Core Insights - NVIDIA emphasizes that its AI growth is still in the early stages, particularly in enterprise and industrial applications, which are described as being in a "nascent" phase [1] - The company maintains an "Overweight" rating, supported by strong arguments regarding the stability of its competitive landscape [1] AI Demand and Market Potential - Current AI demand growth is primarily driven by the migration from CPU to GPU in cloud computing capital expenditures, with significant expansion potential ahead [3] - NVIDIA's CEO forecasts that the AI infrastructure market could reach $3 to $5 trillion by 2030, exceeding common market assumptions [3] - The company notes that wafer capacity is no longer a major bottleneck, with current limitations focused on data center space, power, and supporting facilities [3] Strategic Partnerships and Investments - NVIDIA positions itself as an "accelerator" rather than a "financier," investing in projects like CoreWeave and sovereign fund initiatives in London to expedite data center development [3] - The company views OpenAI's recent 10GW data center construction goal as an incremental opportunity, advising caution in interpreting such targets from competitors [3] Competitive Landscape - In response to AMD's collaboration with OpenAI, NVIDIA highlights its own agreements with non-cloud data centers and equity investments per GW, suggesting a more direct approach to market engagement [4] - Concerns regarding NVIDIA's long-term forecasts, financing cycles, and rising competitor confidence are acknowledged, but the company believes that even partial realization of high assumptions in three-year agreements will support future growth [4] - NVIDIA expects to maintain an 80% market share in the XPU market by 2026, with its valuation still lower than competitors [4] AI Infrastructure and Growth Drivers - The AI infrastructure spending is projected to be between $3 to $4 trillion by 2030, driven by three AI scaling laws that are expected to exponentially increase computing demand [6][8] - The company excels in pre-training, post-training, and inference processes, which are critical for AI applications [8] Token Generation and Usage Growth - Token generation is reportedly doubling every two months, indicating a rapid increase in AI application usage [10] - OpenAI has seen significant growth in its user base, with 5 million paying business users as of recent reports, up from 3 million in June [10]
大摩:英伟达NDR点评,AI基础设施市场前景广阔,竞争动态仍具优势