Dave Salvator · Nov 19, 2026
NVIDIA and Microsoft Expand AI Computing Across Azure, Enterprise Tools and PCs
Markdown · JSON · Original source
describes an expanded collaboration with Microsoft spanning Azure infrastructure, enterprise model customization, on-premises and edge deployments, scientific research, and Windows PCs. The companies say Vera Rubin NVL72 systems are running in Azure and describe plans to extend confidential computing protections across AI infrastructure. Microsoft Foundry is adding access to NVIDIA open models and customization and deployment capabilities; NVIDIA reports that EY improved accuracy on its evaluation dataset by 55% and lowered cost by 15x using Programmed Training for a purchasing-record classification task. The announcement also covers Azure Local support for NVIDIA hardware, scientific tools in Microsoft Discovery, and NVIDIA software used in Microsoft AI’s post-training work. These are distinct stages of availability: some capabilities are described as available, while others are previews or coming soon. The article’s examples and performance figures are presented by NVIDIA, and the EY result concerns its evaluation dataset rather than general enterprise performance.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
· Article
Source details
Original source date: Nov 19, 2026 (2026-11-19T20:00:15+00:00).
Date basis: Date stated by the publisher.
Added to Escape Meridian: 2026-11-19T20:17:37.486Z. First observed: 2026-11-19T20:16:12.056Z. Retrieved: 2026-11-19T20:16:59.199Z.
Referenced by
Recorded citations in public readings. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
The announcement is published by and presents the companies’ capabilities and reported results; the supplied text does not include independent verification of those claims. The reported EY accuracy and cost changes are specific to its evaluation dataset and purchasing-record classification task; the text does not give results across other tasks or datasets. Availability varies: the text describes some capabilities as previews, limited previews, or coming soon, rather than generally available.
Open question
How will the reported EY accuracy and cost results compare with other tasks and datasets beyond the purchasing-record evaluation described here?
Source-stated
NVIDIA says the expanded collaboration covers infrastructure, model customization and deployment, and AI at the edge.
the companies are expanding access to Azure virtual machines powered by NVIDIA Vera Rubin NVL72 systems, as well as adding customer availability to NVIDIA open models and customization capabilities through Microsoft Foundry, and new options for running AI at the edge
Source-stated
NVIDIA reports that EY used Microsoft Foundry Programmed Training to adapt Nemotron open models for purchasing-record classification, improving evaluation-dataset accuracy and lowering cost compared with proprietary frontier models.
improving accuracy on its evaluation dataset by 55% and lowering cost by 15x compared with proprietary frontier models.
Source-stated
NVIDIA says Azure Local will support NVIDIA hardware and edge platforms, while Windows PCs powered by NVIDIA RTX Spark will be used to showcase local large language model experiences.
Azure Local is expanding to NVIDIA Jetson Thor and Jetson Orin edge AI platforms to support industrial applications such as machine vision, robotics and autonomous operations.
Source-stated
NVIDIA says Microsoft AI uses NVIDIA Dynamo during post-training, coordinating inference across a large GPU fleet.
Microsoft AI uses NVIDIA Dynamo to support reinforcement learning rollout inference during post-training.
Source and contributor context
Dave Salvator · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
Media declared by the original publisher; availability checked separately from the editorial reading.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
NVIDIA
NVIDIA develops processors, networking and software used for AI, graphics and scientific computing.
Editorial provenance
Version 1. Publication ID: 2d7aac0a-7fac-4f55-8b5f-65d96c1f2f6c.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-19T20:17:37.486Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: e861494232c31d73ebc5ef042a67eb1c0790bf3c2ff4ff9064e60c10eaa23f6e. Source text hash: 522fb3c4f635522f36f60a9cafca6349349cd886adeea583d222437640e0850d.
Revision and source-check history
Version 1 · 2026-11-19T20:17:37.486Z
NVIDIA and Microsoft Expand AI Computing Across Azure, Enterprise Tools and PCs
describes an expanded collaboration with Microsoft spanning Azure infrastructure, enterprise model customization, on-premises and edge deployments, scientific research, and Windows PCs. The companies say Vera Rubin NVL72 systems are running in Azure and describe plans to extend confidential computing protections across AI infrastructure. Microsoft Foundry is adding access to NVIDIA open models and customization and deployment capabilities; NVIDIA reports that EY improved accuracy on its evaluation dataset by 55% and lowered cost by 15x using Programmed Training for a purchasing-record classification task. The announcement also covers Azure Local support for NVIDIA hardware, scientific tools in Microsoft Discovery, and NVIDIA software used in Microsoft AI’s post-training work. These are distinct stages of availability: some capabilities are described as available, while others are previews or coming soon. The article’s examples and performance figures are presented by NVIDIA, and the EY result concerns its evaluation dataset rather than general enterprise performance.
Original source
· Article
Source details
Original source date: Nov 19, 2026 (2026-11-19T20:00:15+00:00).
Date basis: Date stated by the publisher.
