Post-Cutoff

Official

Satya Nadella and Jensen Huang discuss Hybrid Intelligence and the future of PC

WindowsYouTube45,122 views as of 11 October 2026

Watch on YouTubePlay loads YouTube’s player from youtube-nocookie.com.

Why it is here

Official Windows channel: Nadella and Jensen Huang on Windows as the home for ‘hybrid intelligence’, moderated by former White House AI adviser Sriram Krishnan; ~45k views.

Description

Description written by Gemini from the videoGemini 3.8 Flash, 11 October 2026

Summary
This video captures a stage panel moderated by Sriram Krishnan featuring NVIDIA CEO Jensen Huang and Microsoft CEO Satya Nadella at an official Microsoft Windows event. The discussion covers the history of NVIDIA GPUs and Windows, agent security primitives like Microsoft Execution Containers (MXC) and OpenShell, hybrid intelligence spanning local and cloud models, and newly announced hardware including the Surface Laptop Ultra and RTX Spark workstations.

What is shown

  • [00:00 - 00:50] Moderator Sriram Krishnan introduces the session and mentions testing the Surface Laptop Ultra.
  • [00:54 - 02:45] Jensen Huang reflects on NVIDIA’s founding alongside Windows 3.1 in 1992–1993, Direct3D, and co-inventing programmable shaders with DirectX 8 that paved the way for CUDA.
  • [03:00 - 05:15] Satya Nadella recounts early cloud collaboration on InfiniBand in Azure HPC that later powered OpenAI’s training clusters.
  • [05:20 - 06:33] Jensen discusses planning a new PC architecture four years prior to run local AI agents alongside heavy creative software like Blender, Unreal Engine, and Omniverse.
  • [06:50 - 09:35] Satya explains the rationale behind MXC (Microsoft Execution Containers), detailing isolation at the microVM/session level, agent identity, auditability/observability, and governance.
  • [10:12 - 13:20] Satya details “hybrid intelligence,” referencing local/cloud pipelines demonstrated earlier (using ComfyUI, Blender, Photoshop, and cloud models like Astra).
  • [14:44 - 18:14] Jensen describes the hardware architecture of the dual-chip superchip inside the Surface Laptop Ultra and Spark machines, contrasting its form factor with the historical DGX-1.
  • [19:35 - 20:28] Jensen notes industry-wide testing across OEM laptop and workstation makers.
  • [20:30 - 23:35] Satya and Jensen debate the delegation of authority, liability, and trust mechanisms for personal agents.

Claims & numbers

  • Jensen states NVIDIA was founded in 1993, inspired by Windows 3.1 in 1992.
  • Jensen notes the collaboration on the new PC architecture took four years of planning and execution.
  • Jensen claims the engineering effort spanned over 4,000 engineering years between the two companies and industry partners.
  • Jensen states the dual-chip processor delivers 1 petaflops of compute, matching the original DGX-1 server, which weighed 500 pounds.
  • Jensen claims they performance- and compatibility-tested over 1,200 intensive applications for the new platform.
  • Satya states that agent security requires containment across every level—microVM, VM, and the OS session—alongside complete identity tracking and token spend monitoring.

Notable quotes

  • [00:54] Jensen Huang: “Nvidia was founded because of Windows. It was 1992, late 92, Windows 3.1, and we were just starting to see the vision of what a personal computer really can be.”
  • [07:55] Satya Nadella: “We needed to make the desktop the most secure place for agents to execute, and so that’s where MXC comes from.”
  • [17:54] Jensen Huang: “It has one petaflops. Just to put in perspective, the computer that started the AI revolution called DGX-1 was one petaflops... and now it’s in your laptop.”

Assessment
This is an authentic, unedited recording of an executive fireside panel from an official Microsoft Windows hardware and platform launch event. No live benchmark runs or live desktop software are demonstrated in this segment, as the speakers reflect on and discuss announcements and demos presented earlier in the keynote.

Described by gemini-3.8-flash on 2026-10-11 from the video’s audio and frames.

Related

  1. Chips & compute 99 days after the cutoff

    Microsoft and Nvidia launch RTX Spark Windows PCs for local AI