31-05-2025 |
NVIDIA Blackwell Boosts AI Factories with 40x Performance Gain |
Feature |
NVIDIA Blackwell architecture powers AI factories, delivering a remarkable 40x improvement in AI reasoning performance while using the same power as the Hopper architecture. This advancement, driven by full-stack optimizations, enhances efficiency and speed for data centers over time. Ideal for premium subscription users, NVIDIA Blackwell offers industry insights for cutting-edge AI inference. Explore more about NVIDIA Blackwell at the provided link to elevate your AI infrastructure today! |
|
30-05-2025 |
NVIDIA Showcases AI Innovations at CVPR 2025 in Nashville |
Feature |
NVIDIA highlights groundbreaking computer vision research at CVPR 2025, featuring 61 papers on topics like 3D modeling and human motion forecasting. Attendees can explore NVIDIA’s advanced tools, including Omniverse for simulation and generative AI for edge applications. Gain industry insights through workshops, demos, and networking events like the Researcher Celebration. Register now for the best rates to join NVIDIA in advancing AI and computer vision. |
|
29-05-2025 |
NVIDIA GTC Paris Offers Advanced AI Workshops |
Workshops |
NVIDIA’s GTC Paris, held June 10–12, 2025, introduces hands-on workshops like NVIDIA Isaac for robotics and Omniverse for digital twins, equipping developers with cutting-edge AI skills. These full-day sessions, priced at €275, provide technical certifications and practical experience in AI agents, CUDA programming, and RAG pipelines. Attendees can also access free two-hour training labs and proctored certification exams during the conference. Register now at nvidia.com/en-eu/gtc/training to secure your spot and elevate your AI expertise. |
|
29-05-2025 |
NVIDIA NeMo Guardrails Boosts Real-Time AI Safety |
AI Safety and Security |
NVIDIA NeMo Guardrails enhances generative AI applications by ensuring safe, low-latency streaming of large language model outputs. It reduces user wait times through incremental token streaming while maintaining robust safety checks, preventing issues like prompt injections or data leaks. The platform’s configurable chunk processing and context-aware moderation optimize responsiveness and compliance for enterprise use. |
|
29-05-2025 |
NVIDIA-Backed DreaMS Accelerates Molecular Discovery |
Company News |
Researchers from IOCB Prague and CIIRC CTU, supported by NVIDIA’s AI technology, have developed DreaMS, a machine learning model that rapidly analyzes unknown molecules using mass spectrometry data, as published in Nature Biotechnology. Led by Dr. Tomáš Pluskal and Dr. Josef Šivic, with contributions from Roman and Anton Bushuiev, DreaMS leverages NVIDIA’s advanced AI to interpret complex molecular spectra, uncovering chemical similarities and detecting elements like fluorine, vital for drug development. This innovation could revolutionize medicine, environmental science, and space exploration by identifying new natural compounds. Explore the DreaMS Atlas and its potential at uochb.cz. |
|
23-05-2025 |
El Agente Q Simplifies Quantum Chemistry Workflows |
AI Innovation Update |
El Agente Q, developed by NVIDIA and the University of Toronto’s Matter Lab, is an innovative LLM-based system that streamlines quantum chemistry tasks through natural language prompts. Its multi-agent architecture efficiently decomposes tasks, selects appropriate tools, and executes workflows, achieving an 87% success rate in university-level exercises. Designed for accessibility, it supports researchers and students in exploring molecular properties with ease. Visit the platform to experience its user-friendly interface for computational chemistry. |
|
23-05-2025 |
AceReason-Nemotron Enhances Math and Code Reasoning |
New Releases |
NVIDIA has released AceReason-Nemotron-14B, a model improving math and code reasoning through reinforcement learning, starting from the DeepSeek-R1-Distill-Qwen-14B base. It shows notable gains in benchmarks like AIME 2024 (69.7 to 78.6) and LiveCodeBench v5 (53.1 to 61.1), using separate math and code-focused RL strategies. The approach boosts foundational reasoning and tackles complex problems, with minimal impact on math skills during code RL iterations. Explore the model and its training details on Hugging Face. |
|
22-05-2025 |
cuML Powers Kaggle Win with GPU-Accelerated Stacking |
Company News |
NVIDIA’s cuML library enabled Kaggle Grandmaster Chris Deotte to secure first place in the April 2025 Kaggle Playground competition. Using GPU-accelerated stacking, cuML rapidly trained diverse models to predict podcast listening times with high accuracy. The technique combined hundreds of models into a layered ensemble for superior performance. Discover how cuML’s speed and versatility can enhance your data science projects |
|
22-05-2025 |
Linker Vision Enhances Kaohsiung with NVIDIA AI Solutions |
Case Studies |
