TellWorldAI World Archive

Home›News›Partnership

PartnershipGenerative AI for everyone

NVIDIA and Microsoft expand accelerated computing from Azure to Windows PCs at Ignite

On 19 November 2026 NVIDIA said its expanded collaboration with Microsoft, announced at Microsoft Ignite, covers Azure data centers, open models, edge hardware and Windows PCs.

Background

NVIDIA and Microsoft announced an expanded collaboration at Microsoft Ignite. NVIDIA published its account on 19 November 2026. The collaboration covers Azure infrastructure, open models, edge deployment and Windows PCs [1]. Azure is Microsoft's cloud computing platform, where customers rent computing power over the internet. NVIDIA makes GPUs, the processors widely used to train and run large AI models.

What happened

NVIDIA said that NVIDIA Vera Rubin NVL72 systems are running in Azure data centers, and that Azure is expanding access to virtual machines powered by those systems [1]. A virtual machine is a software-based computer that a customer can rent from a cloud provider and use as if it were its own.

Key details

On benchmark results, NVIDIA said that in MLPerf Training v6.1, Azure scaled to 10,240 NVIDIA GPUs with GB200 NVL72 systems on the DeepSeek-V3 model, the largest scale in that round [1]. MLPerf is a set of standard benchmarks for measuring how fast hardware trains and runs AI models. In MLPerf Inference v6.1, Azure's GB300 NVL72 systems led cloud submissions on DeepSeek-R1 interactive throughput by over 10%, according to Signal65 [1].

On models, NVIDIA said Nemotron 3.5 Super and Lightning are available on Microsoft Foundry Managed Compute, while Nemotron 3.5 Bolt and Transcribe are coming soon [1]. Nemotron is NVIDIA's family of open models.

One customer example is EY. EY used Foundry Programmed Training to adapt Nemotron open models for classifying corporate purchasing records. NVIDIA said this improved accuracy on EY's evaluation dataset by 55% and lowered cost 15x compared with proprietary frontier models [1].

For edge deployment, expanded Nemotron 3.5 support in Foundry Local brings the models to Azure Local, which is also expanding to NVIDIA Jetson Thor and Jetson Orin [1]. Edge deployment means running software on local hardware close to where data is produced, rather than only in a central cloud.

For PCs, Windows 365 Direct Stream for GPU-accelerated Cloud PCs is in limited preview, with a public preview planned for early 2027 [1]. A Cloud PC is a Windows desktop hosted in the cloud and streamed to the user's device.

Impact

The announcement puts cloud data centers, open models, edge hardware and Windows PCs under one collaboration. Some items are already available, such as the Nemotron 3.5 Super and Lightning models on Foundry. Others are still in preview or coming soon, and each is described with its current status.

Sources

  1. 1

Cite this page: NVIDIA and Microsoft expand accelerated computing from Azure to Windows PCs at Ignite. TellWorld, https://www.tellworld.net/news/nvidia-microsoft-ignite-2026. Updated 2026-11-20

Related news

View all