Nvidia: latest AI news
All Oossa news about Nvidia, newest first.
Known for: GPU, Vera Rubin, CUDA
Follow via RSSToday
Business
NVIDIA launches Minerva for AI factory monitoring
Minerva combines infrastructure and workload data to help operators spot faults, trace incidents and manage AI factories from one view.
Business
QumulusAI Expands GPU Lease Financing to $123.6 Million
QumulusAI says its expanded three-year leases with TFC will finance NVIDIA Blackwell GPUs without issuing new equity.
Business
Supermicro, Nokia and NVIDIA launch AI‑RAN edge platform
Supermicro teams with Nokia to ship NVIDIA Aerial RAN computers for low‑latency AI at 5G network edge.
Business
Genedata adds NVIDIA RFD4-Proteina model to its AI discovery platform
Genedata now lets biopharma researchers use NVIDIA's new RFD4-Proteina binder design model within its AI‑enabled workflows.
Business
Supermicro launches 256‑CPU liquid‑cooled rack for agentic AI
The NVIDIA Vera CPU rack, with 256 Vera CPUs and 100% liquid cooling, is now available for quote, targeting large AI factories.
Business
Perplexity and NVIDIA unveil JOLT learning method
The new post‑training technique lets a model teach itself, cutting token use by 6.5× on the MATH benchmark.
Business
NVIDIA adds StormScope EU model to Earth‑2 open weather platform
The new StormScope EU model predicts cloud conditions across Europe every 10 minutes, letting developers build local forecasts for energy, transport and insurance.
Models
NVIDIA releases TensorRT‑LLM 1.4.0rc1 with KV cache manager V2
The new release adds support for several models, makes the KV cache manager V2 the default, and removes unused request fields.
Yesterday
Business
Vivix uses a cloud‑based digital twin to cut glass furnace energy use
Vivix, Siemens, Snowflake and NVIDIA built an executable digital twin that could lower furnace energy by up to 5% and add a year to its life.
Business
Fleet adds management for NVIDIA RTX PRO workstations and DGX Station
Fleet Device Management now supports NVIDIA's full AI hardware line, letting IT teams govern RTX PRO PCs and DGX Station supercomputers alongside regular devices.
Business
Volkswagen Group launches AI data flywheel with Voxel51, NVIDIA, Microsoft
Porsche leads Volkswagen's new stack that turns fleet data into simulation, promising up to 90% fewer test drives and 85% faster AI test cycles.
Business
Tanium adds enterprise management to NVIDIA’s on‑prem AI hardware
Tanium says its endpoint‑management software will now support NVIDIA’s RTX Spark PCs and DGX Station AI machines.
Business
Red Hat and Nvidia expand partnership to secure autonomous AI agents
Red Hat added new runtime and hardware controls to Nvidia’s platform, aiming to stop AI agents from overstepping their limits or being hijacked.
Models
Mistral and Nvidia team up on physics AI for design
Mistral Large 4 and Nvidia's tools will link design, simulation and AI training for aerospace, automotive and semiconductor firms.
Models
DDN and Nebul use Infinia storage to speed up sovereign AI inference
At GTC Berlin 2026, DDN said its Infinia system will store KV cache for Nebul's NVIDIA‑based AI cloud, cutting GPU work and cost.
Business
Verda and Compan expand AI cloud hardware partnership
Verda and Compal will use Compal's new SGX30-2 server with eight NVIDIA Blackwell Ultra GPUs to speed AI cloud rollout worldwide.
Business
Samsung Electro-Mechanics to Supply NVIDIA AI Chip Substrates from 2026
The two firms will jointly expand production lines in South Korea to secure a stable supply of advanced package substrates for NVIDIA's AI processors.
Business
Cadence launches Europe’s first sovereign AI platform for chip design
Cadence, with NVIDIA, opens a Munich data center that bundles its cloud tools, a digital twin platform, and a Blackwell‑based supercomputer.
Business
Armada to ship factory‑built AI data centers in months
Armada and Johnson Controls will use NVIDIA’s DSU POD design to deliver modular AI infrastructure, aiming for 500 MW capacity in 2027.
Business
ServiceNow and Nvidia make Arc AI agent generally available
Arc, the joint AI assistant, is now generally available and runs on Nvidia RTX Spark hardware with multi‑level security controls.
Business
Nvidia ships AI racks in a truck‑size container
Nvidia’s new DSX Pod packs seven Vera Rubin AI racks into a standard shipping container, cutting set‑up time for data centers to two weeks.
Business
NVIDIA launches 64 GB DGX Spark for local AI development
DGX Spark systems with 64 GB memory ship on 2026‑10‑23, letting developers run large language models on‑premises.
Business
Vertiv and NVIDIA unveil AI‑factory method promising up to 30% more token output
The new pre‑engineered, factory‑built approach could cut design time from 30 days to 3 and deployment time from 90 to 45 days.
Business
NVIDIA launches cuDSS library for faster engineering simulations
The new CUDA‑X library speeds up sparse‑matrix solves on GPUs, with reported 4‑10× gains in several industry tools.
Business
European cloud providers adopt NVIDIA Vera CPUs
Nebius, Nscale and Verda will use NVIDIA's new Vera CPUs to power AI agents and data‑center workloads across Europe.
Business
Google Cloud adds NVIDIA Vera Rubin A5X instances for AI training and inference
Google Cloud will make NVIDIA Vera Rubin‑based A5X instances generally available next month, with plans to expand to Europe and scale up to 960,000 Rubin GPUs.
Business
Greenpeace disrupts Nvidia GTC keynote in Berlin
Protesters interrupted Jensen Huang’s speech on Oct 21, 2026, demanding answers about Nvidia chips found in Russian drones.
Business
NVIDIA adds new open models for language, data, robotics and quantum computing
NVIDIA announced a suite of open‑weight models—including Nemotron 3.5, Kumo, Cosmos and BioNeMo—available on Hugging Face and NVIDIA’s own service.
Tools & agents
Google launches SynthID Detector for watermark detection across multiple AI providers
The tool lets anyone upload media to check if it carries a watermark from Google, OpenAI, NVIDIA, Kakao and soon Apple.
Business
F5 and Telenor choose F5 BIG-IP Next for AI Service Plane
F5's BIG-IP Next for Kubernetes will run on Telenor AI Factory's NVIDIA‑powered sovereign AI infrastructure.
