周末吹国产芯片太狠了。各种角度，各种姿势吹，吹太过了。 但是真正深入研究过的人会发现，吹的太狠，但整个产业链少了一个必要环节！ 代工厂！国产代工厂！ 大部分人根本不知道3月国产芯片真正跌的原因（包括hwj）。 表观上是几个大厂capex不行，再深一步，发觉是h20断供！ 所以3-5月国产算力跌的是"巧妇难为无米之炊"，不是没需求，是没供给。看懂了吗？ 现在要讲国产自主强的逻辑，上周五我们写个那篇也只点评ds 3.1适配国产FP8，需求会变好。这是个边 际变好。 但是没有解读一个东西，供给！ 所以这个逻辑里面还有一环没补上！供给！ 万众瞩目，需要一个东西，需要良率提升！ 看懂了吗？ 再进一步可以挖到台积电早断了所有你们想的到的代工！不然那么多大厂自研芯片 眼巴巴的望着国产 良率？它们都是大户。 ...

Core Viewpoint - The recent surge in activity within investment groups, which had been dormant for three years, indicates a renewed interest and excitement among investors [1]. Group 1 - Investment groups that had previously seen minimal communication are now experiencing a significant increase in discussions about stocks, with conversations becoming more frequent and engaging [2].

Core Viewpoint - The successful implementation of the deepseek R1 is attributed to the use of the FP8 data format within a fine-grained mixed precision framework, which allows most compute-intensive operations to be performed at FP8 precision while retaining original data formats for a few critical operations [1]. Group 1 - The previous libraries were optimized based on CUDA, giving NVIDIA cards an advantage, while domestic chips only supported FP16, resulting in a 37% efficiency loss when using R1 [1]. - The recent adaptation of FP8 for domestic chips is expected to reduce costs and benefit local hardware [1]. - Future advancements in both software and hardware are anticipated, with NVIDIA's B cards potentially lowering precision to FP4, while several domestic companies are expected to support native FP8 in their next-generation products [2]. Group 2 - The focus on low-cost solutions is aimed at expanding into global markets [3].

Core Insights - The article highlights the significant advancements of DeepSeek V3.1, particularly in its ability to handle long contexts and improve programming capabilities, which positions it as a leading open-source model in the industry [1][3][4]. Performance Breakthroughs - DeepSeek V3.1 has achieved a breakthrough in context processing, expanding its context window to 128K tokens, doubling the previous version's capacity, allowing it to handle approximately 100,000 to 130,000 Chinese characters [1]. - The model's enhancements in memory management and attention mechanism have resolved issues related to context loss and fragmented responses in long text processing [1]. Application Scenarios - The model's 128K context capability significantly improves efficiency in legal document review and academic paper summaries, allowing for the input of complete lengthy documents while maintaining logical coherence and detail accuracy [2]. - In developer scenarios, the model supports large codebase dependency analysis and technical document parsing, demonstrating superior context retention and solving previous issues of output loops and information fragmentation [2]. Programming Capabilities - DeepSeek V3.1 has made comprehensive advancements in programming, redefining the performance boundaries of open-source programming models [3]. - In benchmark tests, it scored 71.6% in the Aider Polyglot multi-language programming assessment, outperforming competitors and showing improved accuracy in Python and Bash code generation [4]. Cost Efficiency - The model has achieved a significant cost reduction, with the average cost for completing typical programming tasks being only $1.01, which is 1/68 of closed-source models [7]. - This cost advantage is expected to disrupt the development processes of small and medium enterprises, promoting a shift towards localized, high-efficiency, and low-barrier programming tools [7]. Enhanced Agent Capabilities - DeepSeek V3.1 has improved its tool usage and function calling capabilities, transitioning from "cognitive" to "execution" roles, enhancing its task processing abilities [8]. - The model's compatibility with existing APIs reduces migration costs and enhances cross-platform collaboration efficiency [9]. Reliability and Development Efficiency - The introduction of the Beta version of Strict Mode ensures high accuracy in output formats, particularly in sensitive fields like finance and healthcare, achieving a 99% accuracy rate in data structure compliance [10]. - The model's template-based tool calling reduces integration time by 50%, significantly improving development efficiency [11]. Vertical Capabilities and Practical Applications - The model demonstrates high efficiency in code generation and repair tasks, with costs significantly lower than closed-source competitors [14]. - In enterprise DevOps processes, it automates the generation of deployment scripts, achieving a cost reduction of 1/30 compared to using other models [15]. API Pricing Adjustments - Starting September 6, 2025, DeepSeek V3.1 will adjust its API pricing strategy, with input prices set at 0.5 yuan per million tokens for cache hits and 4 yuan for misses, while output prices will be 12 yuan per million tokens [16]. - Despite some increases in single-call costs, the overall cost-effectiveness remains competitive due to improved token efficiency and faster inference speeds [17].

