costs by over 50%, which has been adapted by several domestic chip manufacturers [3] - Analysts from Huaxin Securities express optimism about the domestic AI chip industry, highlighting a complete industry chain from advanced processes to model acceleration by major companies [3] - Zhongyin Securities notes that the commercialization of AI applications is accelerating, leading to increased demand for computing power in the domestic market [3] Group 4: New AI Models and Market Trends - Anthropic launched a new large model, Claude Sonnet 4.5, capable of running autonomously for 30 hours, excelling in cybersecurity and financial services [4] - Tencent released and open-sourced its native multimodal image model, HunyuanImage 3.0, with a parameter scale of 80 billion [4] - TrendForce predicts a shift in AI infrastructure focus towards efficient inference services, with increasing demand for Nearline SSDs due to shortages in traditional HDDs [4]

9月30日早盘，AI相关概念股集体拉升。AI语料方向，当虹科技20CM涨停，开普云、拓尔思（300229）、值得买（300785）等 涨超5%；半导体硬件方向，德明利（001309）涨停，江波龙（301308）、联芸科技等涨超5%，寒武纪、东芯股份等跟涨。 随后，多家国产芯片厂商宣布完成对DeepSeek-V3.2-Exp的适配。寒武纪发文称：已同步实现对深度求索公司最新模型DeepSeek- V3.2-Exp的适配，并开源大模型推理引擎vLLM-MLU源代码。 华为则表示，昇腾已快速基于vLLM/SGLang等推理框架完成适配部署，实现DeepSeek-V3.2-Exp0day支持，并面向开发者开源所 有推理代码和算子实现。海光信息同日宣布基于GPGPU架构强大的生态优势，与编程开发软件栈DTK的特性，DeepSeek-V3.2- Exp在海光DCU上展现出优异的性能。 华鑫证券研报表示，国产AI芯片大时代已经来临，国产AI产业链从上游先进制程到先进封装，到下游字节阿里腾讯的模型加速 迭代升级已经实现全产业链打通，坚定看好国产AI算力设施的加速突破。 中银证券（601696）分析，AI应用商业化拐点临近，应 ...

证券研究报告——晨会聚焦 2025 年 9 月 30 日 9 月金股组合 股票代码 股票名称 601816.SH 京沪高铁 中银晨会聚焦-20250930 601233.SH 桐昆股份 002409.SZ 雅克科技 ■重点关注 300750.SZ 宁德时代 600276.SH 恒瑞医药 688085.SH 三友医疗 【策略研究】策略深度报告*王君 高天然。AI 应用商业化拐点临近，重点 关注国产算力与商业化进程较快的应用垂类赛道。 | 市场指数 | | | | | --- | --- | --- | --- | | 指数名称 | | 收盘价 | 涨跌% | | 上证综指 | | 3862.53 | 0.90 | | 深证成指 | | 13479.43 | 2.05 | | 沪深 | 300 | 4620.05 | 1.54 | | 中小 | 100 | 8328.93 | 1.85 | 行业表现（申万一级） | 指数名称 | 涨跌% | 指数名称 | 涨跌% | | --- | --- | --- | --- | | 非银金融 | 3.84 | 煤炭 | (0.84) | | 有色金属 | 3.78 | 银行 ...