TechnologyNews· Developing story

Nvidia and Microsoft expand AI computing from Azure to Windows PCs

The companies announced cloud, on-premises and device options at Microsoft Ignite, with Azure virtual machine access in preview.

The entrance to Nvidia headquarters, with the company’s sign beside a staircase.
Nvidia headquarters in Santa Clara, California, in 2018. File photograph. NVIDIA Headquarters — Coolcaesar; corrected by Jacek Halicki (resized and converted to WebP). CC BY-SA 4.0.Image sourceImage licence: CC BY-SA 4.0
WhatsAppXLinkedInEmail
Save for later

Nvidia and Microsoft announced an expansion of their AI computing collaboration at Microsoft Ignite on November 19, extending options from Azure data centres to local systems and Windows PCs. The changes give organisations more ways to build and run AI applications in the cloud and closer to their own data.

Nvidia said Vera Rubin NVL72 systems are running in Azure data centres, with access to virtual machines powered by them expanding through Azure’s Vera Rubin preview. The infrastructure is designed for demanding workloads including large-scale AI model training and inference.

The companies also plan to demonstrate local large language model experiences on Windows PCs powered by Nvidia RTX Spark at Ignite. Crypto Briefing described the broader approach as keeping computing capability on PCs while retaining access to cloud services.

Azure infrastructure and model customisation

The companies are also working to extend confidential computing to bare-metal and rack-scale AI infrastructure, according to Nvidia. Its confidential computing and attestation technology helps customers verify a computing environment before releasing sensitive models, data or keys.

Through Microsoft Foundry, developers can use serverless Programmed Training to control post-training workflows while the service manages the underlying infrastructure. Foundry Managed Compute is adding support for deploying custom model weights, connecting customisation with deployment through Microsoft’s managed services.

Nvidia said EY used Programmed Training to adapt Nemotron open models to classify corporate purchasing records. On EY’s evaluation dataset, Nvidia reported a 55% improvement in accuracy and a 15-fold reduction in cost compared with proprietary frontier models.

Local AI and Windows 365 previews

Expanded Nemotron 3.5 support in Foundry Local brings reasoning, coding and tool-using capabilities to Azure Local environments, Nvidia said. Azure Local will also support RTX PRO 6000 Blackwell GPUs from hardware partners, giving organisations another option for running AI on premises.

Azure Local is expanding to Nvidia’s Jetson Thor and Jetson Orin platforms for industrial uses such as machine vision and robotics. Nvidia’s Cosmos physical AI models are also being brought to Foundry Local.

For GPU-accelerated Windows 365 Cloud PCs, the companies are bringing Direct Stream capabilities intended for demanding graphics applications. Nvidia said they support up to two 4K streams at 60 frames per second. A limited preview is open for sign-ups, with a public preview planned for early 2027.

Microsoft AI is also using Nvidia’s Dynamo software during post-training to coordinate inference across a large fleet of GPUs, Nvidia said. The software helps manage requests, model updates and recovery from failures.

Read in depth: Technology

What this story rests on

3 sources from 3 organisations: 3 news outlets.

  1. 1NVIDIA and Microsoft Expand Accelerated Computing From Azure to Windows PCsNVIDIA · blogs.nvidia.comNews outlet
  2. 2Microsoft and Nvidia expand accelerated computing from Azure to Windows PCsCrypto Briefing · cryptobriefing.comNews outlet
  3. 3NVIDIA Expands Microsoft AI Collaboration with Vera Rubin and NemotronUnite.AI · unite.aiNews outlet

AI-assisted article checked against the listed sources. NewsJaws did not conduct interviews or attend the reported events.

Researched and drafted with AI tools from these sources, then checked against them paragraph by paragraph before publication. How we work · Everyone we cite · Report an error

Topics in this story

Add NewsJaws as a preferred source on Google

About NewsJaws Desk

AI-assisted reporting and explainers reviewed against the linked source documents. No claim of on-scene reporting or original interviews.

More from NewsJaws Desk

More from Technology

See all