Jensen Huang and Satya Nadella's Conversation at Microsoft Build

2026-06-03 · Source: NVIDIA · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Robotics & Autonomous Systems · Depth: Advanced, medium

Summary

Jensen Huang and Satya Nadella discussed the evolving NVIDIA-Microsoft AI partnership, focusing on edge AI and cloud innovation. They highlighted the transformation of PCs into "personal AIs" via systems like RTX Spark, which offers a petaflop of AI performance and 128 GB of memory to run hundreds of billions of parameter models. On the cloud side, their collaboration built the first AI supercomputers, evolving from Ampere and Hopper for pre-training to Grace Blackwell for post-training and inference, achieving a 30x token generation rate increase over Hopper. The upcoming Vera Rubin system is designed specifically for agentic AIs, emphasizing extremely low latency and confidential computing with encrypted data in transit and use. This partnership also extends to accelerating all Azure tools, including Fabric, SQL, and Spark, with GPUs to meet the surging demand for agentic compute.

Key takeaway

For CTOs and AI/ML Directors evaluating future AI infrastructure, NVIDIA and Microsoft's deep collaboration signals a critical shift towards agentic systems on both edge and cloud. Your strategy should account for personal AI PCs like RTX Spark and cloud systems like Vera Rubin, which prioritize low-latency, confidential computing, and accelerated tools. Plan for integrated hardware-software stacks to maximize agent performance and cost-efficiency, ensuring your deployments are ready for the next generation of autonomous AI.

Key insights

NVIDIA and Microsoft are transforming PCs into personal AIs and scaling cloud agentic systems through deep collaboration.

Principles

• AI is now useful due to agentic systems and powerful models.
• The entire rack can function as one computer via NVLink 72.
• CPUs are being redesigned for low-latency agentic workloads.

In practice

• Run autonomous AI agents directly on RTX Spark PCs with 128 GB memory.
• Deploy large language models (hundreds of billions of parameters) on edge devices.
• Utilize GPU-accelerated tools on Azure for faster agent iteration and token generation.

Topics

• Edge AI
• Personal AI
• Agentic Systems
• Cloud AI Infrastructure
• NVIDIA-Microsoft Partnership
• GPU Acceleration

Best for: Investor, Executive, AI Architect, CTO, VP of Engineering/Data, Director of AI/ML

Related on AIssential

• Nvidia introduces Vera Rubin, a seven-chip AI platform with OpenAI, Anthropic and Meta on board — Nvidia's Vera Rubin platform and ecosystem target agentic AI with a comprehensive, integrated hardware and software stack.
• Dell’s Microsoft/AMD collaboration: Three insights you may have missed from theCUBE’s coverage of Dell Technologies World — AI cost and architectural demands drive strategic partnerships for hybrid, optimized infrastructure solutions.
• Building the AI Grid with NVIDIA: Orchestrating Intelligence Everywhere — AI-native services are exposing a new bottleneck in AI infrastructure, shifting the challenge from peak training throughput to delivering deterministic inference at scale with predictable latency and…
• Nvidia stretches compute power to Windows PCs in support of agents — Nvidia is expanding its full-stack AI compute strategy by bringing powerful agentic AI capabilities directly to Windows PCs and deskside supercomputers.
• The Nvidia AI PC, Project Solara, Microsoft AI — The AI landscape is shifting from local inference PCs to cloud-centric agent platforms and customizable enterprise models.
• NVIDIA GTC Taipei 2026 Moments: Agents, AI PCs, and Physical AI — NVIDIA is driving AI innovation through new hardware, software, and agent development tools.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by NVIDIA.