#nvidia

Google releases DiffusionGemma, a 26-billion-parameter model that generates text via diffusion, producing 256 tokens in parallel for up to 4x speedup on single GPUs.

The IEEE Honors Ceremony in New York City recognized over 20 laureates, including Nvidia CEO Jensen Huang, for groundbreaking contributions to technology and connectivity.

NVIDIA introduced RTX Spark, an Arm-based Windows PC chip it says can deliver up to 1 petaflop of AI performance for laptops and compact desktops.

Reports suggest Nvidia-based Windows PCs from Dell and Microsoft are coming, alongside software that lets AI agents handle tasks locally on the device.

Nvidia says its annual spending with suppliers in Taiwan has climbed to as much as $150 billion, underscoring how sharply the AI hardware boom has scaled.

FANUC says deeper integration between ROBOGUIDE and NVIDIA Isaac Sim will let manufacturers test, teach, and refine robots inside a more practical digital twin environment.

Nvidia says it will split data center revenue into hyperscalers and a broader enterprise category, a reporting shift aimed at showing investors its business is not tied only to the largest AI cloud buyers.

Nvidia CEO Jensen Huang joined President Donald Trump’s business delegation to Beijing after early reports said he would miss the trip, placing the chipmaker at the center of high-level U.S.-China commercial diplomacy.

A Wired analysis argues that Nvidia’s enduring advantage in AI comes less from chips alone than from CUDA, the software platform that made those chips indispensable to parallel computing.

Nvidia has already committed more than $40 billion to AI-related equity deals in 2026, reinforcing its influence far beyond chips and data centers.

Nvidia has released an open multimodal model for text, image, video and audio processing, along with details showing how heavily synthetic data from rival models now shapes frontier AI training

At Google Cloud Next, Google and Nvidia outlined infrastructure plans aimed at lowering the cost of AI inference at scale, highlighting how the economics of serving models are becoming a primary battleground.

Recursive Superintelligence, founded just four months ago, has reportedly raised at least $500 million to pursue self-improving AI systems.

Nvidia researchers say Lyra 2.0 can generate coherent 3D environments extending roughly 90 meters from a single image, with exports to simulation tools such as Isaac Sim for robotics training.

Cadence Design Systems used its CadenceLIVE event to expand AI-related work with Nvidia and introduce new integrations with Google Cloud, underscoring how engineering software is being rebuilt around AI infrastructure.

Span says it is working with Nvidia and partners on XFRA, a network that would place high-powered compute nodes at homes and small businesses, using smart panels to turn spare electrical capacity into distributed AI infr

Singapore-based Firmus has raised another $505 million at a $5.5 billion post-money valuation, underscoring investor appetite for AI infrastructure and for data center projects built around Nvidia’s next-generation plans

China has approved Nvidia to sell its H200 AI chip to Chinese customers, ending a prolonged regulatory freeze, as the company simultaneously develops a downgraded inference chip for the Chinese market.

At GTC 2026, Nvidia announced major expansions to its physical AI platform, with autonomous vehicles launching in Los Angeles through an Uber partnership, FANUC and ABB industrial robots getting Nvidia brains, and new foundation models targeting humanoid robotics.

Nvidia will supply its full autonomous driving software stack to power Uber robotaxis starting in the first half of 2027, targeting 28 global markets across four continents by the end of 2028.