Market Overview - US stock market saw gains with the Dow Jones up 0.15%, Nasdaq up 0.48%, and S&P 500 up 0.26% [1] - Notable stock movements include Nvidia rising 2.05%, Amazon up 1.09%, and Tesla and Microsoft increasing by up to 0.64% [1] Storage Sector Performance - US storage stocks experienced significant increases, with SanDisk rising nearly 17%, Western Digital up over 9%, Seagate Technology up over 5%, and Micron Technology up over 4% [2] Labor Market Data - The US Labor Department announced that the Bureau of Labor Statistics will pause operations during a government shutdown, affecting the collection and release of non-farm employment data [3] Trade Policy - Former President Trump proposed a 100% tariff on all films produced outside the US [4] Gold Reserves - The value of US gold reserves has reached $1 trillion, exceeding the official book value by over 90 times [5] AI Developments - OpenAI announced its third annual developer conference scheduled for October 6 in San Francisco, expecting 1,500 developers to attend and will unveil a new version of the Sora video generation model [6] - Anthropic launched its latest AI model, Claude Sonnet 4.5, claiming it to be the "best coding model globally" [7] Commodity Market Trends - Spot gold rose by 1.9% to surpass $3,820, reaching a historical high, while silver increased over 1.7%, briefly breaking the $47 mark [7] Securities Strategy Insights - Pacific Securities indicated that the TMT sector is at an extreme in terms of trading volume, suggesting that further investment in tech stocks may not be cost-effective [9] - The report recommends reallocating investments towards high-dividend, anti-involution, and commodity resource sectors as tech stocks may face a temporary slowdown [9] Industry Developments - DeepSeek announced a significant update to its services, reducing API costs by over 50%, which may enhance the development of custom agents in AI applications [10] - The Ministry of Industry and Information Technology aims for the mechanical industry to achieve an average annual revenue growth rate of around 3.5% from 2025 to 2026, targeting a revenue surpassing 1 trillion yuan [11] - Tesla and Apple are exploring the use of glass substrates to enhance semiconductor chip and data center performance, indicating potential growth in the glass substrate market [12] Superconducting Technology - A new world record was set for a fully superconducting magnet achieving a steady-state magnetic field of 35.1 Tesla, which could drive advancements in various high-tech applications [14] Corporate Announcements - Companies such as Gelaun Electronics and Sailyus have made significant acquisitions, indicating active M&A activity in the market [15][18] - Yinglian Co. expects a net profit of 34.5 million to 37.5 million yuan for the first three quarters, reflecting a substantial year-on-year growth [16]

Core Insights - DeepSeek has launched its latest experimental model, DeepSeek-V3.2-Exp, featuring a new attention mechanism called DeepSeek Sparse Attention (DSA), which improves training and inference efficiency while reducing API costs by over 50% [1][19]. Model Features - The V3.2 model builds on DeepSeek-V3.1-Terminus and introduces DSA, achieving faster and more efficient training and inference for long contexts [3][5]. - DSA is the first key technology branded under "DeepSeek" and is an improvement over the Native Sparse Attention (NSA) from a previous collaboration with Peking University [3][5]. - The DSA mechanism allows the model to focus on a small subset of important tokens rather than all tokens, significantly reducing computational complexity from O(L²) to O(Lk), where k is much smaller than L [8][10]. Performance Evaluation - Evaluation results indicate that DeepSeek-V3.2-Exp maintains performance levels comparable to its predecessor, with no significant decline in effectiveness across both short and long text tasks [14][15]. - Specific benchmark results show that while some metrics slightly decreased, others improved, indicating a balanced performance across various tasks [15]. Cost Efficiency - The introduction of DSA has led to substantial reductions in operational costs, with the API price being lowered by over 50% for developers [19]. - The model's deployment has demonstrated significant end-to-end acceleration and cost savings in inference [18]. Future Implications - Although still an experimental model, DeepSeek-V3.2-Exp presents a promising engineering pathway for overcoming long text processing challenges without sacrificing performance [18].

Group 1 - DeepSeek has announced the update of its official App, web version, and mini-program to DeepSeek-V3.2-Exp, resulting in a significant reduction in API costs by over 50% for developers [1] - The cost of AI inference computing power has been decreasing due to advancements in AI large models and improvements in the performance and cost-effectiveness of inference chips, with hardware costs dropping approximately 30% annually and energy efficiency improving by about 40% [1] - The continuous decline in costs for large models, represented by DeepSeek, supports the commercialization of AI applications and enhances the efficiency of distilled models [1] Group 2 - The rapid iteration of large models and enhanced inference capabilities are creating opportunities for customized Agent applications, allowing users to tailor agents based on personal data and needs [2] - Companies like Cambricon and Huawei Ascend have announced their compatibility with DeepSeek-V3.2-Exp and have open-sourced the vLLM-MLU inference engine [2] - Companies such as Fanwei Network, Kingsoft Office, and Dingjie Smart are involved in the development of Agent and AI applications [3] Group 3 - Huawei Ascend has achieved software and hardware adaptation with companies like Softcom and Changshan Beiming [4] - Jiuqi Software plans to upgrade its Nüwa GPT in early 2025, integrating deeply with mainstream large models and launching various intelligent applications [4] - Jiuqi Software's AI distillation technology is similar to that of DeepSeek, indicating a trend in the industry towards efficient model optimization [4]

