Morgan Stanley released the report 'Global Technology: Next-Generation Storage Innovation' on July 16, stating that the key bottleneck in AI development is shifting from 'computing scalability' to the 'storage wall.' The report emphasizes that the surging demand for data volume in AI models, coupled with systemic limitations in storage system capacity, bandwidth, and cost, is constraining AI expansion speed due to storage performance. Morgan Stanley believes that storage has become the core of AI infrastructure competition, and future investment logic will shift from computing chips like GPUs to the entire storage ecosystem, encompassing advanced processes, packaging, interconnects, in-memory computing, and new materials. The report projects that by 2030, Agentic AI will account for 26% to 77% of the new global DRAM demand, cloud storage spending will rise to $418 billion, with a compound annual growth rate of approximately 8% from 2026 to 2030. Storage's share in cloud providers' capital expenditures will increase from 12% in 2023 to 40% in 2027. Morgan Stanley raised its target price for Montage Technology (HK stock: HK$310 → HK$432; A stock: RMB 274 → RMB 377) and forecasts that the market size for new storage types excluding HBM will expand from $1.2 billion in 2025 to $23 billion in 2030. Including HBM, the overall market size could reach $276 billion.