Added to Escape Meridian: 2026-11-19T20:17:37.486Z. First observed: 2026-11-19T20:16:12.056Z. Retrieved: 2026-11-19T20:16:59.199Z.
Limits
The announcement is published by and presents the companies’ capabilities and reported results; the supplied text does not include independent verification of those claims. The reported EY accuracy and cost changes are specific to its evaluation dataset and purchasing-record classification task; the text does not give results across other tasks or datasets. Availability varies: the text describes some capabilities as previews, limited previews, or coming soon, rather than generally available.
Open question
How will the reported EY accuracy and cost results compare with other tasks and datasets beyond the purchasing-record evaluation described here?
Source-stated
NVIDIA says the expanded collaboration covers infrastructure, model customization and deployment, and AI at the edge.
the companies are expanding access to Azure virtual machines powered by NVIDIA Vera Rubin NVL72 systems, as well as adding customer availability to NVIDIA open models and customization capabilities through Microsoft Foundry, and new options for running AI at the edge
Source-stated
NVIDIA reports that EY used Microsoft Foundry Programmed Training to adapt Nemotron open models for purchasing-record classification, improving evaluation-dataset accuracy and lowering cost compared with proprietary frontier models.
improving accuracy on its evaluation dataset by 55% and lowering cost by 15x compared with proprietary frontier models.
Source-stated
NVIDIA says Azure Local will support NVIDIA hardware and edge platforms, while Windows PCs powered by NVIDIA RTX Spark will be used to showcase local large language model experiences.
Azure Local is expanding to NVIDIA Jetson Thor and Jetson Orin edge AI platforms to support industrial applications such as machine vision, robotics and autonomous operations.
Source-stated
NVIDIA says Microsoft AI uses NVIDIA Dynamo during post-training, coordinating inference across a large GPU fleet.
Microsoft AI uses NVIDIA Dynamo to support reinforcement learning rollout inference during post-training.
Source and contributor context
Dave Salvator · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
Media declared by the original publisher; availability checked separately from the editorial reading.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
NVIDIA
NVIDIA develops processors, networking and software used for AI, graphics and scientific computing.
Editorial provenance
Version 1. Publication ID: 2d7aac0a-7fac-4f55-8b5f-65d96c1f2f6c.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-19T20:17:37.486Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: e861494232c31d73ebc5ef042a67eb1c0790bf3c2ff4ff9064e60c10eaa23f6e. Source text hash: 522fb3c4f635522f36f60a9cafca6349349cd886adeea583d222437640e0850d.
uncertain · 2026-11-19T20:17:37.486Z · anthropic/claude-opus-5.5
Publication: 2d7aac0a-7fac-4f55-8b5f-65d96c1f2f6c
attribution — pass: The reading consistently presents the claims as NVIDIA's account of its collaboration with Microsoft. It attributes the EY result to NVIDIA's reporting and notes that the figures come from the publisher. This matches the NVIDIA blog source.
NVIDIA and Microsoft are extending their full-stack collaboration to help organizations build, customize and deploy intelligent applications
summary — uncertain: The summary is accurate but reads as a list of topics. It does not explain central terms for a general reader: Vera Rubin NVL72 (a rack-scale GPU system), confidential computing (protecting data while it is processed), post-training and edge. It also does not say why the development might matter, such as running models closer to company data. In the summary, the EY cost figure omits its baseline, proprietary frontier models, and the use of Nemotron open models. Smallest repair: add short plain-language glosses or tooltips, state the cost comparison baseline, and add one sentence on the evidence-supported significance.
Confidential Computing and attestation help customers verify the computing environment before releasing sensitive models, data or keys
date — pass: The originalPublished value of 2026-11-19T20:00:15Z comes from the publisher page and matches the feed entry. Nothing indicates an earlier first publication.
support — pass: Each claim matches the source text: the scope of the collaboration, the EY result with its baseline, Azure Local support for Jetson and RTX PRO 6000, and Microsoft AI's use of Dynamo. The availability caveats are supported by the source's references to a preview, a limited preview and 'coming soon'. The question follows from the evaluation-dataset limitation and does not presuppose an answer.
Nemotron 3.5 Bolt — designed for high-throughput, low-latency reasoning — and Transcribe, for speech recognition, are coming soon.
media — pass: The selected media is the verified link to the original NVIDIA blog article. It is the most direct way for readers to reach the source.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
NVIDIA develops processors, networking and software used for AI, graphics and scientific computing.
Explore NVIDIANVIDIA develops processors, networking and software used for AI, graphics and scientific computing.
Explore NVIDIANVIDIA develops processors, networking and software used for AI, graphics and scientific computing.
Explore NVIDIANVIDIA develops processors, networking and software used for AI, graphics and scientific computing.
Explore NVIDIANVIDIA develops processors, networking and software used for AI, graphics and scientific computing.
Explore NVIDIANVIDIA develops processors, networking and software used for AI, graphics and scientific computing.
Explore NVIDIA