Core Insights - IBM z17 is designed to redefine AI at scale, enabling enterprises to process 100% of their transactions in real-time and handle 50% more AI inference operations per day compared to its predecessor, z16 [1][12] - The system supports over 250 AI use cases across various industries, enhancing business value through applications like loan risk mitigation and medical image analysis [1][2] Product Development - The IBM z17 is the result of five years of design and development, with over 300 patent applications filed and input from more than 100 clients [2] - It introduces multi-model AI capabilities, enhanced security features, and tools for improved system usability and management [2][3] AI Integration - The z17 integrates AI into the core of enterprise operations, providing the necessary software, processing power, and storage to operationalize AI quickly [3] - It features a second-generation on-chip AI accelerator, capable of performing over 450 billion inferencing operations daily with a response time of one millisecond [4][12] Future Enhancements - The IBM Spyre™ Accelerator, expected in Q4 2025, will further enhance AI compute capabilities, particularly for generative AI applications [4][6] - z/OS 3.2, the next version of IBM's operating system, will support hardware-accelerated AI capabilities and modern data access methods, set to release in Q3 2025 [5] Operational Efficiency - IBM Z Operations Unite will streamline operations by consolidating performance metrics and logs, enhancing anomaly detection and incident resolution [5] - The integration of AI assistants and agents aims to improve developer and IT operations efficiency [4][6] Security Features - IBM z17 enhances security and resiliency, incorporating AI-driven processes for incident remediation and case resolution [7][11] - New capabilities for discovering and classifying sensitive data will utilize natural language processing to protect mission-critical data [11] Storage Solutions - The 10th Generation IBM Storage DS8000 is designed to work seamlessly with IBM z17, providing optimized data performance and modular architecture for business growth [8] Availability - IBM z17 is set to be generally available on June 18, 2025, with the IBM Spyre™ Accelerator expected in Q4 2025 [9]

" 戴尔的非 GAAP 预期市盈率低于行业中值。 " 作者 | Noah's Arc Capital Management 编译 | 华尔街大事件 自11月份以来，戴尔( NYSE: DELL )的股价一直落后于市场。股价落后的主要原因是市场担 心人工智能数据中心建设放缓，上周末传闻微软将改变资本支出战略就是明证。 虽然这在市场 某些领域可能确实如此，但关键是推理计算仍将表现良好。当数据中心为此而建时，戴尔可以 成为推理计算的主要赢家。鉴于此，分析师认为股价还有上涨空间。 此前，许多 AI 芯片的效率都太高了，以至于这些基础模型公司不需要那么多芯片就能完成必 要的任务。这导致整体 GPU 需求增长放缓，这显然会损害销售增长预期。 AI 领域的大多数 公司（无论其估值倍数如何）都根据对 AI 设备销售的预期进行交易。戴尔就是其中之一。 现在， 市场担心 Blackwell 支持的预训练集群已经变得高效，而 AI 模型在过去三个月中并没 有呈指数级增长。 每个代币的成本继续下降，这对开发者来说是好事。一些关键因素已经朝着 有利于公司的方向转变。 Atreides Management 的执行合伙人兼首席信息官 G ...

Dell Technologies (DELL) Q3 2025 Earnings Call November 26, 2024 04:30 PM ET Company Participants Paul Frantz - Vice President of Investor RelationsJeffrey Clarke - COO & Vice ChairmanYvonne McGill - Chief Financial OfficerAmit Daryanani - Senior Managing DirectorDavid Vogt - Managing DirectorErik Woodring - Managing Director - Equity ResearchBen Reitzes - Managing Director – Head of Technology ResearchAsiya Merchant - Director - Equity ResearchKrish Sankar - Managing DirectorMichael Ng - Managing Director ...