Core Insights - The first MUSA Developer Conference (MDC2025) was held by Moore Threads in Beijing, emphasizing the advanced and leading technology of the MUSA architecture as the foundation for full-function GPUs [1] - The company aims to enhance its research and development investments to tackle core technological challenges from hardware to software, focusing on building a self-reliant domestic computing industry ecosystem [1] Group 1 - The MUSA architecture supports full precision computing from FP4 to FP64, with a 50% increase in density and a 10-fold improvement in efficiency, capable of supporting computing clusters exceeding 100,000 cards [1] - Moore Threads plans to launch two high-performance chips based on the MUSA architecture: the "Hua Shan" chip for AI training and inference, and the "Lu Shan" chip for high-performance graphics rendering, both set to be mass-produced next year [1] Group 2 - The company showcased the "Kua E Wan Ka" intelligent computing cluster, demonstrating its capability to support trillion-parameter model training with reliability and achieving international mainstream levels in several key precision metrics [2] - Moore Threads shared its planning for the MTT C256 super-node architecture aimed at the next generation of ultra-large-scale intelligent computing centers [2] - The company also released the MTTAIBOOK, an AI computing device equipped with the intelligent SoC chip "Chang Jiang" [2]

Core Insights - The term "ecosystem" has been frequently mentioned during the first MUSA Developer Conference held by Moore Threads, highlighting its importance in the GPU industry [1][3] Group 1: Company Initiatives - Moore Threads emphasizes that ecosystem construction is a core barrier and value in the GPU industry, aiming to build a comprehensive empowerment system from hardware tools to talent cultivation [3] - The company plans to increase R&D investment to tackle core technological challenges from hardware to software, leveraging the advantages of the MUSA architecture [3] - Moore Threads has established a developer growth system that has gathered nearly 200,000 developers and learners, integrating cutting-edge technology and industry practices into over 200 universities across China, attracting more than 100,000 students [3][6] Group 2: Ecosystem Development - The first MUSA Ecosystem Center will be established in Haidian, aiming to foster more ecological applications and industrial projects [5] - The MUSA Ecosystem Center will serve as a main venue for collaboration among developers, partners, and academic institutions, focusing on four key missions: providing core technical support, showcasing results and innovations, building a talent training base, and creating an innovation incubation and service platform [6] Group 3: Regional Support and Strategy - Haidian District is committed to nurturing leading hard-tech enterprises and building a competitive industrial ecosystem, with over 240 integrated circuit companies currently established in the area [7] - The district has implemented a comprehensive policy system to support innovation, including a maximum annual R&D subsidy of 15 million yuan for qualifying companies [7] - Future plans for Haidian include focusing on becoming a source of independent innovation and a hub for emerging industries, particularly in the AI chip sector [9]

Core Insights - The article highlights the launch of the new generation GPU architecture "Huagang" by Moer Thread, which supports full precision computing from FP4 to FP64, with a 50% increase in density and a 10-fold improvement in efficiency [1][2]. Group 1: Product Features - The new GPU architecture "Huagang" features significant performance enhancements, with a 50% increase in computing density and optimized energy efficiency, supporting full precision calculations from FP4 to FP64 [2]. - It integrates a new asynchronous programming model and supports large-scale interconnectivity, enabling the expansion of intelligent computing clusters with over 100,000 cards through the self-developed MTLink high-speed interconnect technology [2]. - The architecture includes a built-in AI generative rendering framework and enhances hardware ray tracing acceleration, fully supporting DirectX 12 Ultimate for high synergy between graphics rendering and intelligent computing [2]. Group 2: Future Developments - The company plans to launch high-performance AI training and inference chip "Huashan" and a chip specialized in high-performance graphics rendering named "Lushan" based on the MUSA architecture [2]. - Moer Thread introduced the "Kua'e Wanka" intelligent computing cluster, which provides full precision and general computing capabilities, achieving efficient and stable AI training and inference at a scale of 10,000 cards [2]. - The company also unveiled the AI computing device MTT AIBOOK, equipped with the intelligent SoC chip "Changjiang" [2]. Group 3: Strategic Vision - The CEO emphasized that the ecosystem is the core moat and value of the GPU industry, and the company is committed to increasing R&D investment to tackle core technological challenges from hardware to software [3]. - Moer Thread aims to deepen collaboration with ecosystem partners through open innovation to build a self-reliant and robust domestic computing industry ecosystem [3].