Investment Rating - The report does not explicitly state an investment rating for the industry Core Insights - The integration of 5G-A and AI is driving a new wave of technological revolution and industrial transformation globally, with a focus on creating a unified infrastructure that supports digital transformation across various sectors [4][9] - The report emphasizes the need for a "computing-network-intelligence" integrated solution to address the challenges faced by traditional network architectures in supporting industrial control and smart manufacturing [5][11] - China Mobile proposes a "5G-A × AI integrated solution" that leverages its network advantages and technological innovation to enhance production efficiency, optimize operational costs, and improve security [5][19] Summary by Sections 1. Background of 5G-A and AI Integration - The integration of 5G-A and AI is positioned as a core driver for new industrialization in China, supported by national policies aimed at enhancing the synergy between AI and the real economy [9][10] - Key challenges include the lack of integration among diverse network types, difficulties in real-time response to business needs, and the need for intelligent operations management [11][12] 2. 5G-A × AI Integrated Technology System - The report outlines a technology system based on the principle of "computing as the engine, network as the foundation, and intelligence as the brain," promoting deep integration of connectivity, computing power, and capabilities [14][15] - The architecture includes an edge intelligent core network, heterogeneous computing platform, and intelligent driving center, which collectively enhance the capabilities of enterprises [16][19] 3. Key Capabilities of Computing-Network-Intelligence - The edge intelligent core network is supported by a computing platform that allows for lightweight and elastic deployment, ensuring efficient resource allocation and high reliability [21][22] - Features such as heterogeneous access, intent-driven networking, and multi-modal perception are highlighted as essential for enhancing enterprise private networks [27][28][33] 4. Application Cases - The report presents case studies demonstrating the effectiveness of the integrated solution, such as reducing PLC control transmission latency from 60ms to under 20ms, significantly improving operational efficiency [48][53] - Another case illustrates the implementation of intent-driven intelligent management for 5G private networks, addressing challenges in traditional network operations [54][55]

Investment Rating - The report does not explicitly provide an investment rating for the industry. Core Insights - The integration of 5G-A and AI is expected to create significant value across various sectors, enhancing network performance and efficiency while unlocking new services and applications [9][50]. - The report highlights the transformative potential of 5G-AxAI technologies in meeting evolving network demands and driving industrial innovation [10][11]. Summary by Sections 1. Executive Summary - The report emphasizes the rapid development of AI technologies and their integration with 5G-A, which is anticipated to drive a multiplier effect across industries, enhancing network performance and creating new services [9]. 2. 5G-AxAI: New Capabilities to Meet New Demands - 5G-A has reached a stage where it can provide high-speed, low-latency connectivity, with actual speeds increasing to 3-5 Gbps and expected to exceed 10 Gbps [12][13]. - The report notes that over 60 operators globally have announced commercial plans for 5G-A, with significant deployments in China [14][16]. 3. 5G-AxAI Emerging Technologies - The report identifies four key areas of innovation: Network Intelligence, Digital Twin Network Intelligence, Application Intelligence, and Sustainable Intelligence, all contributing to new applications and services [60]. 4. 5G-AxAI Implementation of New Cases - Various new applications are highlighted, including differentiated experience assurance, industrial certainty services, and immersive experiences, showcasing the versatility of 5G-AxAI technologies [4][4][4]. 5. The Fifth New Model Industrial Revolution - The report discusses innovative business models emerging from the integration of 5G-A and AI, indicating a shift from traditional connectivity to a more integrated approach combining connectivity, computation, and intelligence [59]. 6. Global Industry Cooperation Proposal - The report suggests that the collaboration among global industry players is crucial for the successful implementation and scaling of 5G-AxAI technologies [8]. 7. Glossary - A glossary is provided to clarify terms used throughout the report, aiding in understanding the technical aspects of 5G-AxAI [7]. 8. References - The report includes a comprehensive list of references to support the data and claims made throughout the document [8].

Investment Rating - The report does not explicitly provide an investment rating for the industry Core Insights - The report discusses the innovative nature of multimodal large language models (MLLMs) that integrate language processing with multimodal capabilities, enabling them to handle various data types such as text, images, and videos [2][4] - It highlights the challenges posed by the large scale and high costs of training and inference for MLLMs, which limit their widespread application in academia and industry [4][29] - The focus is on the development of efficient and lightweight MLLMs, particularly for edge computing scenarios, which presents significant potential for future advancements [4][29] Summary by Sections Overview of Multimodal Large Language Models - MLLMs have gained success due to the scaling law, where increased resource investment leads to better performance, but high resource demands restrict their development and deployment [29] - The report emphasizes the need for lightweight MLLMs to reduce resource consumption while maintaining performance [29][54] Lightweight Optimization Methods - The report identifies three core modules of MLLMs: visual encoder, pretrained large language model, and visual-language projector, with optimization efforts focused on these areas [30][54] - Techniques for lightweight optimization include model compression methods such as quantization, pruning, and knowledge distillation, which have been explored in traditional deep learning networks [7][29] Visual Token Compression - Visual token compression is crucial for reducing computational load caused by large token sequences, which is essential for efficient MLLMs [8][57] - The report discusses various methods for multi-scale information fusion to enhance visual feature extraction, allowing models to capture fine-grained details and broader contexts [40] Efficient Structural Design - The report outlines the importance of optimizing model structures or algorithm designs to achieve high performance with fewer resources, focusing on expert mixture models and inference acceleration [9][41] - It mentions the potential of deploying lightweight MLLMs on edge devices, which could significantly enhance the capabilities of intelligent devices and robots [61]

Investment Rating - The report does not explicitly provide an investment rating for the industry Core Insights - The development of 5G and AI technologies has created new opportunities for video encoding technology, which faces challenges due to the emergence of new video content types such as VR and panoramic videos [7][11] - Video semantic encoding technology aims to encode based on video content and semantic features, potentially overcoming the performance limitations of traditional video encoding methods and driving high-quality development in the video industry [23][35] Summary by Sections 1. Overall Development Trends in Video Encoding Technology - The video industry is experiencing significant innovation driven by multimedia communication advancements, with video encoding technology becoming increasingly important [11] - The need for more efficient and intelligent video encoding technology is urgent due to the rapid growth of video data and diverse application scenarios [12] 2. Overview of Video Encoding Technology Development - Video encoding standards have evolved over decades, with a focus on hybrid coding frameworks [28][31] - New generation video encoding standards like VVC and AVS3 offer improved performance but also increased complexity, posing challenges for real-time encoding [31] 3. Key Technologies in Video Semantic Encoding Transmission - Video semantic encoding utilizes semantic information to enhance encoding efficiency, with various technical routes including visual perception coding, generative coding, and machine vision coding [47][72] - The integration of AI in video encoding processes is highlighted as a future direction for enhancing performance and efficiency [49][61] 4. Standardization Progress and Recommendations - The report discusses the need for standardization in AI video encoding, VR video encoding, and multi-view video encoding to support industry growth [10][36] 5. Summary and Outlook - The report concludes that video semantic encoding technology is essential for addressing the challenges posed by the explosive growth of video data and diverse application scenarios, positioning it as a new driving force for the video industry [23][75]