NVIDIA Dev Blog

As AI infrastructure scales, enterprise expectations for operational maturity are increasing. Organizations expect these systems to be provisionable,... As AI infrastructure scales, enterprise expectations for operational maturity are increasing. Organizations expect these systems to be provisionable, observable, secure, and manageable at scale—the same standard applied to all critical infrastructure. The moment an AI system moves from development into enterprise deployment, that operational foundation is essential. NVIDIA DGX Spark and… Source

Converting a quantized checkpoint into an NVIDIA TensorRT engine bridges the gap between model optimization and production deployment, enabling faster... Converting a quantized checkpoint into an NVIDIA TensorRT engine bridges the gap between model optimization and production deployment, enabling faster inference, higher throughput, and more efficient GPU utilization at scale. In a previous post, we produced a high-quality FP8-quantized Contrastive Language-Image Pretraining (CLIP) checkpoint with NVIDIA TensorRT Model Optimizer. Source

Federated learning (FL) research often begins with a deceptively simple question: What should we try next? A new aggregation rule, a FedProx coefficient, a... Federated learning (FL) research often begins with a deceptively simple question: What should we try next? A new aggregation rule, a FedProx coefficient, a server optimizer setting, a SCAFFOLD variant, or a model architecture tweak may all look promising before an experiment starts. After the run finishes, the harder questions begin: Did the change actually improve the metric? Source

Training a speech AI model to correctly recognize or synthesize clinical terminology is surprisingly difficult. Drug names like Acetaminophen, Amlodipine,... Training a speech AI model to correctly recognize or synthesize clinical terminology is surprisingly difficult. Drug names like Acetaminophen, Amlodipine, Cefazolin, and Biktarvy are not part of everyday vocabulary. Procedure names, anatomy terms, and specialty-specific diagnoses introduce the same problem in a different form. Off-the-shelf speech systems can sound fluent and still miss the words… Source

Pre-training frontier LLMs comes down to throughput. When training spans trillions of tokens across thousands of accelerators, every percentage point of step... Pre-training frontier LLMs comes down to throughput. When training spans trillions of tokens across thousands of accelerators, every percentage point of step time can add up to days of training and substantial compute costs. Numerical precision is one of the highest-leverage knobs available, but low- bit mixed-precision pretraining is hard to get right. To address this… Source

Single-turn chatbots are evolving into long-running agents that can reason, maintain context, use tools, and run efficiently across many turns to complete... Single-turn chatbots are evolving into long-running agents that can reason, maintain context, use tools, and run efficiently across many turns to complete complex workflows. However, these multi-agent workflows cause token counts to grow quickly. Agents plan, call tools, invoke sub-agents, receive information, and then pass history, outputs, and reasoning steps back into the model… Source

AI agents are changing how you interact with your PC. Creators, developers, and AI enthusiasts are already using these agents extensively to assist with... AI agents are changing how you interact with your PC. Creators, developers, and AI enthusiasts are already using these agents extensively to assist with day-to-day tasks such as coding, video editing, and content management. NVIDIA and Microsoft are teaming up to enable the next generation of developers to build on-device agents on the Windows platform, with easier setup, native security… Source

AI agents are a powerful tool for synthesizing data to accelerate research, summarize information, and help teams make decisions faster. But combining internal... AI agents are a powerful tool for synthesizing data to accelerate research, summarize information, and help teams make decisions faster. But combining internal data with public sources poses security challenges. This post shares an open source example using Hermes Agent with NVIDIA NemoClaw for product research across Outlook, Slack, and GitHub. NVIDIA OpenShell enforces a security-approved… Source

As AI agents move from the digital world to the physical environment, they can readily use NVIDIA Jetson to accelerate real-world deployment with optimized... As AI agents move from the digital world to the physical environment, they can readily use NVIDIA Jetson to accelerate real-world deployment with optimized memory and performance. NVIDIA JetPack 7.2 directly supports one-command deployment of NVIDIA NemoClaw, an open source stack that adds privacy and security controls to OpenClaw. It introduces NVIDIA agent skills for Jetson—Jetson device… Source

The rise of autonomous, long-running AI agents has introduced a new class of compute demand, namely tasks that maintain large context windows, spawn concurrent... The rise of autonomous, long-running AI agents has introduced a new class of compute demand, namely tasks that maintain large context windows, spawn concurrent subagents, and iterate continuously without cloud dependency. Security and privacy concerns are also accelerating the shift toward local agents. Developers, by running autonomous agents on hardware they own with NVIDIA NemoClaw… Source

