As the global market expands its investment in AI computing and the development of related applications, humanoid robots are expected to become the largest AI application by market size. The introduction of torque and inertia sensing technology by manufacturers is critical for humanoid robots to walk flexibly and mimic human behavior in various scenarios. These two sensor systems need to work perfectly in sync, driving robots from simple programmed actions towards more natural human behavior. Humanoid robots need to operate in complex and variable environments and possess reaction speeds similar to humans. Their sensing systems must accurately perceive and respond to their surroundings. Currently, several robot manufacturers are specializing in improving their walking performance.
DeepSeek’s rapid rise not only highlights the role of algorithm optimization in balancing model performance and hardware efficiency but also signals a shift in AI development toward inference-driven applications. AI agents are expected to become the primary form of such applications, gradually evolving into systematic Agentic AI, enabling greater automation and goal-oriented task execution.
Near-eye displays (NEDs) adopt different optical engine modules and display components due to variations in interaction functionality and immersive experience requirements across practical application scenarios. NED are further categorized into virtual reality/mixed reality (VR/MR) and augmented reality (AR) devices. Typically, NEDs that achieve an immersive experience through a single display component are classified as VR/MR devices. Devices that emphasize transparency and pass-through and focus on hybrid virtual-physical application scenarios are classified as AR devices. With increasingly diverse metaverse applications and enhanced AI functionality, the overall shipment volume of NEDs, after experiencing a temporary downturn due to inventory consumption, is expected to steadily increase year by year. NED shipments are projected to reach 62.8 million units by 2030, with a compound annual growth rate (CAGR) of 31% from 2023 to 2030.
Benefiting from deep learning and AIGC technologies, humanoid robots have undergone a remarkable transformation after decades of development. With leading companies successively announcing mass production plans, the humanoid robot market is set to accelerate its expansion. On the demand side, this report explores the potential demand for humanoid robots. On the supply side, this report focuses on the developments of leading companies, analyzes market trends, and examines the potential roles that Taiwanese companies may play.
The increase of smart home appliances and additional projects supported by Matter, as well as the emergence of generative AI, have prompted the improvement of human-computer interaction to be the key development for the next phase of smart homes. Smart speakers, as carriers of generative AI assistants, could potentially become the entry of home applications for large models that would in turn improve user experience.
1. Global Trends in Mobile Base Station Development
2. AI Applications: Mobile Base Stations
3. TRI’s View
1. Overview of Global Memory Supply
2. Analysis of Mobile Memory Demand
3. Overview of the Smartphone Market
4. TRI’s View
As international geopolitical risks rise, Apple is beginning to face the issue of supply chain relocation. Since 2000, Apple has established a complete supply chain layout in China. Before the outbreak of the epidemic, the assembly factory located in Zhengzhou of Henan provided sufficient production capacity for iPhones every year, allowing the products to be exported worldwide. However, the factory's capacity utilization rate continued to decline after the outbreak of the epidemic. In November 2022, a large-scale protest erupted due to some employees' dissatisfaction with wages and epidemic restrictions, leading to serious problems in the supply chain. Apple realized that it could no longer concentrate a large amount of production capacity on a single supplier or region, so it accelerated the relocation of its supply chain.
Demand for large panel driver ICs is set to rise sequentially throughout 2024, with ample overall supply. However, uncertainty in end-user demand has led to conservative stocking practices in the supply chain, fostering a reliance on short-term, rush orders—a practice that threatens long-term supply chain health.
In 2023, mobile TDDI chips benefited significantly from active repair and second-hand markets, resulting in annual shipments far exceeding intial forecasts. The release of lower-cost capacities by Chinese wafer foundries has helped sustain robust demand for mobile TDDIs through 2023 to 2024. Nonetheless, intense comptition in this segment has escalated price wars.
For AMOLED driver ICs, the introduction of new production capacities has boosted demand as AMOLED smartphone panel penetration increases. Currently, the market faces no-short-term supply shortages. As more suppliers enter the AMOLED DDI market, prices continue to decline. Maintaining profitability in AMOLED DDI production remains a critical challenge for IC manufacturers, especially with stable wafer prices.
The anticipation of AI-Generated Content (AIGC) applications approaching large-scale commercial deployment has triggered a sharp rise in demand for AI chips, significantly impacting the global high bandwidth (HBM) market. This report will provide analysis from two fronts: First, it will explore the development of AIGC applications and large language models (LLMs). Then, it will examine the evolving needs for ASICs, GPUs, and HBM crucial for AI-powered cloud computing.
