华尔街点评AMD财报：运营支出限制盈利能力，市场焦点转向下半年机架规模AI

Core Viewpoint - AMD's Q4 2025 revenue exceeded expectations, but concerns remain regarding high operational expenditures limiting profitability release [1][3][12] Financial Performance - AMD reported Q4 2025 revenue of $10.27 billion, a 34% year-over-year increase and an 11% quarter-over-quarter increase, surpassing the market expectation of $9.67 billion [3][7] - The revenue growth was primarily driven by the data center business, which generated $5.38 billion in Q4, a 24% quarter-over-quarter increase [3][10] - Excluding a one-time special sales revenue of approximately $390 million, the actual revenue growth was only slightly above the guidance upper limit [3][10] Operational Expenditures - Operational expenditures reached $3.05 billion, significantly higher than the market expectation of $2.94 billion, marking several consecutive quarters of exceeding expectations [7][12] - Analysts express concern that high operational expenditures will continue to limit AMD's operational leverage improvement [4][12] Future Guidance - AMD's Q1 2026 revenue guidance is approximately $9.8 billion, above the market consensus of $9.4 billion, despite a projected 5% quarter-over-quarter decline [6][7] - The data center business is expected to see quarter-over-quarter growth, with server CPU revenue anticipated to achieve double-digit growth in the traditional off-season [6][10] Product Development and Market Position - AMD confirmed the MI455 series GPU will be delivered to customers as planned in Q3 2026, with significant revenue growth expected in Q4 2026 [9][10] - Key clients, including OpenAI and Oracle, are expected to deploy AI computing clusters based on the MI-4XX architecture, supporting AMD's long-term revenue goals [9][10] Market Sentiment and Analyst Ratings - Major investment banks maintain a neutral rating on AMD, citing that while growth potential exists, current valuations reflect this, and short-term operational efficiency remains a core constraint [12][13] - Concerns persist regarding execution risks associated with upcoming AI products and potential customer concentration issues [13]