Core Insights - Oracle plans to build a new data center utilizing 50,000 AMD chips to enhance its AI capabilities, marking a significant move in the AI sector [1][2] - The partnership with AMD is seen as a strategic choice for Oracle, particularly in the inference space, as they have a long-standing relationship and a robust software stack [3][4] - Oracle's investment in AMD chips is a substantial commitment, reflecting founder Larry Ellison's vision for the company's competitive stance against major players like Microsoft, Amazon, and Google [5][6] Company Developments - Oracle is deploying AMD's latest AI chips to create a supercluster aimed at enabling customers to run larger and more complex AI models by the second half of next year [2] - The market is responding positively to AMD, with its shares rising as it is increasingly recognized as a viable competitor in the inference market, especially following OpenAI's recent investment in AMD [4] - Larry Ellison is expected to provide insights into a $300 billion deal with OpenAI and discuss Oracle's new cloud offerings that aim to enhance efficiency and speed for customers [5][6] Financial Considerations - Oracle's financial analyst day will reveal how the company plans to fund its significant chip purchases, indicating the importance of this week for Oracle's financial strategy [6]

We're back on Morning Trade Live. Perhaps lost in the start of earnings season this week. Oracle is in the spotlight of the next couple of days in Las Vegas as the company's AI World 2025 event is underway.The stock has rallied nearly 30% since early September after the company projected massive cloud and AI growth over the next few years. So for more on Oracle, joining us now, I'm pleased to say, Steven Dickens, CEO and principal analyst, Hyperframe Research. It's always such a pleasure to talk to you, Ste ...

LewisTsePuiLung/iStock Editorial via Getty Images Listen below or on the go on Apple Podcasts and Spotify J.P. Morgan, Goldman, Wells Fargo top estimates. (0:16) Oracle deploying 50,000 AMD chips. (1:41) Geopolitical arbitrage increasing as Americans seed golden visas. (2:14) This is an abridged transcript of the podcast: Our top story so far, big banks are kicking off earnings season with consensus-bearing results – as usual. Among the results for the heavy hitters: JPMorgan Chase (NYSE:JPM) nudged ...

My conviction in Amazon.com (NASDAQ: NASDAQ: AMZN ) has grown following the company's recent execution and also management guidance that points to continued acceleration in AWS. While Q2 2025 results confirmed AWS's strong 17.5% y/y andKennedy is a GARP-themed investor with a bias towards companies with aggressive growth prospects, en route to becoming highly profitable in 1-2 years.His investment philosophy emphasizes long-term discipline, consistent alpha, and a pinch of salt (risk).He writes to empower t ...

Key Points Alphabet has multiple growth avenues, from digital ads to AI, that should boost its results. Veeva Systems has found success by focusing on a small corner of the giant cloud market. 10 stocks we like better than Alphabet › There aren't many guarantees in life, but the stock market's ability to deliver competitive returns over long periods may be one of them. So, despite some volatility related to the threat of tariffs and other factors this year, it's still a good idea to invest in stock ...

Valued at a market cap of $2.3 trillion, Amazon.com, Inc. (AMZN) is a leading global technology company that operates in multiple industries, including e-commerce, cloud computing, advertising, digital streaming, and AI. The Seattle, Washington-based company is scheduled to announce its fiscal Q3 earnings for 2025 in the near future. Ahead of this event, analysts expect this e-commerce giant to report a profit of $1.58 per share, up 10.5% from $1.43 per share in the year-ago quarter. The company has a pro ...

大语言模型（LLM）与强化学习（RL）的融合趋势，正加速催生 "自主智能体"（能自主决 策、执行任务的 AI 系统）的发展。 目前 CoreWeave 的业务覆盖范围正快速扩张，这使其能将基础设施与服务推向更多市场和企 业，为智能体时代的规模化服务奠定基础。 从 " 算 力 供 应 商 " 到 " 智 能 体 运 行 平 台 " 收购 OpenPipe 是 CoreWeave 向 "价值链上游" 突破的关键动作。 OpenPipe 的核心竞争力是一套 "强化学习工具包"—— 开发者可借助它训练智能体，还能让 模型适配新任务需求。 此次收购后，CoreWeave 不仅掌握了智能体训练的核心技术，更获得了开发者群体的认可， 彻底打通了智能体训练的全流程。 这并非 "小幅升级"，而是从 "硬件层 + API 接口" 到 "智能体全周期支持平台" 的质变。 智能体相关工作负载呈指数级增长，算力需求持续飙升； 自研强化学习工具与运行时服务（Runtime）将显著扩大利润率； 在电力供应、散热效率与 GPU 资源获取上，相比超大规模云厂商（Hyperscalers）具 备持久竞争优势。 过去，每个开发团队都需自行搭建 " ...

