NVIDIA Benchmarking LLM Inference at Scale with AIPerf
NVIDIA introduced AIPerf, a benchmarking tool designed to measure and evaluate large language model inference performance at scale.
Source: NVIDIA Developer Blog Silicon and supply chain
NVIDIA
AI news about NVIDIA: product launches, model updates, and industry moves. Latest: Benchmarking LLM Inference at Scale with AIPerf
NVIDIA NVIDIA Cute Critters Come to the Cloud: ‘Aniimo’ Launches on GeForce NOW
Pawprint Studio's creature-catching game Aniimo is now available on NVIDIA's GeForce NOW cloud gaming service, alongside a path-tracing update for 007 First Light.
Source: NVIDIA BlogNVIDIA How to Use AI Agents to Prepare 3D Scenes for Simulation
AI agents can be deployed to prepare and validate 3D scenes and digital twins for physical AI systems and simulation use cases.
Source: NVIDIA Developer BlogNVIDIA TensorRT Edge-LLM Completes the MLPerf Edge Agentic Benchmark 6.4x Faster on Jetson AGX Thor
TensorRT Edge-LLM completed the MLPerf Edge Agentic Benchmark 6.4x faster on Jetson AGX Thor hardware, marking advancement in edge AI agent deployment.
Source: NVIDIA Developer BlogNVIDIA Translating CUDA Tile Operations from Python to Rust Using Agentic AI
Researchers used agentic AI to translate CUDA tile operations from Python to Rust, leveraging AI to improve Rust-based GPU kernel development.
Source: NVIDIA Developer BlogNVIDIA NVIDIA Vera Rubin NVL72 Delivers Leading Performance in MLPerf Inference v6.1 Debut
NVIDIA's Vera Rubin NVL72 system achieved leading performance in MLPerf Inference v6.1 benchmarks, demonstrating advances in AI inference efficiency.
Source: NVIDIA BlogNVIDIA Emerald AI, Google and NVIDIA Launch Alliance to Advance Flexible AI Data Centers
Emerald AI, Google, and NVIDIA announced the formation of the AI Energy Management Alliance to advance data centers with dynamic electricity management capabilities.
Source: NVIDIA BlogNVIDIA University of Manchester Uses NVIDIA Earth-2 to Forecast Air Pollution Across the UK
University of Manchester uses NVIDIA Earth-2 to forecast air pollution across the UK.
Source: NVIDIA BlogNVIDIA ‘Now We Can Know Everything and Do Anything,’ Jensen Huang Says at Dreamforce
NVIDIA CEO Jensen Huang presents at Salesforce Dreamforce, introducing Koa, Salesforce's first CRM reasoning model built on NVIDIA Nemotron 3 Super.
Source: NVIDIA BlogNVIDIA Dense vs. MoE Models: Active Parameters, Throughput, and When to Choose Each
NVIDIA explains how mixture-of-experts models like Nemotron 3.5 Lightning activate only 3B parameters despite having 30B total parameters.
Source: NVIDIA Developer BlogNVIDIA From Megawatts to Tokens: How NVIDIA Maximizes AI Factory Production
NVIDIA discusses power management and efficiency optimization strategies for AI factories.
Source: NVIDIA BlogNVIDIA AI Infra Summit: NVIDIA Vera Rubin and DSX Platform Advancements Showcase Energy Efficiencies of Optimizing Tokens Per Watt for AI Factories
NVIDIA AI Infra Summit showcases AI factory efficiency advancements and tokens-per-watt optimization.
Source: NVIDIA BlogNVIDIA How NVIDIA Groq 3 LPX Deterministic Execution Drives Power-Efficient High-Interactivity Inference on NVIDIA Vera Rubin
NVIDIA describes how Groq 3 LPX deterministic execution enables power-efficient inference on Vera Rubin.
Source: NVIDIA Developer BlogNVIDIA How NVIDIA NVLink 6 Delivers Multi-Layer Resiliency for AI Factories
NVIDIA NVLink 6 provides multi-layer resiliency for large-scale AI factory operations.
Source: NVIDIA Developer BlogNVIDIA Scaling Federated Learning Across Docker, Kubernetes, and Slurm with NVIDIA FLARE
NVIDIA FLARE enables federated learning to scale across Docker, Kubernetes, and Slurm environments.
Source: NVIDIA Developer BlogNVIDIA Heart of the Matter: How a Major Children’s Hospital Uses Open Source NVIDIA AI for Cardiac Care
A major children's hospital is utilizing open-source NVIDIA AI technology to enhance cardiac care capabilities.