Developing autonomous vehicle (AV) policies requires bridging an important gap between training and deployment. Vision-language-action (VLA) models that can... Developing autonomous vehicle (AV) policies requires bridging an important gap between training and deployment. Vision-language-action (VLA) models that can reason over more complex driving scenes and produce richer intermediate reasoning are predominantly trained in open-loop, where model outputs are directly compared to ground-truth behaviors without considering their effect on the environment. Source

Physical AI systems must understand the real world before they can act within it. Robots, autonomous vehicles, and smart spaces need to understand what's... Physical AI systems must understand the real world before they can act within it. Robots, autonomous vehicles, and smart spaces need to understand what’s happening in their world, predict what’s likely to happen next, and generate actions for specific environments, embodiments, and tasks. NVIDIA Cosmos 3 is a frontier foundation model for physical AI that combines physical reasoning… Source

The AI era is driving a new class of infrastructure: AI factories that transform data into intelligence for autonomous AI agents operating at unprecedented... The AI era is driving a new class of infrastructure: AI factories that transform data into intelligence for autonomous AI agents operating at unprecedented scale. Powered by accelerated computing, AI factories enable enterprises to train, fine-tune, and deploy AI with greater speed and efficiency. This new class of infrastructure also introduces a fundamentally new attack surface spanning… Source

Each wave of AI has created a new scaling law. Pretraining scaled intelligence through larger datasets, more parameters, and massively parallel GPU systems.... Each wave of AI has created a new scaling law. Pretraining scaled intelligence through larger datasets, more parameters, and massively parallel GPU systems. Post-training scaled usefulness through instruction tuning, and re-balancing GPUs for generative inference. Test-time scaling improved reasoning by giving models more generated tokens for thinking. Now, agentic AI and reinforcement… Source

AI is now essential infrastructure, powered by AI factories that generate intelligence in the form of tokens. As demand grows, these factories must scale... AI is now essential infrastructure, powered by AI factories that generate intelligence in the form of tokens. As demand grows, these factories must scale faster, operate more efficiently, and lower the cost of intelligence across the five-layer stack: energy, chips, infrastructure, models, and applications. NVIDIA DSX platform provides the complete playbook for designing, simulating, building… Source

Modern LLM serving is hard to tune because each deployment is a stack of interacting choices: model backend, tensor-parallel shape, prefill/decode split, worker... Modern LLM serving is hard to tune because each deployment is a stack of interacting choices: model backend, tensor-parallel shape, prefill/decode split, worker counts, scheduler settings, routing policy, KV cache behavior, autoscaling thresholds, and topology. Those choices interact across layers, and a local improvement can shift the bottleneck somewhere else. For larger models… Source

As AI models grow in complexity and regulatory scrutiny intensifies under frameworks including California’s AB-2013 and the EU AI Act, software teams... As AI models grow in complexity and regulatory scrutiny intensifies under frameworks including California’s AB-2013 and the EU AI Act, software teams face a challenge beyond delivering great code: They need to produce comprehensive, auditable model documentation before the models are released. Model cards describe how a model works, its intended use and license, training data, performance… Source

AI applications are moving beyond text generation to multimodal systems that can perceive, search, and reason across images, documents, video, and... AI applications are moving beyond text generation to multimodal systems that can perceive, search, and reason across images, documents, video, and language in real time—turning fragmented information into actionable insights. Step 3.7 Flash, the latest from StepFun, brings these capabilities to production and enterprise-scale, available on NVIDIA-accelerated infrastructure. It is a 198B… Source

The cold-start problem In production inference deployments, demand fluctuates over time, requiring inference replicas to scale elastically. However,... In production inference deployments, demand fluctuates over time, requiring inference replicas to scale elastically. However, cold-starting inference workloads on Kubernetes can take several minutes. During that time, GPUs are allocated but idle, generating no tokens and serving no requests. This delay increases the risk of service level agreement (SLA) violations during traffic spikes… Source

Large language models (LLMs) are revolutionizing the financial trading landscape by enabling sophisticated analysis of vast amounts of unstructured data to... Large language models (LLMs) are revolutionizing the financial trading landscape by enabling sophisticated analysis of vast amounts of unstructured data to generate actionable trading insights. These advanced AI systems can process financial news, social media sentiment, earnings reports, and market data to predict stock price movements and automate investment strategies with unprecedented… Source

