Microsoft Launches New Nvidia-Powered Laptops for AI Workloads

<p>Microsoft on Wednesday unveiled a new class of Nvidia-powered Windows laptops, aiming to integrate the operating system more closely with the AI sector.</p><p><strong>Why it matters: </strong>Microsoft is attempting to establish a high-end laptop category for users who wish to run demanding AI workloads locally instead of relying solely on cloud services.</p><hr /><p><strong>Driving the news:</strong> Microsoft CEO Satya Nadella, Nvidia CEO Jensen Huang, and former White House AI official Sriram Krishnan were present in San Francisco for the official launch of Microsoft Surface Ultra, a high-end laptop featuring Nvidia's RTX Spark chip, starting at a price of $2,599.</p><ul><li>"We have reinvented the computer as we know it," Huang stated.</li><li>Microsoft previewed the laptop in June but did not disclose specific details regarding pricing and availability at that time. It indicated that Nvidia-powered devices would also be released by companies such as Asus, Dell, HP, and Lenovo. The Surface Ultra and other devices will be available starting October 16.</li><li>On the software front, Microsoft introduced a version of Copilot that can operate using local models in conjunction with those in the cloud, and announced that Meta will bring its Muse personal agent to Windows.</li></ul><p><strong>The big picture:</strong> Microsoft's initial foray into AI-powered Windows machines, referred to as Copilot+ PCs, faced challenges after the delay of Recall, a key feature.</p><p><strong>Between the lines:</strong> Microsoft suggests that developers' demand for AI capabilities surpasses their companies' AI budgets, advocating for a hybrid approach that utilizes both advanced cloud models and effective local models.</p><ul><li>"Hybrid intelligence exists because people's appetite for intelligence is effectively unlimited," stated Microsoft executive VP Pavan Davuluri.</li><li>Before discussing hardware, Microsoft emphasized its efforts to enhance Windows security, ensuring that the actions of AI agents are distinct from those of humans and remain within defined limits.</li></ul>