AI news for builders and product teamsUpdated Nov 17, 2026, 13:01 UTC
NVIDIA Blog
First-party releases and research from NVIDIA Blog. Headlines and excerpts link to the original articles.
Latest stories
Newest first
S-GAME's Phantom Blade Zero launched on GeForce NOW on Oct. 28, leading a week of nine new library titles, mod.io support for Cyberpunk 2077 and The Witcher 3: Wild Hunt — Remastered, and new four-month memberships in India.

C-BRAIN, a consortium of 18 institutions launched in November 2025 and led by Dr. Randall Bateman, is working with NVIDIA to apply federated AI to Alzheimer's research data, including unpublished and negative results it calls "dark data." Its open source OpenScientist tool completed analyses in minutes that scientists estimated would take weeks to months, at under $10 per run.

At GTC Berlin, NVIDIA showcased European banks and financial firms building in-house AI on open weight models and its accelerated computing, including Revolut's PRAGMA foundation model serving 80 million customers and Stripe's tested 42x faster fraud-detection data pipelines with 92% cost savings.

NVIDIA says energy and data-center companies including GE Vernova, Hitachi Energy, Schneider Electric, Siemens Energy, Phaidra and Emerald AI are building on its DSX AI Factory Platform to plan, simulate and operate grid and compute power as one system.

Shopify post-trained NVIDIA's Nemotron 3.5 Lightning for a GraphQL subagent in its Sidekick merchant assistant, reporting higher accuracy than a leading frontier model on an internal benchmark with 90% lower serving cost and 7x faster responses.

NVIDIA outlined a set of European public-sector AI projects built on its Nemotron open models, including a German sovereign cloud for classified data, a 1-million-record synthetic persona dataset released in the UK, and a French workspace open to more than 2,500 organizations.

At GTC Berlin, NVIDIA detailed its local AI hardware lineup: RTX Spark laptops, DGX Spark 64GB systems shipping October 23, and DGX Station, with a Windows version due in December. It also announced Recursive Flint, a 40-billion-parameter open coding model for its local hardware.

NVIDIA's European cloud partners are expanding regional capacity built on Vera Rubin and DSX, and the company has announced an OpenShell Reference Architecture, unveiled at GTC Berlin, for confining what AI agents can access. Nscale reported 37% more GPUs and nearly 50% higher inference throughput within the same power budget using DSX MaxLPS.

NVIDIA released its Factory Operations Blueprint (FOX) in early access on GitHub, letting developers build always-on factory manager and specialist agents that connect to models such as Cosmos and Nemotron. Siemens is integrating FOX-built agents into Intelligence Center X, with an estimated 20% cost saving.

NVIDIA has released cuDSS, a CUDA-X library that moves sparse linear system solving onto GPUs, into general availability. Early adopters report speedups ranging from 4x in an EUV lithography benchmark to as much as 10x versus CPU-only sparse solves.

NVIDIA announced Confidential Bare Metal, a new deployment option in its Confidential Computing line that lets frontier model builders run sensitive workloads on dedicated third-party bare-metal infrastructure while keeping control of models and data. Cohere plans to adopt it in Cohere Model Vault, and Google Cloud will extend it to A4X Max instances powered by GB300 NVL72.

NVIDIA has expanded its open source Agent Toolkit with a new Agent Knowledge Foundation, the Kumo open model family and an agent-optimization NemoClaw blueprint, aimed at helping enterprises ground agents in proprietary data without moving it.

NVIDIA says AI cloud providers Nebius, Nscale and Verda are building infrastructure on its Vera CPU, with server makers Dell, HPE, Lenovo, Supermicro and Bull also building Vera-based systems. Vera is positioned for agentic workloads, especially the fast sandbox startups and data processing that agents rely on.

NVIDIA says some Barcelona Supercomputing Center AI Factory services are now reachable through the NVIDIA Inception portal, and three Inception members were named AI Champions at the programme's pilot demo day.

NVIDIA's RTX Spark laptops are now available starting at $2,399 from ASUS, Dell, HP, Lenovo, Microsoft and MSI, pairing a Blackwell RTX GPU with a Grace CPU, up to 128GB of unified memory and up to 1 petaflop of AI compute.

NVIDIA is launching RTX Spark, a Windows PC platform built for creators, gamers and AI agents, with hands-on launch events in six cities on Friday, Oct. 16. Reviewers cited by NVIDIA praise its local AI, gaming and creative performance, and the announcement cites independent testing from Artificial Analysis.

NVIDIA released Nemotron 3.5 Super VL, an open 120-billion-parameter multimodal agent model it claims matches its largest reasoning model at roughly one-fifth the parameter count, alongside two edge-focused Bolt models, one of which arrives October 23.

NVIDIA added DLSS 4.5 controls to GeForce NOW, giving members in-app settings for Super Resolution, Dynamic Multi Frame Generation and Ray Reconstruction, with Ultimate tier streams on RTX 5080 rigs reaching Dynamic Multi Frame Generation at up to 4X.

At the AI in Science Summit in Dublin, Nvidia backed the European Commission's new AI in Science Enablers Initiative, tied to RAISE, offering open models, training, compute access and startup support. Nvidia says 35 supercomputers are in development across Europe, to serve more than 3 million researchers.

NVIDIA detailed its expanding Nordic footprint, including a Swedish consortium's Sferical One supercomputer with 1,152 GPUs, an expected 1 GW of regional AI cloud capacity by the end of 2027, and deployments by Spotify, Volvo Cars, Einride and several Swedish startups.