Core Viewpoint - The China Securities Information Technology Application Innovation Industry Index has decreased by 2.30% as of February 4, 2026, with mixed performance among constituent stocks, highlighting the volatility in the sector [1]. Group 1: Market Performance - The leading stock, Borui Data, surged by 12.28%, while other notable gainers included Geer Software with a 6.93% increase and Nasda with a 2.06% rise [1]. - Conversely, Zhuoyi Information led the decline with a drop of 7.05%, followed by Panwei Network at 5.73% and Deepin Technology at 5.20% [1]. - The Xinchang ETF (562570) fell by 2.03%, reflecting the overall downward trend in the market [1]. Group 2: Company Developments - Moore Thread has officially launched the AI Coding Plan, an intelligent programming service based on domestically produced full-function GPU computing power [1]. - This service utilizes the MTT S5000's robust full-precision computing capabilities, integrating a silicon-based flow reasoning acceleration engine and the top-tier GLM-4.7 code model, marking a significant breakthrough in AI coding [1]. Group 3: ETF Insights - The Xinchang ETF (562570) closely tracks the China Securities Xinchang Index, emphasizing domestic full-stack substitution across key areas such as storage, CPU, and AIPC, which accounts for 16% of the ETF [1]. - The ETF actively incorporates the DeepSeek ecosystem (72%) and AI applications (58%), aiming to establish a secure and controllable system from foundational hardware to industry intelligence [1]. - Currently, the Xinchang ETF (562570) is the largest ETF in its category by scale [1].

Group 1 - Moore Threads launched the AI Coding Plan, an intelligent programming service based on domestic full-function GPU computing power, featuring the MTT S5000's strong computing capabilities and integrating the GLM-4.7 code model [1] - Longzhong Securities indicates that the shift from Copilot to Agent is underway, which is expected to disrupt the software development ecosystem by transforming tools into decision-making aids and significantly reducing the marginal cost of code generation [1] - The AI Coding market is rapidly expanding, with Spherical Insights predicting it will exceed $29.5 billion by 2032, and current penetration rates show that 91% of developers in the US use AI programming tools compared to 30% in China, which is expected to increase quickly [1] Group 2 - Sanwei Tiandi's low-code development platform has integrated AI large models to enhance development efficiency, with the Sunwaylink product allowing customers to achieve programming goals with minimal learning costs [1] - Zhuoyi Information emphasizes the advantages of its AI programming platform, which utilizes a multi-agent collaborative architecture, industry/language templates, cloud-native resources, and high-quality refined data-driven specialized models for rapid application development [2]

Core Insights - Moole Technology has launched the AI Coding Plan intelligent programming service, marking a significant advancement in the domestic GPU sector as the first solution built on a fully functional domestic GPU computing base [3] - The service leverages the full precision computing capabilities of the MTT S5000 and integrates the top-tier code model GLM-4.7 from Zhiyuan AI, achieving a critical breakthrough in the AI coding field with domestic chips and models [3] Group 1 - Moole Technology collaborated with Silicon-based Flow to develop a system that significantly reduces response latency while ensuring code generation quality through efficient operator fusion and framework optimization [1] - The core model of the system, GLM-4.7, ranks first among open-source and domestic models in the Code Arena, outperforming peers in function completion and vulnerability detection, even surpassing GPT-5.2 [1] - The AI Coding Plan service is compatible with major programming tools such as Claude Code, Cursor, and OpenCode, allowing developers to switch seamlessly between different development environments [1]

Core Viewpoint - The company, Moore Threads, has officially launched its AI Coding Plan, marking a significant advancement in the field of intelligent programming services based on domestic GPU computing capabilities [1] Group 1: Product Launch - The AI Coding Plan is the first intelligent development solution built on a fully functional domestic GPU computing foundation [1] - The service leverages the MTT S5000's strong full-precision computing power as its core driver [1] Group 2: Technological Integration - The solution integrates a silicon-based flow inference acceleration engine and the top-tier GLM-4.7 code model [1] - This integration represents a key breakthrough in the AI Coding field, combining domestic chips with domestic large models [1] Group 3: Industry Impact - The launch signifies a breakthrough in domestic computing power within the AI core productivity tools sector, enabling support for high-complexity application development [1]