Linker Vision leverages NVIDIA Metropolis and Omniverse to transform Kaohsiung’s city operations, reducing emergency response times by up to 80%. Using vision AI and digital twins, their platform breaks departmental silos, enabling real-time data sharing and smarter decision-making. The NVIDIA AI Blueprint for video analytics powers AI agents to detect and respond to urban incidents effectively. Learn how NVIDIA’s advanced tools can optimize smart city infrastructure. |
|
22-05-2025 |
NVIDIA Cosmos Wins Top COMPUTEX 2025 Awards |
Awards & Honours |
NVIDIA Cosmos, a platform for accelerating physical AI development, secured both the Best Choice of the Year and Golden Awards at COMPUTEX 2025. It enables rapid synthetic data generation, processing 20 million hours of video in two weeks, significantly reducing development time for robotics and autonomous vehicles. Built on the NVIDIA Blackwell architecture, Cosmos enhances AI model training with unmatched speed and scale. Discover how NVIDIA Cosmos is shaping the future of autonomous systems. |
|
21-05-2025 |
NVIDIA Llama Nemotron Nano 4B Boosts Edge AI |
AI Innovation Update |
NVIDIA’s Llama Nemotron Nano 4B, an open reasoning model, delivers superior accuracy and 50% higher throughput compared to other open models with 8 billion parameters or fewer, excelling in scientific tasks, coding, and complex math. Optimized for edge deployment on NVIDIA Jetson and RTX GPUs, it supports hybrid reasoning for low-cost, secure inference. This model is ideal for developers building efficient AI agents for real-time applications. Access it now on Hugging Face to enhance your edge AI projects. |
|
20-05-2025 |
NVIDIA Isaac Sim 5.0 Boosts Robotics Development |
AI Innovation Update |
NVIDIA unveiled Isaac Sim 5.0 and Isaac Lab 2.2 at COMPUTEX 2025, introducing advanced sensor simulation, synthetic data generation, and actuator modeling for robotics. These updates, built on NVIDIA Omniverse, enable customizable simulations and faster development cycles, enhancing robot training for real-world applications. Key features include MobilityGen for diverse motion data and standardized ROS 2 interfaces for seamless integration. Developers can access these tools to accelerate innovation in robotics, from humanoid robots to industrial automation. |
|
19-05-2025 |
NVIDIA Omniverse Powers Digital Twins for Industrial AI |
Showcase |
Leading Taiwanese manufacturers, including Foxconn, TSMC, and Wistron, are leveraging NVIDIA Omniverse to create OpenUSD-based digital twins, transforming factory planning and industrial AI. These digital twins enhance operational efficiency, accelerate robotics development, and optimize complex processes like defect detection and thermal assessments. By integrating NVIDIA AI Blueprints and Metropolis, companies achieve real-time analytics and automation. |
|
19-05-2025 |
NVIDIA Metropolis Boosts Video Analysis for Enterprises |
AI Innovation Update |
The NVIDIA Metropolis platform powers the AI Blueprint for Video Search and Summarization (VSS), enabling businesses to deploy AI agents for efficient video data processing. These agents enhance operations in sectors like manufacturing, smart cities, and sports by analyzing extensive video feeds. Companies such as Siemens and Pegatron have achieved notable cost savings and productivity improvements. Visit build.nvidia.com to discover how VSS can unlock valuable insights from video content. |
|
19-05-2025 |
FBC: Firebreak Launches with DLSS 4 Enhancements |
Games Update |
Remedy Entertainment’s FBC: Firebreak, a cooperative first-person shooter, launches on June 17, 2025, featuring NVIDIA’s DLSS 4 with Multi Frame Generation and Ray Reconstruction for enhanced visuals and performance. The game leverages full ray tracing, delivering stunning graphics on GeForce RTX GPUs. An exclusive #RTXOn trailer showcases these advancements in action. Wishlist FBC: Firebreak on Steam or Epic Games Store to experience its immersive gameplay |
|
19-05-2025 |
NVIDIA AI Data Platform Powers Enterprise AI at COMPUTEX 2025 |
Tech Events |
NVIDIA collaborates with top storage leaders like IBM, NetApp, and VAST Data to enhance the NVIDIA AI Data Platform, showcased at COMPUTEX 2025. This platform integrates NVIDIA Blackwell GPUs and BlueField DPUs to accelerate data processing, enabling AI agents to deliver real-time insights from enterprise data. Partners such as Dell Technologies and Hitachi Vantara are building solutions to improve AI-driven decision-making across industries. Explore these advancements at nvidia.com to see how they can transform your business operations. |
|
19-05-2025 |
NVIDIA and Foxconn Partner to Build AI Factory with Blackwell GPUs in Taiwan |
Collaboration |
