For operators of large-scale AI factories, maximizing continuous output is essential for productivity. In massive-scale AI training, every GPU in the cluster must synchronize gradients across…

How can a 30B-parameter model activate only 3B parameters per token, and still use the capacity of the larger model? Nemotron 3.5 Lightning illustrates the answer: It uses a Mixture-of-Experts (MoE)…

At Dreamforce, Salesforce and NVIDIA announced Koa, a CRM reasoning model for Agentforce, post-trained from Nemotron on 27 years of Salesforce CRM intelligence — and running entirely within Salesforce’s own infrastructure.

Manchester physicists David Topping and Hao Zhang applied NVIDIA Earth-2 generative models to air quality forecasting — unlocking predictions fast enough to inform public health.

The AI Energy Management Alliance brings together the full AI and power value chain to accelerate the interconnection of flexible, grid-enhancing data centers, strengthen reliability and protect affordability.

cuTile Rust () is a tile-based system for safe, idiomatic GPU kernel authoring in the Rust programming language. Extending the Rust ownership model to tile-based GPU kernels, it splits mutable outputs…

First preview submission using NVIDIA Vera Rubin NVL72 systems delivers up to 3.7x better throughput than NVIDIA GB300 NVL72.

AI agents are moving from cloud data centers to vehicles, robots, and other edge devices. Unlike a chatbot that answers a single prompt, an agent works through a sequence of steps. It selects tools…

Agentic AI workflows can be used to prepare and validate digital twins for physical AI systems. Agents can inspect 3D scenes, author simulation-relevant data in OpenUSD, add physics properties…

Explore the creature-catching adventure at launch, alongside ‘Active Matter,’ ‘Outer Wilds’ and '007 First Light' with a path-tracing update.

You’re deploying a model on a system. It starts up, prompts are getting responses. Now the hard question: Is this fast? Your instincts might lead you to send curl commands, hand-roll an asyncio script…

Weather-sensitive industries increasingly have access to observations that offer an earlier, more local view of changing conditions. Energy companies collect measurements across wind and solar assets…

From fast-tracking fusion to repurposing recycled EV batteries for energy storage, these clean energy innovators are harnessing AI to help build a lower carbon grid.

Open research, controls across the agent stack and continuous testing help defenders build and operate more secure AI systems.

AI factories are also coming online across South Africa, Morocco and Nigeria, enabling Africa’s developers and enterprises to reach production-scale compute onshore.

Physical AI safety spans hardware, software and validation. See how NVIDIA Halos brings full-stack safety to autonomous vehicles and robots.

The new qualification program helps AI factory builders identify products that meet applicable NVIDIA DSX requirements, starting with battery energy storage systems and cooling distribution units.

When you ship an AI agent, the key question is whether it can execute a chain of work across dozens of sequential tool calls against a live environment, and recover when a step fails.

The compute and memory demands of generative AI increasingly exceed what a single GPU can provide. NVIDIA TensorRT multi-device inference is a new capability that enables a single TensorRT network to…

GPU acceleration can speed up compute-intensive robotics workloads, but a fast CUDA kernel alone does not guarantee a fast ROS 2 graph. As messages move between nodes, they may continue to be…

NVIDIA DLSS 5 introduces DLSS 3D-Guided Neural Rendering and granular controls that help game developers add lifelike lighting and material detail while preserving their artistic intent.