Group 1 - The U.S. stock market exhibited significant polarization recently [1] - Semiconductor stocks in Taiwan experienced an average decline of 3% [3] - The recent developments are linked to Trump's vision of "manufacturing return," emphasizing U.S. self-sufficiency [4] Group 2 - The Trump administration announced an expansion of tariffs on steel and aluminum imports by 50%, including hundreds of derivative products [4] - This expanded tariff list officially took effect on the 18th of the month [4]

Group 1 - The article discusses the recent performance of the Shanghai Composite Index, highlighting a trading volume of 476.3 million and a total transaction amount of 691.787 billion, indicating significant market activity [1] - The index reached a high of 3738.59 and a low of 3702.38, with a closing price of 3738.28, reflecting a 1.12% increase from the previous close of 3696.77 [1] - The price-to-earnings ratio is noted to be around 16.0, while the price-to-book ratio stands at 1.47, suggesting a valuation perspective on the index [1] Group 2 - The article raises questions about whether the current market highs are sustainable or if they are driven by leading technology companies, indicating a focus on sector performance [3]

Core Viewpoint - The current market favors large-cap stocks due to their ability to generate substantial earnings, which are considered rare in the present environment [1] Group 1 - Large-cap stocks are more likely to be integrated into global supply chains compared to smaller companies [1] - The ability to deliver strong earnings is increasingly scarce, making companies with significant performance growth highly valuable [1] - The emphasis on substantial earnings growth highlights the competitive advantage of larger firms in the current market landscape [1]

Core Viewpoint - The article emphasizes the growing trend of liquid cooling technology in data centers, particularly in relation to AI applications and the performance improvements it offers over traditional cooling methods. Group 1: Liquid Cooling Technology - Liquid cooling servers outperform air-cooled versions by 25% in performance and reduce power consumption by 30% [1] - The adoption of liquid cooling is expected to rise significantly, with projections indicating that over 65% of new systems will utilize this technology by next year [1] - The cost-effectiveness of liquid cooling solutions is highlighted, as they are seen as a way to lower overall expenses while maintaining high performance [1][6] Group 2: Company Performance and Projections - NVIDIA is projected to ship approximately 30,000 units of the GB200 and 10,000 units of the GB300, with an additional 200,000 units of the B200 single card expected [5] - The GB300 is anticipated to see increased shipments, with expectations raised to 100,000 units [5] - The server assembly sector is experiencing a resurgence, indicating a positive outlook for companies involved in this space [5] Group 3: Industry Trends - All Azure regions are now equipped to support liquid cooling, enhancing the flexibility and efficiency of data center operations [3] - The industry is moving towards building gigawatt and multi-gigawatt data centers, with over 2 gigawatts of new capacity established in the past year [4] - The trend towards liquid cooling is accelerating, with domestic manufacturers beginning to capture market share [6]

Group 1 - The market has reached new highs, indicating a strong performance overall [1] - Tencent's upcoming financial report is anticipated, with expectations that capital expenditures (capex) will improve in the following quarters [2] - Alibaba Cloud's revenue and capex are expected to be strong, particularly due to significant overseas investments [3] Group 2 - There is potential for AI infrastructure growth in Southeast Asia and the Middle East, driven by strong demand for high-power laser technology [4] - The trend towards liquid cooling in chip technology is becoming essential, with more refined solutions being the only viable option [5]

Core Viewpoint - The article discusses the positive outlook for the semiconductor industry, particularly focusing on NVIDIA and AMD, highlighting strong demand and expected revenue growth in the upcoming quarters [4][5]. Group 1: NVIDIA's Financial Performance - NVIDIA is expected to report a revenue of $45.5 billion for the current quarter, with guidance of $45 billion, and $52.5 billion for the next quarter [4]. - The company anticipates shipping nearly 30,000 units of the GB200 and 10,000 units of the GB300 this year, along with 200,000 units of the B200 single card [4]. - The GB300 is projected to have a significant increase in shipments next year, with expectations raised to 100,000 units [4]. Group 2: Market Dynamics and Product Developments - The server assembly industry is experiencing a resurgence, with all manufacturers showing positive performance [4]. - The GB200 and GB300 cabinets will utilize liquid cooling, which enhances performance by 25% and reduces power consumption by 30% compared to air cooling [4]. - AMD's single-chip performance is comparatively lower, prompting the company to recommend liquid cooling solutions to its clients [4]. Group 3: Competitive Landscape - The TPU (Tensor Processing Unit) is positioned as a leader in ASIC technology, with a significant increase in liquid cooling adoption expected next year, projected to exceed 65% [5]. - The V7 TPU is expected to achieve 4,614 TFlops, a 2.5 times increase over the V6, with memory capacity also increasing by 4.5 times [5]. - The lower cost of the TPU compared to the B200 suggests a potential shift towards more affordable domestic products in the market [5].