Core Insights - The Huawei Connect 2025 conference highlighted the rapid growth of the MaaS (Model as a Service) industry, which is crucial for the large-scale implementation of AI technologies [1] - Wisers, as a key partner of Huawei, showcased its enterprise-level AI productivity engine based on Huawei Cloud, emphasizing the importance of scalable and controllable AI applications [1][3] Industry Trends - The demand for large model services is accelerating, leading to an explosive growth trend in the MaaS industry [1] - The conference featured discussions on the evolution of MaaS platforms, providing actionable solutions for enterprises to enhance productivity in the AI era [1] Company Developments - Wisers has implemented a dual-active LLM and free distillation industry model (2+1) on Huawei Cloud, enabling more accurate and faster industry-specific applications [3] - The company faces challenges in the deployment of large models, including difficulties in computing power acquisition and high technical barriers, which it addresses through six core advantages of Huawei Cloud [3][4] Technological Advancements - Wisers' self-developed media big data mining and analysis model (Wisers Industry LLM) leads the industry in effectiveness and efficiency, unlocking significant AI commercial value [3] - The model can accurately identify complex semantics and filter irrelevant content, enhancing the efficiency and timeliness of alerts [3][4] Future Outlook - Wisers aims to build an AI-driven intelligent data platform and business decision solutions, collaborating with industry partners to promote broader and deeper AI applications [5]

Core Insights - The 4th 828 B2B Enterprise Festival opened in Guiyang, aiming to accelerate AI application across various industries through technology accessibility and ecosystem collaboration [1][2] - Huawei is committed to building a national computing power hub in Guizhou, enhancing its cloud services to support enterprise digitalization and intelligence [2][3] - The festival showcased over 12,000 new products and nearly 600 selected intelligent products and solutions, promoting cost reduction and innovation for enterprises [5] Group 1: Event Overview - The festival was co-initiated by Huawei and 17 leading companies, focusing on AI, intelligent computing, and data technologies [1] - Key figures from the government and Huawei delivered speeches emphasizing the importance of digital transformation and collaboration [1] Group 2: Technological Advancements - Huawei Cloud announced the integration of its Tokens service with the CloudMatrix384 super node, achieving high throughput and low latency performance [3] - The new computing architecture and hardware optimizations are designed to enhance AI application efficiency [3][4] Group 3: Industry Collaboration - Various industry leaders shared their AI innovation practices based on Huawei Cloud, providing benchmarks for other enterprises [4] - The festival included signing agreements for national intelligent enterprise computing power cooperation, promoting resource integration and AI technology implementation [4] Group 4: Future Initiatives - The upcoming 828 National Action Month will include initiatives for accelerating AI applications and supporting over 100,000 enterprises with subsidies [5]

Group 1 - Huawei Cloud announced its Tokens service is fully integrated with CloudMatrix384 super nodes, leveraging a "mixed bag" advantage to enhance performance through system innovation [1] - The xDeepServe architecture innovation allows for a maximum of 2400 TPS and 50 ms TPO, achieving high throughput and low latency performance that surpasses industry standards [1] - In the past 18 months, China's AI computing power demand has grown exponentially, with daily Token consumption increasing from 100 billion at the beginning of 2024 to over 30 trillion by June 2023, reflecting a growth of over 300 times [1] Group 2 - CloudMatrix384 super nodes are a revolutionary AI architecture introduced by Huawei, designed to address traditional computing power bottlenecks with features such as high throughput, low latency, and high elasticity [1] - The integration of CloudMatrix384 super nodes significantly enhances Tokens processing efficiency, accelerating the commercialization of AI [1] - Shengke Communication is identified as a rare domestic Ethernet switch chip design company, targeting large-scale data center and cloud service demands with flagship chips offering switching capacities of 12.8 Tbps and 25.6 Tbps [1] Group 3 - Oulutong is a high-power server power supply provider that benefits from the growth of the AI industry and opportunities for domestic substitution [2]

Core Viewpoint - Huawei Cloud has announced the full integration of its Tokens service with the CloudMatrix384 super node, achieving a significant performance breakthrough with a maximum throughput of 2400 TPS and a low latency of 50 ms, surpassing industry standards [1][2]. Group 1: AI Computing Demand and Tokens Service - Over the past 18 months, the demand for AI computing power in China has grown exponentially, with daily Token consumption increasing from 100 billion at the beginning of 2024 to over 30 trillion by June 2023, a growth of over 300 times in just 1.5 years [2]. - Huawei Cloud launched its Tokens service based on MaaS in March 2023, offering various service specifications to meet different performance and latency requirements for AI tools [2]. - The integration of Tokens service with CloudMatrix384 has led to an increase in throughput from 1920 TPS at the beginning of the year to 2400 TPS [2]. Group 2: Full-Stack Innovation and Architecture - The construction of large computing power is a full-stack innovation encompassing hardware, software, operators, storage, inference frameworks, and super nodes, leveraging Huawei's comprehensive capabilities [4]. - The CloudMatrix384 super node features a new computing architecture that breaks performance bottlenecks and establishes a robust computing foundation [4]. - The CANN Ascend hardware optimizes operators and communication strategies, enabling efficient utilization of cloud computing power [4]. Group 3: xDeepServe and Performance Enhancement - xDeepServe, as a native service of CloudMatrix384, utilizes a Transformerless architecture to decompose large models into independent micro-modules, allowing for parallel processing across different NPUs [5][6]. - The performance of Tokens service has improved from 600 tokens/s on non-super nodes to 2400 tokens/s on super nodes through continuous optimization of xDeepServe [6]. - FlowServe, a restructured decentralized distributed engine, allows for autonomous DP groups within CloudMatrix384, ensuring high concurrency without congestion [6]. Group 4: Model Performance and Industry Applications - Huawei Cloud's MaaS service supports major large models and has developed capabilities for model performance optimization, achieving twice the output speed of mainstream platforms for image generation [8]. - The company has partnered with over 100 organizations to develop AI Agents across various industry scenarios, enhancing efficiency in fields such as analysis, content creation, and smart operations [8][9]. - The introduction of intelligent solutions, such as the talent digital employee solution, demonstrates the application of advanced technologies to improve service efficiency and customer satisfaction [9].