NVIDIA RTX provides game developers with direct paths to AI-driven characters, frame generation, and ray-traced rendering. This post walks through a meaningful... NVIDIA RTX provides game developers with direct paths to AI-driven characters, frame generation, and ray-traced rendering. This post walks through a meaningful set of recent updates across the RTX ecosystem. NVIDIA ACE expands its multilingual AI character capabilities, making it easier to ship conversational NPCs. NVIDIA DLSS 4.5 arrives as an Unreal Engine (UE) plugin… Source

Developers can now use NVIDIA CUDA Tile programming within large existing C++ GPU codebases to develop highly optimized GPU kernels using tile-based... Source

NVIDIA CUDA 13.3 brings new capabilities and performance optimizations to developers across the CUDA ecosystem. The launch of NVIDIA CUDA Tile programming in... NVIDIA CUDA 13.3 brings new capabilities and performance optimizations to developers across the CUDA ecosystem. The launch of NVIDIA CUDA Tile programming in C++, enables high-level, tile-based kernel development that automatically manages complex low-level GPU details for optimal performance and portability. Additionally, CUDA Tile programming is now supported on Compute Capability 9.0… Source

NVIDIA CompileIQ tackles one of the hardest problems in performance engineering: finding the compiler options that unlock the best performance for a specific... NVIDIA CompileIQ tackles one of the hardest problems in performance engineering: finding the compiler options that unlock the best performance for a specific workload. Consider a team that has spent weeks optimizing an LLM inference pipeline on GPUs, tuning batch sizes, quantizing to FP8, adopting flash attention, fusing every kernel they can. The profiler says there’s nothing left to squeeze. Source

Precision medicine depends on two fundamental capabilities: understanding disease at the genomic level and identifying treatments at the molecular level. ... Precision medicine depends on two fundamental capabilities: understanding disease at the genomic level and identifying treatments at the molecular level. NVIDIA’s contributions to precision medicine extend far beyond accelerated computing, delivering a full-stack platform that translates hardware and software advancements directly into healthcare outcomes. Sequencing the human genome… Source

High‑quality 3D medical imaging data is the foundation of modern radiology AI, but access to it is often constrained by data scarcity, privacy restrictions,... High‑quality 3D medical imaging data is the foundation of modern radiology AI, but access to it is often constrained by data scarcity, privacy restrictions, and the high cost of expert annotation. As a result, training reliable 3D medical imaging models is frequently bottlenecked by small, narrow, and hard‑to‑share datasets, limiting model robustness and generalization. To help teams overcome… Source

In quantitative finance, researchers build algorithms to trade assets, derivatives, and other financial instruments. A key part of that work is finding signals:... In quantitative finance, researchers build algorithms to trade assets, derivatives, and other financial instruments. A key part of that work is finding signals: patterns in messy market data that may help predict future returns. These signals can come from price and volume data, economic indicators, fundamentals, or alternative sources like news sentiment. For years… Source

As AI models grow in scale and complexity, realizing the full performance of modern accelerated infrastructure depends as much on how workloads are placed as on... As AI models grow in scale and complexity, realizing the full performance of modern accelerated infrastructure depends as much on how workloads are placed as on the hardware itself. NVIDIA GB200 NVL72 delivers exascale compute in a single rack, unlocking real-time trillion-parameter models. Yet capturing that performance in a shared cluster requires schedulers that understand the system… Source

Telcos around the world are building sovereign AI factories based on the NVIDIA Cloud Partner (NCP) reference architecture, giving governments, enterprises, and... Telcos around the world are building sovereign AI factories based on the NVIDIA Cloud Partner (NCP) reference architecture, giving governments, enterprises, and startups access to in‑country AI infrastructure with the right controls, trust, and performance. But infrastructure alone doesn’t get you to high-margin, production-ready enterprise AI services. Model sizes and reasoning workloads… Source

Autonomous AI agents are taking on all types of work for businesses: routing logistics fleets, triaging support tickets, generating code, and orchestrating... Autonomous AI agents are taking on all types of work for businesses: routing logistics fleets, triaging support tickets, generating code, and orchestrating multistep workflows. How do you take a general-purpose model and make it excel at your specific task? Customization provides an agent with the right capabilities. This post explains nine techniques for customizing AI agents… Source

