CSC Financial: Focus on high-speed optical modules, AI servers, and data center infrastructure
CSC Financial's research report suggests focusing on high-speed optical modules, AI servers, and data center infrastructure, while also paying attention to order and commercialization progress for physical AI-related sensors, edge computing, and core robot components. Meta's third-quarter revenue was $64.601 billion, up 26% year-over-year, and its full-year capital expenditure guidance was raised to $135–145 billion; Moonshot AI released Kimi K3.1 Pro, supporting a 1 million Token context.
Body · AI translation
Article body
CSC Financial's research report states that Meta's third-quarter revenue was $64.601 billion, up 26% year-over-year, and it raised its full-year capital expenditure guidance to $135–145 billion, lifting the lower bound by $5 billion, continuing to ramp up AI investment. Moonshot AI released Kimi K3.1 Pro, supporting a 1 million Token context and strengthening coding, 3D creation, and Agent capabilities. Unitree Robotics held the World Physical AI Conference, where Wang Xingxing predicted that physical AI may achieve a breakthrough as early as 2027; Tesla's Optimus has started production and is in the early ramp-up phase, while Teradyne and Sereact are collaborating to advance industrial applications. The expansion of overseas AI capital expenditure and upgrades to domestic models continue to support computing power demand, and frontier models are extending into the physical world, which is expected to further open up space for robot applications as well as simulation training, edge computing, and on-device inference. We recommend paying attention to high-speed optical modules, AI servers, and data center supporting facilities, while also attaching importance to order and commercialization progress for physical AI-related sensors, edge computing, and core robot components.
(Source: Yicai) Middle of the article Embedded advertisement