NVIDIA collaborates with Foxconn and Taiwan to launch an AI factory featuring 10,000 NVIDIA Blackwell GPUs, boosting innovation in smart cities, electric vehicles, and manufacturing. The Big Innovation Cloud AI factory will empower TSMC researchers and Taiwan’s tech ecosystem with advanced AI computing resources. Join the virtual COMPUTEX 2025 keynote to explore how this initiative accelerates AI-driven breakthroughs. Register now to learn more about NVIDIA’s cutting-edge infrastructure. |
|
19-05-2025 |
NVIDIA Unveils DGX Spark and DGX Station for AI-Powered Personal Computing |
AI Innovation Update |
NVIDIA launches DGX Spark and DGX Station, AI-first personal computing systems delivering 1000 AI TOPS for developers and researchers. Available from partners like Acer, ASUS, Dell, and HP starting July 2025, DGX Spark won the COMPUTEX 2025 Best Choice Award. Reserve now to prototype, fine-tune, and deploy AI models up to 200 billion parameters. Explore how these systems accelerate innovation at NVIDIA’s GTC 2025 keynote replay. |
|
19-05-2025 |
NVIDIA and Microsoft Enhance RTX AI PCs with TensorRT and Windows ML |
Company News |
NVIDIA and Microsoft unveiled advancements for RTX AI PCs at MSBuild and COMPUTEX 2025, integrating TensorRT for RTX with Windows ML to boost AI app performance by over 50% compared to DirectML. This collaboration streamlines development with just-in-time engine building and supports NVIDIA NIM microservices and Project G-Assist plug-ins. Developers can access these tools via Windows ML preview or NVIDIA Developer in June. Explore the NVIDIA RTX AI Garage blog for community-driven AI innovations. |
|
19-05-2025 |
NVIDIA TensorRT for RTX Boosts AI Performance on Windows 11 PCs |
Feature |
NVIDIA’s TensorRT for RTX, integrated with Windows ML, delivers over 50% faster AI inference on RTX GPUs compared to DirectML, optimizing apps in gaming, creativity, and productivity. This lightweight 200 MB library uses just-in-time compilation for tailored performance in under 30 seconds. Developers can access it via Windows ML preview or as a standalone SDK in June. Visit NVIDIA’s developer site to explore and enhance AI workflows. |
|
19-05-2025 |
NVIDIA GeForce RTX 5060 Laptop GPUs Launch with Advanced AI Capabilities |
Feature |
NVIDIA’s GeForce RTX 5060 Laptop GPUs, powered by the Blackwell architecture, are available starting May 19, 2025, in laptops from major manufacturers like ASUS, MSI, and Razer, priced from $1,099. These ultra-thin laptops, as light as 1.46kg, deliver double the performance of previous generations with DLSS 4 Multi Frame Generation and Max-Q optimization, offering over 100 FPS in games. Creators and students benefit from enhanced AI-driven workflows in apps like DaVinci Resolve. Visit NVIDIA’s Marketplace to explore models and upgrade for cutting-edge gaming and productivity. |
|
19-05-2025 |
NVIDIA’s AI Robotics Breakthrough Unveiled at COMPUTEX |
AI Conferences |
NVIDIA’s CEO Jensen Huang showcased advancements in AI-driven robotics during his COMPUTEX 2025 keynote, emphasizing technologies that accelerate robotic automation across industries. The keynote highlighted NVIDIA’s latest tools for robotics development, including AI workflows for manufacturing and logistics. Attendees can explore workshops and sessions to dive deeper into these innovations. Watch the keynote replay to learn how NVIDIA is shaping the future of robotics. |
|
17-05-2025 |
NVIDIA Showcases AI Innovation at Red Hat Summit 2025 |
Summits |
NVIDIA will present six sessions at Red Hat Summit 2025, held May 19-22 in Boston, highlighting AI integration with open source technologies. Key sessions include a keynote on modernized infrastructure for enterprise AI, a breakout on generative AI at Bank of New York using Red Hat Enterprise Linux, and a panel on cloud-native AI model serving with Red Hat and Google. These sessions demonstrate how NVIDIA’s AI solutions, paired with Red Hat’s platforms, enhance performance and scalability in industries like finance and telecom. Visit the Red Hat Summit session catalog to explore NVIDIA’s full lineup and register. |
|
16-05-2025 |
NVIDIA CEO Jensen Huang to Deliver COMPUTEX 2025 Keynote |
Tech Events |
NVIDIA's Jensen Huang will present the COMPUTEX 2025 keynote on May 19 at 11:00 a.m. Taipei Time, livestreamed globally. The event, hosted at Taipei Music Center, includes a pre-keynote broadcast with industry experts. Attendees can explore workshops, sessions, and a chance to win an NVIDIA RTX 5080 graphics card. Tune in to the livestream to catch the latest NVIDIA innovations. |
|
16-05-2025 |
Mistral Medium 3 Enhances Multimodal AI with NVIDIA NIM |
Service |
Mistral Medium 3, now available through NVIDIA NIM, offers advanced multimodal capabilities, processing both text and images with a 128K context window for enterprise applications. It supports over 45 languages, excelling in tasks like document parsing, coding, and visual question answering. The NVIDIA NIM microservice ensures efficient deployment on NVIDIA Hopper GPUs, simplifying integration for developers. Try it at build.nvidia.com to elevate your AI projects. |
|