Agent harnesses like Claude Code, Codex, and LangChain Deep Agents are excellent orchestrators. They manage sessions, chain tools, execute code, and respond to... Agent harnesses like Claude Code, Codex, and LangChain Deep Agents are excellent orchestrators. They manage sessions, chain tools, execute code, and respond to developer intent. But when these harnesses need to do deep research, such as multi-document synthesis, decision briefs backed by enterprise data, and long-horizon analysis with source attribution, the complexity of deep research shifts back… Source

Autonomous AI agents are becoming more capable. Open models, Model Context Protocol (MCP)-connected tools, and portable skills are also making agents easier to... Autonomous AI agents are becoming more capable. Open models, Model Context Protocol (MCP)-connected tools, and portable skills are also making agents easier to extend. But scaling agent use with structural transparency and operational integrity requires more than runtime guardrails. Organizations and teams need to understand and trust the skills, or instructions, an agent is using. Source

Evaluating an AI model and evaluating an AI agent are related—but they answer fundamentally different questions. A model benchmark tests the capability of a... Evaluating an AI model and evaluating an AI agent are related—but they answer fundamentally different questions. A model benchmark tests the capability of a foundation model (how well it understands language, follows instructions, or solves problems on static tasks). An agent evaluation tests the behavior of a system operating end-to-end—planning, calling tools, handling uncertainty… Source

Maximizing the value of AI infrastructure demands deep visibility into GPU utilization. Yet many platform teams running AI workloads on Kubernetes operate with... Maximizing the value of AI infrastructure demands deep visibility into GPU utilization. Yet many platform teams running AI workloads on Kubernetes operate with limited visibility into how their GPUs are used. Most don’t know who’s consuming them, how much memory is in use, and whether Kubernetes pods are pending or silently idle. Without a signal, GPU fleets are routinely underutilized and slow to… Source

Agentic inference has fundamentally changed the runtime dynamics of inference workloads by introducing non-deterministic trajectories—actions, observations,... Agentic inference has fundamentally changed the runtime dynamics of inference workloads by introducing non-deterministic trajectories—actions, observations, and decisions that an AI agent produces while working through a task. These trajectories compound end-to-end latency across hundreds of inference requests per session. NVIDIA Vera Rubin NVL72 handles the bulk of that inference load as… Source

A massive-scale X-ray free-electron laser (XFEL) enables tracking structural and electron dynamics in novel systems, including fusion materials, semiconductors,... A massive-scale X-ray free-electron laser (XFEL) enables tracking structural and electron dynamics in novel systems, including fusion materials, semiconductors, batteries, and catalysis. It produces ultrashort X-ray pulses that can record the movements of atoms and electrons. These instruments can detect the smallest change in material structure caused by defects and other influences. Source

In today’s data-driven world, organizations increasingly rely on video to capture critical information, yet extracting meaningful, real-time insights from... In today’s data-driven world, organizations increasingly rely on video to capture critical information, yet extracting meaningful, real-time insights from massive amounts of footage remains a challenge. NVIDIA Metropolis Blueprint for video search and summarization (VSS) overcomes this hurdle by transforming millions of live video streams or hours of recorded video into instantly searchable… Source

The path from a trained AI model to production should be smooth, but rarely is. Many teams invest weeks fine-tuning models, only to discover that exporting to a... The path from a trained AI model to production should be smooth, but rarely is. Many teams invest weeks fine-tuning models, only to discover that exporting to a deployment format breaks layers, input shapes cause runtime failures, or version mismatches silently degrade performance. These issues are collectively known as pipeline friction, and they cost organizations time, money… Source

The compute capability of large GPU fleets presents unprecedented opportunities to innovate and provide value to customers in record time. Yet these... The compute capability of large GPU fleets presents unprecedented opportunities to innovate and provide value to customers in record time. Yet these advancements come with a variety of challenges. At scale, teams are juggling heterogeneous hardware, fast‑moving software stacks, tight power envelopes, and spiky, multitenant workloads. A single hotspot, misconfigured driver, or subtle hardware fault… Source

Bash is one of the most flexible and powerful interfaces exposed to AI agents. In the right system, a model that emits grep, curl, tar, or a shell pipeline is... Bash is one of the most flexible and powerful interfaces exposed to AI agents. In the right system, a model that emits , , , or a shell pipeline is producing an executable action that can read files, mutate a workspace, open network connections, and chain tools together. For the NVIDIA AI Red Team, this makes command generation a useful research target. If smaller language models can be guided… Source

