Investment Rating - The report assigns an "Overweight" rating to the computer industry, indicating an expected price increase exceeding 10% over the next 6-12 months compared to the market benchmark index [7]. Core Insights - Xiaomi has launched three new models in the MiMo-V2 family: MiMo-V2-Pro, MiMo-V2-Omni, and MiMo-V2-TTS, with MiMo-V2-Pro being a flagship model designed for high-intensity agent work scenarios [1][2]. - MiMo-V2-Pro features over 1 trillion total parameters (42 billion active parameters) and supports a context length of 1 million tokens, showcasing significant advancements in AI capabilities [2]. - The model ranks eighth globally and second domestically on the Artificial Analysis leaderboard, demonstrating superior end-to-end task completion capabilities in various frameworks [2]. - Xiaomi claims that the user experience of MiMo-V2-Pro surpasses that of Claude Sonnet 4.6 while being priced at only one-fifth of Claude Opus 4.6, significantly lowering the barrier to access advanced AI [2]. Summary by Sections Product Launch - Xiaomi officially released the MiMo-V2 family models on March 19, confirming that the previously high-profile anonymous model Hunter Alpha is an early internal version of MiMo-V2-Pro, while Healer Alpha corresponds to MiMo-V2-Omni [1]. Technical Specifications - MiMo-V2-Pro has tripled the parameter count compared to its predecessor MiMo-V2-Flash and improved the hybrid attention mechanism ratio from 5:1 to 7:1, maintaining high inference efficiency while supporting ultra-long context [2]. Market Trends - On March 18, Alibaba Cloud announced price increases for AI computing and storage products due to rising global AI demand and supply chain costs, with increases up to 34% [3]. - Baidu and Tencent Cloud also announced similar price adjustments, reflecting the ongoing trend of rising costs in AI infrastructure [3].

一、盘前资讯 1、要闻简讯 一、证监会：坚决清除影响资本市场改革发展的"拦路虎""绊脚石"（证券时报） 3月17日，证监会召开工作会议，会议强调，2026年是"十五五"开局之年，要扎实推进会系统和资本市场廉洁文化 建设，持续发力铲除腐败滋生的土壤和条件，进一步深化重点领域腐败问题专项治理，突出重点人员、重点问 题，加大案件查办力度，特别是严肃查处破坏资本市场秩序、侵害中小投资者利益的人和事，坚决清除影响资本 市场改革发展的"拦路虎""绊脚石" 一、盘前资讯 1、要闻简讯 一、证监会：坚决清除影响资本市场改革发展的"拦路虎""绊脚石"（证券时报） 3月17日，证监会召开工作会议，会议强调，2026年是"十五五"开局之年，要扎实推进会系统和资本市场廉洁文化 建设，持续发力铲除腐败滋生的土壤和条件，进一步深化重点领域腐败问题专项治理，突出重点人员、重点问 题，加大案件查办力度，特别是严肃查处破坏资本市场秩序、侵害中小投资者利益的人和事，坚决清除影响资本 市场改革发展的"拦路虎""绊脚石"。 3月18日，科技部副部长在国新办新闻发布会上表示，要建立健全创新资源向企业集聚的政策，完善院校、企业、 科研人员双向流动、互相 ...

Core Viewpoint - The rapid development of global AI applications has led to a significant increase in demand for computing power, prompting major cloud service providers like Baidu and Alibaba Cloud to raise prices for their AI computing and storage products by 5% to 34% [1][3]. Group 1: Price Adjustments - Baidu announced a price increase of approximately 5% to 30% for AI computing services and a 30% increase for parallel file storage services, effective from April 18, 2026 [1]. - Alibaba Cloud's price adjustments include a maximum increase of 34% for AI computing and storage products, with specific products like the PingTouGe Zhenwu 810E computing card seeing increases between 5% and 34% [3]. Group 2: Market Reactions - Following the price hikes, Alibaba's stock surged over 4% in Hong Kong, with other companies like Kingsoft Cloud and MINIMAX also experiencing significant stock price increases of over 20% [3]. - The A-share computing power industry saw a strong performance, with concepts related to computing power leasing and storage chips gaining traction [3]. Group 3: Token Consumption and Demand - The surge in token consumption, particularly for the OpenClaw application, has led to a doubling of platform processing volume to 14.8 trillion tokens in the first week of March 2026 compared to the beginning of the year [6]. - The increase in token consumption indicates a shift in AI industry value from model capability to task completion capability, creating structural impacts on computing power infrastructure and driving demand for both centralized and edge computing [6]. Group 4: Industry Outlook - With the explosion of token consumption in the domestic market, domestic computing power is expected to gradually dominate the infrastructure layer due to cost advantages and a well-developed ecosystem [6].