Core Insights - DeepSeek has launched the experimental version DeepSeek-V3.2-Exp, which introduces the DeepSeekSparseAttention mechanism to enhance training and inference efficiency for long texts [1] - The AI community now hosts over 700 high-quality open-source models, providing developers with various services including API calls and distributed training [2] Group 1 - DeepSeek-V3.2-Exp is available for free download in the national supercomputing internet AI community, allowing enterprises and developers to quickly develop applications [1] - The new model is a step towards a next-generation architecture, building on the previous version V3.1-Terminus [1] - DeepSeekSparseAttention achieves significant improvements in long text training and inference efficiency with minimal impact on model output [1] Group 2 - The supercomputing internet AI community features a collection of over 700 models, including various versions of the DeepSeek series [2] - Developers can utilize the community for a range of services, including online inference dialogue and model fine-tuning [2] - The community supports a comprehensive MaaS (Model as a Service) offering for developers [2]

Core Insights - DeepSeek company has released the DeepSeek-V3.2-Exp model, introducing a sparse attention mechanism that significantly reduces computational resource consumption and enhances inference efficiency [1][6] - The new model has led to a price reduction of API services by 50% to 75% [1] - The release has prompted immediate recognition and adaptation from several domestic chip manufacturers, including Huawei, Cambricon, and Haiguang, indicating a growing synergy within the domestic AI hardware and software ecosystem [2][4] Summary by Sections Model Release and Features - The DeepSeek-V3.2-Exp model incorporates the DeepSeek Sparse Attention mechanism, optimizing training and inference efficiency for long texts [6] - The model is compatible with CUDA and utilizes TileLang for rapid prototyping, which is designed specifically for AI operator development [6] Industry Response - Cambricon was the first to claim adaptation of the new model, followed by Huawei and Haiguang, showcasing a collaborative effort among domestic manufacturers [2] - The rapid response from these companies indicates a consensus within the domestic AI industry regarding the significance of the DeepSeek model [6] Ecosystem Development - DeepSeek is emerging as a key player in building a new ecosystem for domestic AI, with its model becoming a benchmark for open-source models in China [2][4] - The collaboration among major internet companies like Tencent and Alibaba in adapting domestic chips further accelerates the establishment of this ecosystem [7] Historical Context - The previous version, DeepSeek-V3.1, did not receive any proactive claims from companies regarding its adaptation, highlighting the significant shift in industry dynamics with the latest release [5] - Experts believe that the rapid development of domestic chips by 2025 can be attributed to the emergence of DeepSeek as a standard-setting entity [3]

Group 1 - DeepSeek has launched a new experimental model version, DeepSeek-V3.2-Exp, which incorporates the "DeepSeek Sparse Attention" mechanism to enhance training and inference efficiency in long context scenarios [1] - The new architecture of DeepSeek-V3.2-Exp has significantly reduced API pricing, with costs dropping by 75%, making it more affordable for developers to utilize DeepSeek API [1] - PPIO platform offers high-performance API services and features a variety of open-source models, achieving the top rank in throughput tests for DeepSeek-R1-0528 according to the "2025 Large Model Service Performance Ranking" [2] Group 2 - PPIO has successfully achieved over 10 times cost reduction in large model inference through practices in 2024, balancing inference efficiency and resource usage dynamically [2]

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 ...

Group 1: Technology Developments - DeepSeek has released version V3.2-Exp, significantly reducing API prices by over 50% due to lower service costs [2] - Huawei's CEO confirmed that the company has adapted to the new DeepSeek model and is supporting it with their Ascend architecture [13] - The latest AI model Qwen3-Omni from Alibaba has topped the global open-source model rankings, showcasing advanced multimodal capabilities [5] Group 2: Corporate Announcements - Apple CEO Tim Cook disclosed personal investments in Bitcoin and Ethereum but clarified that Apple will not accept cryptocurrency for product purchases [3] - Saisys has completed the payment for acquiring a 10% stake in Shenzhen Yiwang Intelligent Technology from Huawei, totaling RMB 11.5 billion [4] - Xiaomi's public relations manager stated that there are no plans to reduce orders for the Xiaomi 17 series, indicating an increase in overall product orders [9] Group 3: Financial Performance - Saisys announced a cash dividend of RMB 3.10 per 10 shares, totaling RMB 506 million, representing 17.22% of its net profit for the first half of the year [14] - Yinglian Co. expects a significant increase in net profit for the first three quarters of 2025, projecting a growth of 1531.13% to 1672.97% [15] Group 4: Infrastructure and Standards - CATL plans to invest in the construction of 100 battery swap stations in Hainan by 2030 to enhance green transportation [10] - A new national standard for cross-border personal information security management will be implemented in March 2026, aimed at regulating data processing activities [11]

Core Insights - DeepSeek v3.2 introduces a significant change by adopting TileLang, a domain-specific language for GPU kernel development, which has garnered substantial attention in the tech community [1][4][6] - TileLang is noted for its performance, allowing developers to implement attention mechanisms faster than existing solutions, with claims of achieving a 30% speed increase over Flash Attention 2 [3][5] Group 1: TileLang Overview - TileLang is designed to simplify the development of high-performance GPU/CPU kernels, comparable to NVIDIA's CUDA, and is recommended by DeepSeek for experiments due to its debugging and rapid iteration advantages [4][13] - The language is built on a Python-like syntax and operates on top of the TVM compiler infrastructure, enabling developers to focus on productivity without sacrificing performance [13] - TileLang features three programming interfaces catering to different developer skill levels, from high-level abstractions for beginners to low-level controls for performance experts [15] Group 2: DeepSeek's Adoption of TileLang - DeepSeek's collaboration with TileLang was first highlighted at the Beijing Zhiyuan Conference in June, where a report indicated that TileLang's operator implementation could be faster [6][19] - The DeepSeek team has utilized TileLang for rapid prototype development, subsequently optimizing performance with lower-level methods [17][23] - Following the release of DeepSeek v3.2, TileLang's capabilities were validated, demonstrating its effectiveness in model training [23]