Core Insights - Oracle has announced new networking capabilities in Oracle Acceleron, enhancing data movement with direct paths, reducing overhead, and providing high-bandwidth, ultra-low-latency connections [1][2] - The new features aim to help customers run workloads faster and more cost-effectively, leveraging over a decade of cloud networking innovation [1][3] Networking Capabilities - Oracle Acceleron integrates dedicated network fabrics, direct data paths, and a converged NIC to enhance performance and reduce costs, while also implementing host-level zero-trust routing and multi-planar designs for improved security and resilience [2][3] - The architecture allows for line-rate encryption throughput, ultra-low latency, and up to 2x potential network processing capacity and storage IOPS, benefiting a wide range of OCI customers [2] Fabric Network Architecture - The dedicated fabric network architecture provides predictable low-latency, high-bandwidth connectivity by isolating traffic and supporting multiple fabrics, optimizing throughput and latency for various workloads [7] - Multi-planar networking enhances network resiliency and flexibility by connecting customer NICs to multiple isolated network planes, allowing for instant traffic shifts in case of issues [7] Converged NIC and Security Features - The converged NIC increases throughput and lowers costs while maintaining strong isolation between customer and provider planes, enabling storage acceleration and line-rate encryption [8] - Zero-Trust Packet Routing (ZPR) secures workloads with least-privilege enforcement and simplifies policy management, enhancing security for sensitive endpoints [8][9] Industry Collaboration - Industry leaders like Arista Networks and AMD are leveraging Oracle Acceleron to enhance AI initiatives, improve efficiency, and strengthen security, indicating strong partnerships and collaborative innovation [10]

Core Insights - Oracle has announced the launch of OCI Zettascale10, the largest AI supercomputer in the cloud, capable of delivering up to 16 zettaFLOPS of peak performance [1][2][3] - The supercomputer is built on Oracle's Acceleron RoCE networking architecture and NVIDIA AI infrastructure, designed to support multi-gigawatt AI workloads with improved efficiency and reliability [1][3][4] Group 1: Technical Specifications - OCI Zettascale10 connects hundreds of thousands of NVIDIA GPUs across multiple data centers, forming multi-gigawatt clusters optimized for low GPU-GPU latency [1][2] - The architecture is hyper-optimized for density within a two-kilometer radius, enhancing performance for large-scale AI training workloads [2] - Initial deployments will target up to 800,000 NVIDIA GPUs, ensuring predictable performance and strong cost efficiency [3][5] Group 2: Strategic Partnerships and Development - The supercomputer is part of a collaboration with OpenAI at the Stargate site in Abilene, Texas, focusing on advancing AI capabilities [1][3] - Oracle and NVIDIA are combining their technologies to provide a robust compute fabric for state-of-the-art AI research and industrial applications [4] Group 3: Networking Innovations - Oracle Acceleron RoCE networking architecture enhances scale, reliability, and efficiency for AI workloads, allowing GPUs to connect to multiple switches simultaneously [4][11] - Key features include a wide, shallow, resilient fabric that reduces costs and power while increasing deployment speed for larger AI clusters [4][11] - The design improves reliability by eliminating data sharing across network planes, maintaining stability for AI jobs [4][11]

First cross-cloud consumption model  will enable customers to quickly and easily buy Oracle AI Database and OCI services in the cloud of their choice Customers will be able to streamline procurement with flexible terms and consistent contracts across AWS, Google Cloud, Microsoft Azure, and OCI , /PRNewswire/ -- Oracle AI World -- Oracle today introduced Oracle Multicloud Universal Credits, a new licensing option that will enable customers to procure Oracle AI Database and Oracle Cloud Infrastructure (OC ...