Volume 343, Issue 3 · 11 articles
NVIDIA launches Minerva, an AIOps platform that applies AI models to infrastructure and workload telemetry to keep AI factories productive.
NVIDIA has launched Minerva, an AIOps platform designed to help teams find problems earlier and keep AI factories productive.
Minerva brings together telemetry from infrastructure and AI workloads and applies purpose-built AI models to turn that data into actionable insight.
NVIDIA notes that AI factories run vast numbers of GPUs across high-speed networks, so a single degraded component, such as a hot GPU or a slowing NVMe drive, can slow an entire job.
Minerva is meant to surface those easily missed warning signs.
NVIDIA expands Cosmos with new training tools, simulation checkpoints and Cosmos Data Factory blueprints for world model builders.
NVIDIA has expanded its Cosmos platform with new training tools, simulation checkpoints and Cosmos Data Factory blueprints to help developers build world models for physical AI.
NVIDIA says Cosmos ranks in the top five on six open-model leaderboards and gives builders frontier open models plus data, training and optimization recipes.
The tools are meant to help teams scale training data, adapt models, test behavior and optimize performance before deployment.
Cosmos Data Factory brings together multimodal, multisensory data as the starting point for world models.
NVIDIA adds Confidential Bare Metal so frontier model builders can run sensitive workloads on third-party infrastructure while keeping control.
NVIDIA is expanding NVIDIA Confidential Computing with Confidential Bare Metal, a new deployment option for frontier model builders.
It lets them run sensitive workloads on dedicated third-party infrastructure while keeping control of their models and data.
NVIDIA positions confidential computing as the infrastructure layer protecting the models and data agents depend on.
The expansion comes alongside the newly announced NVIDIA OpenShell Reference Architecture for deploying agents safely at scale.
NVIDIA expands its Isaac robotics platform with a reference humanoid, datasets, runtime and robot learning framework updates.
NVIDIA announced a major expansion of its Isaac open robotics development platform at GTC Berlin.
The update adds a reference humanoid, datasets and a runtime, along with updates to NVIDIA's robot learning framework.
NVIDIA says the pieces give developers an integrated path from data to deployment for general-purpose robots.
Jensen Huang said the ChatGPT moment for robotics is fast approaching.
NVIDIA says companies use its DSX AI Factory Platform to plan, simulate and run the grid and AI factory as one system.
NVIDIA says companies worldwide are building on the NVIDIA DSX AI Factory Platform to plan, simulate and operate the power grid and the AI factory as a single system, from substation to rack.
NVIDIA calls power the biggest bottleneck in infrastructure, citing more than $300 billion in backlogged equipment and over 200 gigawatts of projects in U.S. interconnection queues.
It blames siloed design of grid, facility and compute for delays, risk and stranded capacity, which DSX aims to address.
Google Cloud A5X instances built on NVIDIA Vera Rubin NVL72 reach general availability next month, with plans for up to 960,000 Rubin GPUs.
NVIDIA says Google Cloud A5X instances built on NVIDIA Vera Rubin will be generally available next month, with planned expansion to Europe.
The instances combine Vera Rubin NVL72, ConnectX-9 SuperNICs, the Google Cloud Virgo Network and fourth-generation liquid cooling for agentic inference and reinforcement learning at scale.
Google Cloud plans to scale its Vera Rubin infrastructure to as many as 960,000 Rubin GPUs across multiple sites.
London-based Ineffable Intelligence is cited as an early user.
NVIDIA says European banks and payment firms at GTC Berlin are running open-weight models on their own NVIDIA infrastructure.
At GTC Berlin, NVIDIA highlighted banks, payment companies and financial data providers building production-scale intelligence on NVIDIA accelerated computing and open-weight models running on their own infrastructure.
Revolut is rebuilding its technology stack on a single foundation model, PRAGMA, using NVIDIA cuDF and Dynamo-Triton on Blackwell Ultra GPUs.
NVIDIA says Revolut is scaling PRAGMA in production to serve an initial 80 million customers, with headroom to scale 3x.
NVIDIA makes cuDSS generally available, a CUDA-X library that moves large sparse system solving for simulations onto GPUs.
NVIDIA has launched cuDSS in general availability, a CUDA-X library that shifts a critical part of engineering and scientific simulation onto GPUs.
NVIDIA notes that complex designs for chips, aircraft and industrial systems depend on simulations that grow slower as designs get more complex.
cuDSS accelerates solving large sparse linear systems, the math-heavy step that repeats throughout these simulations.
NVIDIA says this lets teams test more design ideas in the same time.
NVIDIA says telecom operators are adding accelerated computing across their networks, and T-Mobile completed a live AI-RAN call with Nokia.
NVIDIA announced that leading telecom operators are adding NVIDIA accelerated computing to central data centers, metro sites and cell sites.
The goal is to serve AI models and applications closer to the people and machines that use them.
T-Mobile completed what NVIDIA calls the world's first AI-RAN call on a live network in Bellevue, Washington, with NVIDIA and Nokia.
Deutsche Telekom is advancing physical AI, Orange is adding voice intelligence services, and Elisa reported a European first.
NVIDIA says Amkor, IBIDEN, KYEC, Samsung Electro-Mechanics, SPIL and Unimicron will expand manufacturing capacity for its AI platform.
NVIDIA announced at GTC Berlin plans with supply chain partners to expand global manufacturing capacity for its accelerated computing and AI factory platform.
The partners are Amkor Technology, IBIDEN, King Yuan Electronics, Samsung Electro-Mechanics, Siliconware Precision Industries and Unimicron.
NVIDIA says it is helping new and existing suppliers expand capabilities in substrates, advanced packaging, testing and optical engines.
The goal is to make NVIDIA's global supply chain more resilient and redundant.
NVIDIA expands its open-source Agent Toolkit, adding Agent Knowledge Foundation to ground enterprise agents in company data.
NVIDIA says enterprises are using its open-source NVIDIA Agent Toolkit, now expanded with new tools, to customize, control, monitor and govern AI agents.
The company argues that agents are only as useful as what they know and only as trustworthy as the boundaries they run within.
Builders need an open agent stack that combines models, harness, runtime, data and tools.
The new NVIDIA Agent Knowledge Foundation helps enterprises turn their data into company intelligence agents can use.