Source: NVIDIA BlogNVIDIA NVIDIA NVLink Fusion Brings NVHBM to Next-Generation AI Infrastructure
NVIDIA announces NVLink Fusion with NVHBM technology to meet growing compute requirements for larger AI models and complex reasoning tasks in modern AI infrastructure.
Source: NVIDIA Developer BlogNVIDIA How to Train a Cross-Embodiment Robot Navigation Policy with AI Agents
NVIDIA explores training robot navigation policies using cross-embodiment AI agents, enabling robots to convert perception into autonomous navigation capability.
Source: NVIDIA Developer BlogNVIDIA Experiment with Qwen3.8-Flash-Next 176B Model on NVIDIA GB300 NVL72 for Agentic Coding
Alibaba released preview weights of the Qwen3.8-Flash-Next 176B model for developers to experiment with before the full Qwen4 architecture launch.
Source: NVIDIA Developer BlogNVIDIA Restore LLM Inference Capacity in Seconds with Shadow Engine Recovery in NVIDIA Dynamo
NVIDIA introduces Shadow Engine Recovery for Dynamo, a technique to restore LLM inference capacity in seconds instead of performing lengthy cold restarts that require reloading weights and recompiling kernels.
Source: NVIDIA Developer BlogNVIDIA Leading Publishers Bring Blockbuster PC Games and Technology to NVIDIA RTX Spark
NVIDIA showcases RTX gaming support at Gamescom with new titles from publishers including EA, Embark, and Ubisoft, featuring improved anti-cheat technology and visual quality.
Source: NVIDIA BlogNVIDIA CUDA Python 1.0: Stable APIs, One Foundation, Full Platform Access
NVIDIA releases CUDA Python 1.0, providing stable APIs that enable Python developers to access GPU computing without requiring deep C++ expertise or complex build toolchain setup.
Source: NVIDIA Developer BlogNVIDIA Giga-Scale AI and the Ethernet Evolution: How Spectrum-X Ethernet Rewrites the Rules
NVIDIA Spectrum-X Ethernet infrastructure addresses data center design challenges for distributed AI model training spanning hundreds of thousands of GPUs.
Source: NVIDIA Developer BlogNVIDIA How XPUs Meet a World-Class AI Factory
NVIDIA discusses XPU design for AI infrastructure, emphasizing that AI factories should be built as integrated systems optimized for tokens per second, energy efficiency, and cost per token—rather than collections of individual accelerators.
Source: NVIDIA BlogNVIDIA With Groq 3 LPX in Full Production, NVIDIA Extends Vera Rubin Inference for Agents
NVIDIA is extending its Vera Rubin NVL72 inference system with faster token generation capabilities to support agentic AI systems in full production.
Source: NVIDIA BlogNVIDIA Up to 30x More Work Per Watt: NVIDIA Vera Rubin NVL72 Sets a New Efficiency Standard for AI Agents
NVIDIA's Vera Rubin NVL72 achieves up to 30x more work per watt for AI agents. The system addresses the fact that agentic workloads consume approximately 15x more tokens than simple chat requests due to activities like database queries and sub-agent invocations.
Source: NVIDIA BlogNVIDIA NVIDIA Vera Rubin and Blackwell Set a New Standard for Agentic AI Performance per Watt
NVIDIA Vera Rubin and Blackwell platforms optimize performance per watt for agentic AI systems performing multi-step workflows with reasoning, tool invocation, and subagent coordination.
Source: NVIDIA Developer BlogNVIDIA NVIDIA BlueField-4 Powers New Scale-In Network Infrastructure for Agentic AI Factories
NVIDIA BlueField-4 network accelerator enables new scale-in infrastructure architecture for agentic AI factories connecting diverse users and workloads.
Source: NVIDIA Developer BlogNVIDIA Solving Agentic AI Fleet Challenges with NVIDIA Vera CPU
NVIDIA Vera CPU optimizes fleet economics in AI factories by improving the efficiency of converting power and capital investment into completed agent tasks.
Source: NVIDIA Developer BlogNVIDIA How NVIDIA Groq 3 LPX Unlocks Ultrafast Interactivity at Long Context on NVIDIA Vera Rubin
NVIDIA Groq 3 LPX is an interactive inference accelerator for the NVIDIA Vera Rubin platform, delivering ultrafast performance for long-context AI inference.
Source: NVIDIA Developer BlogNothing here yet.