Core Insights - The company Moer Thread has officially launched the AI Coding Plan, an intelligent programming service based on domestically produced full-function GPU computing capabilities, marking a significant breakthrough in the AI coding field with domestic chips and models [1][3]. Group 1: Product Features - The AI Coding Plan leverages the MTT S5000's full precision computing power, achieving a significant increase in computational efficiency through a software-hardware collaborative architecture [3]. - The system integrates the GLM-4.7 model, which ranks first among open-source and domestic models in the Code Arena, outperforming peers in function completion and vulnerability detection [3]. - The service is compatible with multiple mainstream programming tools, including Claude Code, Cursor, and OpenCode, allowing developers to switch seamlessly between different development environments [3]. Group 2: Pricing and Packages - The AI Coding Plan offers a tiered package system to cater to different team sizes and usage needs: - Free Trial: 30-day free trial suitable for lightweight projects [3]. - Lite Plan: Priced at 120 yuan per quarter, offering three times the usage of the Claude Pro package for medium-frequency development scenarios [3]. - Pro Plan: Priced at 600 yuan per quarter, providing five times the usage of the Lite package, designed for complex project architectures [3]. - Max Plan: Priced at 1200 yuan per quarter, ensuring peak traffic priority for enterprise-level high-frequency usage [3]. Group 3: Strategic Vision - As the AGI era accelerates, Moer Thread aims to deepen its full-stack layout of "chip + model + toolchain," supporting Chinese developers in their intelligent transformation [5].

Group 1 - The core point of the article is that Moore Threads has officially launched its AI Coding Plan, which is an intelligent programming service [2] Group 2 - The service aims to enhance programming efficiency and support developers in their coding tasks [2] - This launch reflects the growing trend of integrating AI technologies into software development [2]

Core Insights - Moore Threads officially launched the AI Coding Plan, an intelligent programming service driven by the MTT S5000 full-precision computing capability [1] - The service integrates a silicon-based flow inference acceleration engine and the GLM-4.7 code model, designed to support high-complexity application development [1] - GLM-4.7 ranks first among open-source and domestic models in the Code Arena global developer blind evaluation system, reportedly outperforming GPT-5.2 in function completion scenarios [1] Service Offerings - The AI Coding Plan offers a 30-day free trial period, with tiered pricing ranging from 120 yuan per quarter to 1200 yuan per quarter [1] - This launch is viewed as an integration attempt of domestic chips and domestic large models in the AI programming tools sector [1]

Summary of Key Points Core Viewpoint - Moer Technology (688795.SH) has seen its stock price decline, closing at 554.92 yuan with a drop of 2.65%, reaching a new low of 536.20 yuan since its listing [1] Group 1: IPO Details - Moer Technology was listed on the Shanghai Stock Exchange's Sci-Tech Innovation Board on December 5, 2025, with a total of 70 million shares issued, representing 14.89% of the post-issue total share capital [1] - The IPO was entirely new shares, with no existing shareholders selling shares, and the issue price was set at 114.28 yuan per share [1] - The total funds raised from the IPO amounted to 799.96 million yuan, with a net amount of 757.61 million yuan after deducting issuance costs, which was 42.39 million yuan less than the original plan [1] Group 2: Use of Proceeds - The funds raised are intended for the development of Moer Technology's next-generation self-controlled AI training and inference integrated chips, next-generation self-controlled graphics chips, next-generation self-controlled AI SoC chips, and to supplement working capital [1]

Group 1 - The company, Donghua Software (002065.SZ), maintains good strategic partnerships with Moore Threads and Muxi [1] - The collaborations are progressing in an orderly manner according to the established plans [1]

Core Viewpoint - Cambricon is expected to achieve its first profitable year in 2025, but it may be experiencing a decline in quarterly revenue [1][3]. Group 1: Financial Performance - Cambricon forecasts revenue between 6 billion to 7 billion yuan for 2025, representing a year-on-year growth of 410.87% to 496.02%, with a net profit of 1.85 billion to 2.15 billion yuan, marking a turnaround from losses [3]. - In Q3 2025, Cambricon reported revenue of 1.727 billion yuan, a year-on-year increase of 1332.52%, and a net profit of 567 million yuan, also showing a positive shift [5]. - However, on a quarter-on-quarter basis, Q3 revenue declined by 2.4% and net profit decreased by 17% [5]. Group 2: Market Reactions and Comparisons - Following the earnings forecast announcement, Cambricon's stock price dropped significantly, falling over 13% and bringing its market capitalization below 460 billion yuan [1]. - Other domestic AI chip stocks also experienced declines, with stocks like Moer Technology and Muxi Shares dropping by over 4% and 5% respectively [6]. - Moer Technology and Muxi Shares have also released their 2025 earnings forecasts, with Moer expecting revenue between 1.45 billion to 1.52 billion yuan and Muxi expecting revenue between 1.6 billion to 1.7 billion yuan, both showing significant year-on-year growth but still reporting losses [6].