Group 1 - The National Development and Reform Commission announced a new policy financial tool with a total scale of 500 billion yuan, all allocated to supplement project capital [1][2] - The Ministry of Industry and Information Technology and five other departments released the "Mechanical Industry Stabilization Growth Work Plan (2025-2026)", aiming for an average annual revenue growth rate of about 3.5% and total revenue exceeding 10 trillion yuan by 2026 [2] - In 2024, China's cultural industry is projected to achieve a revenue of 19.14 trillion yuan, a 37.7% increase compared to 2020 [2] Group 2 - The DeepSeek-V3.2-Exp model was officially released, reducing the cost of using the DeepSeek API by over 50% [3] - The total net subscription amount for multiple broad-based equity ETFs reached 22.2 billion yuan on September 26, marking a new high in over five months [3] Group 3 - The securities industry is expected to continue its high growth in Q3, with 42 listed securities firms reporting a total revenue of 251.87 billion yuan in the first half of the year, a year-on-year increase of 11.37% [5] - The average annual revenue growth rate for the securities industry is anticipated to further increase due to the active stock market and low base effects [5] Group 4 - OpenAI's upcoming developer conference on October 6 is expected to focus on the application of AI technology in hardware, potentially boosting the consumer electronics supply chain [7] - The demand for lithium batteries is surging, with production expected to grow by 10% month-on-month in October, leading to a projected annual demand growth rate exceeding 35% [8] Group 5 - China CNR Corporation announced that its total contract amount for Q3 exceeded 50 billion yuan, with significant contracts signed for various types of vehicles [9] - Huayou Cobalt signed a major supply agreement with LGES for a total of 76,000 tons of ternary precursor products from 2026 to 2030 [10] Group 6 - Tianqi Lithium received significant institutional buying, with two institutions purchasing a total of 221 million yuan worth of shares, driven by strong growth in its electrolyte business [16] - GF Securities saw institutional buying of 254 million yuan, reflecting positive performance in its brokerage and asset management businesses [17]

Core Viewpoint - DeepSeek-V3.2-Exp has been released and open-sourced, introducing a sparse Attention architecture, with Haiguang DCU achieving seamless adaptation and deep optimization for zero-wait deployment of large model computing [1] Group 1: Company Developments - Haiguang Information is committed to building an AI software stack ecosystem under the "Deep Computing Intelligence" strategy, fully supporting the global mainstream open-source large models led by DeepSeek [1] - Haiguang DCU has rapidly completed "Day0" level efficient adaptation and optimization for DeepSeek-V3.2-Exp, benefiting from long-term and active technological accumulation [1] Group 2: Technical Performance - DeepSeek-V3.2-Exp demonstrates excellent performance on Haiguang DCU, leveraging the strong ecological advantages of the GPGPU architecture and the characteristics of the programming development software stack DTK [1] - The successful performance of DeepSeek-V3.2-Exp on Haiguang DCU validates the high versatility, high ecological compatibility, and self-controllable technological advantages of Haiguang DCU, establishing it as a key infrastructure for AI large model training and inference [1]

Core Insights - DeepSeek has released the V3.2-Exp model, which introduces a Sparse Attention mechanism aimed at optimizing training and inference efficiency for long texts [1] - The official app, web version, and mini-program have all been updated to DeepSeek-V3.2-Exp, and the API has seen a significant price reduction [1] - Under the new pricing policy, the cost for developers to access the DeepSeek API will decrease by over 50% [1] - The performance of DeepSeek-V3.2-Exp on various public evaluation datasets is comparable to that of V3.1-Terminus [1]

Core Insights - The new pricing policy will reduce the cost for developers using the DeepSeek API by over 50% [2][3] - The release of the DeepSeek-V3.2-Exp model on September 29, 2025, introduces the DeepSeek Sparse Attention mechanism, enhancing training and inference efficiency for long texts [2] - The V3.2-Exp model maintains performance levels comparable to the previous V3.1-Terminus model across various benchmarks [2][3] Performance Comparison - In the MMLU-Pro benchmark, DeepSeek-V3.1-Terminus scored 85.0, while V3.2-Exp maintained the same score [3] - For the BrowseComp search benchmark, V3.2-Exp improved to 40.1 from 38.5 in V3.1-Terminus [3] - The Codeforces-Div1 benchmark saw an increase from 2046 in V3.1-Terminus to 2121 in V3.2-Exp [3] Accessibility and Development - The V3.2-Exp model has been made open-source on Huggingface and Modao platforms, allowing users to access and develop further [5] - The updated version is available on the official app, web, and mini-programs [2][3]

Core Insights - DeepSeek has released an experimental version of its model, DeepSeek-V3.2-Exp, which introduces Sparse Attention for improved training and inference efficiency on long texts [1] Group 1: Model Development - The new version, V3.2-Exp, is a step towards a next-generation architecture, building on the previous V3.1-Terminus [1] - The Sparse Attention mechanism is aimed at optimizing the model's performance for long text processing [1] Group 2: Pricing and Accessibility - The API pricing has been significantly reduced, with costs now at 0.2 yuan per million tokens for cache hits, 2 yuan for cache misses, and 3 yuan for output [1] - This pricing represents a reduction of over 50% compared to previous costs for developers using the DeepSeek API [1]

Core Insights - DeepSeek has released the DeepSeek-V3.2-Exp model, which is an experimental version aimed at transitioning to a new generation architecture [1] - The new model introduces DeepSeek Sparse Attention, focusing on optimizing training and inference efficiency for long texts [1] - The official app, web version, and mini-program have all been updated to DeepSeek-V3.2-Exp, and the API costs have been significantly reduced by over 50% for developers [1] - The performance of DeepSeek-V3.2-Exp on various public evaluation sets is comparable to that of V3.1-Terminus [1]

Core Insights - DeepSeek officially released the DeepSeek-V3.2-Exp model, which introduces a Sparse Attention mechanism aimed at optimizing training and inference efficiency for long texts [1] - The official app, web version, and mini-program have all been updated to DeepSeek-V3.2-Exp, reflecting the latest enhancements [1] - The pricing policy for the DeepSeek API has been significantly reduced, with costs for developers decreasing by over 50% [1]