An agentic exchange must preserve a structured interaction: assistant turns interleave reasoning with one or more tool calls, and subsequent user turns return... An agentic exchange must preserve a structured interaction: assistant turns interleave reasoning with one or more tool calls, and subsequent user turns return the corresponding tool results to the model context. Reasoning replay is model- and turn-dependent: some reasoning should be retained, while some should be dropped. The inference engine is responsible for supporting this more expressive… Source

NVIDIA GB200 NVL72 introduces a fundamentally new way to build GPU clusters by extending NVIDIA NVLink coherence across an entire rack. This design enables... NVIDIA GB200 NVL72 introduces a fundamentally new way to build GPU clusters by extending NVIDIA NVLink coherence across an entire rack. This design enables exascale performance, but it also changes the assumptions that many scheduling systems were built on. As a result, “rack-scale locality” becomes a hard constraint. When workloads cross domain boundaries, performance drops sharply… Source

Model quantization is an effective method to reduce VRAM usage and improve inference performance on consumer devices such as NVIDIA GeForce RTX GPUs. By... Model quantization is an effective method to reduce VRAM usage and improve inference performance on consumer devices such as NVIDIA GeForce RTX GPUs. By lowering computational and memory requirements while preserving model quality, quantization helps AI models run more efficiently in resource-constrained environments. This post walks through how to use NVIDIA Model Optimizer to quantize a… Source

Distributed deep learning depends on fast, reliable GPU-to-GPU communication using the NVIDIA Collective Communication Library (NCCL). When training slows down,... Distributed deep learning depends on fast, reliable GPU-to-GPU communication using the NVIDIA Collective Communication Library (NCCL). When training slows down, it becomes challenging to determine why and what to do next. A problem can span computation, communication, a specific rank, or underlying hardware. NVIDIA NCCL Inspector accelerates triaging by providing a lightweight and continuous… Source

The automotive cockpit is undergoing a fundamental shift from rule-based interfaces to agentic, multimodal AI systems capable of reasoning, planning, and... The automotive cockpit is undergoing a fundamental shift from rule-based interfaces to agentic, multimodal AI systems capable of reasoning, planning, and acting. In most vehicles on the road today, in-vehicle assistants still rely on fixed command-response patterns: interpret a phrase, trigger an action, reset. While effective for well-defined tasks, this approach doesn’t scale to modern… Source

Generative AI’s explosive first chapter was defined by humans sending requests and models responding. The agentic chapter is different. Agents don't... Generative AI’s explosive first chapter was defined by humans sending requests and models responding. The agentic chapter is different. Agents don’t follow a pre-determined sequence of actions. They call tools, spawn sub-agents with different tasks and models, retain information in memory, manage their own context window, and decide for themselves when they’re finished. In doing so… Source

Modern supply chains operate under the constant pressures of fluctuating demand, volatile costs, constrained capacity, and interdependent decision-making.... Modern supply chains operate under the constant pressures of fluctuating demand, volatile costs, constrained capacity, and interdependent decision-making. Traditionally, specialized operations research (OR) teams solved these problems by translating business questions into mathematical models. This process can take weeks and often produces fragile solutions that struggle to adapt when conditions… Source

Today, game developers can begin integrating NVIDIA DLSS 4.5 with Dynamic Multi Frame Generation, Multi Frame Generation 6X, and the second-generation... Today, game developers can begin integrating NVIDIA DLSS 4.5 with Dynamic Multi Frame Generation, Multi Frame Generation 6X, and the second-generation transformer model for NVIDIA Super Resolution. In this post, we’ll go over new technologies and resources to share with our game-developer community, including: At CES 2026, we introduced DLSS 4.5, extending its AI-driven… Source

Neural network techniques are increasingly used in computer graphics to boost image quality, improve performance, and streamline content creation. Approaches... Neural network techniques are increasingly used in computer graphics to boost image quality, improve performance, and streamline content creation. Approaches like super resolution, denoising, and neural rendering help real-time engines work more efficiently, offering new creative possibilities while keeping performance in mind. Unreal Engine 5 (UE5) has taken several steps in this direction… Source

Creative and visualization teams today produce more assets, in more formats, with leaner teams. Generative AI can accelerate that work – compressing tasks... Creative and visualization teams today produce more assets, in more formats, with leaner teams. Generative AI can accelerate that work – compressing tasks that once took hours of manual effort into automated, repeatable pipelines. ComfyUI is an open-source, node-based creative tool that runs locally on NVIDIA RTX GPUs. It connects image generation, video synthesis, and language models into… Source