Group 1: Technology Developments - DeepSeek has released version V3.2-Exp, significantly reducing API prices by over 50% due to lower service costs, with Huawei Cloud and Cambricon already adapting to the new model [2] - Alibaba's Tongyi has seven models ranking in the top ten of the global open-source model list, with the newly released Qwen3-Omni model achieving the top position, showcasing capabilities across text, image, voice, and video [5] - Huawei's Ascend has announced support for DeepSeek-V3.2-Exp, quickly adapting and deploying the model for developers [13] Group 2: Corporate Announcements - Apple CEO Tim Cook confirmed personal investments in Bitcoin and Ethereum but stated that Apple will not accept cryptocurrency for product purchases or invest its cash reserves in crypto assets [3] - Seres has completed the payment for acquiring a 10% stake in Shenzhen Yiwang Intelligent Technology from Huawei, totaling RMB 11.5 billion [4] - Xiaomi's PR manager announced that there are no plans to reduce orders for the Xiaomi 17 series, with an increase in overall product orders due to new versions [9] Group 3: Financial Performance - Seres announced a cash dividend of RMB 3.10 per 10 shares, totaling RMB 506 million, representing 17.22% of its net profit for the first half of 2025, which saw an 81.03% year-on-year increase [14] - Yinglian Co. expects a significant increase in net profit for the first three quarters of 2025, projecting a rise of 1531.13% to 1672.97% year-on-year, driven by the metal packaging sector [15] Group 4: Strategic Initiatives - CATL plans to build 100 battery swap stations in Hainan by 2030, enhancing the green transportation system in the Hainan Free Trade Port [10] - The State Administration of Taxation reiterated that platform enterprises cannot transfer tax obligations, with a focus on protecting the rights of low-income workers [7] - The market regulatory authority has released China's first national standard for cross-border personal information security management, effective March 1, 2026 [11]

Group 1 - The Shanghai Stock Exchange Sci-Tech Innovation Board Chip Index (000685) has seen a strong increase of 2.14%, with constituent stocks such as Baiwei Storage (688525) rising by 7.66%, Yandong Micro (688172) by 7.10%, and Lexin Technology (688018) by 5.24% [1] - The Sci-Tech Chip ETF Index (588920) also rose by 2.30%, with the latest price reported at 1.65 yuan [1] - Tianfeng Securities highlights that an AI storage revolution is underway, driven by "computing through storage," which significantly reduces computing power consumption and accelerates AI inference, leading to a higher growth rate in SSD demand compared to traditional curves [1] Group 2 - The Sci-Tech Chip ETF Index closely tracks the Shanghai Sci-Tech Innovation Board Chip Index, which selects securities related to semiconductor materials and equipment, chip design, manufacturing, packaging, and testing from listed companies on the Sci-Tech Board [2] - As of August 29, 2025, the top ten weighted stocks in the Sci-Tech Chip Index include Cambricon (688256), Haiguang Information (688041), SMIC (688981), and others, with these ten stocks accounting for a total of 62.02% of the index [2]

Group 1 - DeepSeek announced the upload of its new model DeepSeek-V3.2 to the HuggingFace community platform on September 29 [2] - Zhiyu's new model GLM-4.6 is expected to be released soon, with some users already able to access it via API [2] - Both DeepSeek and Zhiyu, leading companies in the large model sector in China, are poised for significant advancements [2] Group 2 - The DeepSeek-V3.1 model, released in August, features a hybrid inference architecture that supports both thinking and non-thinking modes [2] - DeepSeek-V3.1-Think offers improved thinking efficiency, providing answers in a shorter time compared to DeepSeek-R1-0528 [2] - The new model has enhanced agent capabilities through post-training optimization, showing significant improvements in tool usage and agent tasks [2] Group 3 - Zhiyu's flagship model GLM-4.5, launched in July, integrates reasoning, coding, and agent capabilities within a single model to meet complex application needs [2] - In August, Zhiyu introduced the GLM-4.5V, a top-performing open-source visual reasoning model with 100 billion parameters (total parameters 106 billion, active parameters 12 billion) [2]

华泰证券曾表示，模型降价将吸引更多的开发者开发AI应用，或进一步提振算力需求，提升Super App 出现概率。中银国际认为，AI应用商业化拐点临近。在算力层，推理效率与性价比大幅提升，国产芯 片加速替代；在模型层，通用大模型的能力已逐步达到商用标准；在数据层，行业专属数据的积累与合 成数据技术成熟之下，企业加速实现数据闭环训练与模型微调。三者共同推动AI能力从"单点突破"走 向"体系协同"，为AI应用大规模商业化落地创造条件。 AI概念股早盘多数走高，截至发稿，汇量科技(01860)涨4.47%，报19.88港元；迈富时(02556)涨4.33%， 报51.35港元；创新奇智(02121)涨3.65%，报7.95港元；第四范式(06682)涨3.15%，报65.5港元；美图公 司(01357)涨3.26%，报9.16港元。 消息面上，DeepSeek昨日宣布，官方App、网页端、小程序均已同步更新为DeepSeek-V3.2-Exp。 DeepSeek介绍，得益于新模型服务成本的大幅降低，官方API价格也相应下调，新价格即刻生效。在新 的价格政策下，开发者调用DeepSeek API的成本将降低50%以上。 ...