Core Viewpoint - The debut of Muxi Technology has marked a significant moment in the domestic GPU market, with its stock price soaring over 700% on the first day of trading, reaching a market capitalization of 332 billion yuan, nearly rivaling that of Moer Technology [1][3]. Group 1: Company Overview - Muxi Technology, founded in September 2020 and headquartered in Shanghai, specializes in high-performance general-purpose GPU chips and solutions, targeting AI training, inference, data centers, and high-performance computing (HPC) [4]. - The founding team of Muxi consists of three former AMD scientists, indicating a strong background in GPU development [5]. Group 2: Market Context - The high valuation of Muxi, despite not being profitable and having a sales ratio exceeding 150 times, reflects a market shift towards domestic GPU companies amid increasing demand for computing power and changes in the global supply chain [3][12]. - The focus of the market has shifted from verifying the technical feasibility of domestic GPU firms to assessing their mass production capabilities and revenue generation [3]. Group 3: Financial Performance - Muxi's revenue has shown rapid growth, with figures of 426,000 yuan in 2022, 53.02 million yuan in 2023, and projected revenues of 743 million yuan in 2024, and 1.236 billion yuan in the first three quarters of 2025 [13][14]. - Despite this growth, the company is still operating at a loss, with a projected net loss of approximately 346 million yuan for the first three quarters of 2025 [14][16]. Group 4: Competitive Landscape - Muxi's strategy involves high compatibility with existing CUDA applications, which lowers the initial usage barrier for customers, while also indicating a reliance on existing programming paradigms [11]. - The company aims to achieve breakeven by 2026, earlier than Moer Technology's target of 2027, highlighting a competitive edge in commercializing its products [16]. Group 5: Future Outlook - The market is keenly observing which domestic GPU company can establish a sustainable business model and achieve profitability in the next 2-3 years, as the competition intensifies [18].

Group 1 - The company, Baiwei Storage, is experiencing rapid growth in its enterprise-level business, having obtained core supplier qualifications from major server manufacturers, leading internet companies, and top domestic OEM manufacturers, marking a significant enhancement in its commercial capabilities [1][3] - The enterprise-level products are designed for data centers and critical business scenarios, requiring higher performance stability, sustained write lifespan, and long-term reliability, with a more stringent certification cycle and integration process [3] - The acquisition of core supplier qualifications typically indicates the completion of multiple validation rounds and entry into customer systems, paving the way for subsequent bulk orders [3] Group 2 - The collaboration among server manufacturers, internet companies, and OEMs is expected to accelerate the ramp-up of new enterprise-level products, with the company improving yield and delivery stability during the pre-mass production phase through process optimization and consistency control [3] - Industry trends indicate that AI training and inference are driving data center expansion, leading to changes in enterprise storage demand, with higher bandwidth, lower latency, and greater durability becoming key selection criteria [3] - The company plans to enhance its product matrix and deepen collaborative development with core customers, focusing on performance, reliability, and energy efficiency while optimizing the supply chain to improve large-scale supply capabilities [3] Group 3 - The combination of core qualifications and pre-mass production is seen as a critical step towards commercialization, with future attention needed on the pace of bulk integration, fluctuations in downstream demand, and the impact of price competition on profitability [3] - If the company successfully achieves large-scale mass production, there is potential for further optimization of customer and product structures [3]

Investment Rating - The report maintains a "Buy" rating for the company, indicating an expected stock price increase of over 15% relative to the industry index within the next six months [4][17]. Core Insights - The company achieved steady revenue growth in H1 2025, with a revenue of 1.945 billion yuan, representing a year-on-year increase of 19.57%. The net profit attributable to the parent company was 99.166 million yuan, up 31.83% year-on-year [2][3]. - The company is actively enhancing its product development and market competitiveness through continuous R&D investments, which amounted to 59.486 million yuan in H1 2025, a 10.70% increase year-on-year [3]. - The company is benefiting from expanding demand in various downstream applications within the cable industry, with notable growth in sectors such as new energy cables and automotive cables, which saw revenue increases of 81.23% and 83.54% respectively [2][3]. Financial Summary - The company's projected financial performance shows a steady increase in revenue and net profit from 2023 to 2027, with revenues expected to grow from 3.186 billion yuan in 2023 to 5.647 billion yuan in 2027, reflecting a compound annual growth rate (CAGR) of 19.5% [1][9]. - The net profit attributable to the parent company is forecasted to rise from 144 million yuan in 2023 to 276 million yuan in 2027, with an expected EPS increase from 0.45 yuan to 0.85 yuan over the same period [1][9]. - The company's return on equity (ROE) is projected to improve from 10.8% in 2023 to 14.6% in 2027, indicating enhanced profitability and efficiency [1][9].

Core Viewpoint - DeepSeek's announcement of its new model DeepSeek-V3.1 utilizing UE8M0 FP8 Scale parameter precision has sparked significant interest in the capital market, leading to a surge in stock prices of chip companies like Cambrian. However, industry insiders express a more cautious outlook regarding the practical value and challenges of FP8 in model training and inference [1][4]. Group 1: DeepSeek's Impact on Capital Market - The launch of DeepSeek-V3.1 has led to a strong reaction in the capital market, with stock prices of chip companies rising sharply [1]. - The industry response at the 2025 Computing Power Conference was more subdued, focusing on the actual value and challenges of FP8 rather than the excitement seen in the capital market [1]. Group 2: Understanding FP8 - FP8 is a lower precision format that reduces data width to 8 bits, enhancing computational efficiency compared to previous formats like FP32 and FP16 [2]. - The direct advantages of FP8 include doubling computational efficiency and reducing network bandwidth requirements during training and inference, allowing for larger models to be trained or shorter training times under the same power consumption [2]. Group 3: Limitations of FP8 - While FP8 offers speed advantages, it can lead to calculation errors due to its limited numerical range, necessitating a mixed precision training approach to balance efficiency and accuracy [3]. - Different calculations have varying precision requirements, with some operations being more tolerant of lower precision [3]. Group 4: Future of DeepSeek and FP8 Standards - DeepSeek's use of FP8 is seen as a signal that domestic AI chips are entering a new phase, providing opportunities for local computing power manufacturers [4]. - The industry acknowledges that while FP8 represents a step towards computational optimization, it is not a panacea, and the actual implementation results are crucial [4]. - The transition to FP8 may require an upgrade across the entire domestic computing ecosystem, including chips, frameworks, and applications [4]. Group 5: Challenges in Large Model Training - The core bottlenecks in large model training and inference include not only computational scale but also energy consumption, stability, and cluster utilization [5]. - There is a need for advancements from simple hardware stacking to more efficient single-card performance and optimized cluster scheduling to meet growing demands [5].