NVIDIA Noticias de IA sobre NVIDIA: lanzamientos, novedades de modelos y movimientos del sector. Latest: Scale Agentic Coding Workflows with Meta’s Muse Spark Open on NVIDIA GB300 NVL72
NVIDIA Meta launches Muse Spark Open, a coding-focused mixture-of-experts model with 816B total parameters, 42B active, and a 1M-token context window. NVIDIA's developer blog covers scaling agentic coding workflows with it on GB300 NVL72 systems.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA announces the launch of RTX Spark, a Windows PC platform aimed at creators, gamers, and AI agents. Launch events are planned around the world.
Fuente: NVIDIA BlogNVIDIA NVIDIA released Nemotron 3.5 Super VL, an open 120-billion-parameter model, plus two edge models, Bolt 74B-A9B 24GB and Bolt E4B 4GB, for local agentic workloads.
Fuente: NVIDIA BlogNVIDIA NVIDIA's developer blog describes how Nemotron 3.5 Super VL supports always-on agents with multimodal reasoning for longer, more complex enterprise workflows.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA's DLSS 4.5 controls are now available on GeForce NOW, and Enshrouded has left early access with its 1.0 release.
Fuente: NVIDIA BlogNVIDIA NVIDIA says it is expanding access to advanced technology for Europe's scientific community through open learning resources, collaborations and compute. It notes that smaller labs and early-career scientists often find pre-research work the hardest part.
Fuente: NVIDIA BlogNVIDIA An NVIDIA blog post says the Nordic countries lead Europe in unicorn startup density, with the most $1B+ emerging companies per capita, and describes the region's push to build a major computing ecosystem.
Fuente: NVIDIA BlogNVIDIA An NVIDIA Developer Blog post describes using NeMo Gym to improve agent reliability and tool-calling efficiency in chip design. It notes that AI agents can boost productivity in complex chip design workflows but may generate plausible yet incorrect artifacts.
Fuente: NVIDIA Developer BlogNVIDIA An NVIDIA blog post describes the U.K.'s AI ecosystem moving from ambition to deployment across industries. It cites a five-year, $250 million compute commitment to strengthen the U.K.'s agentic safety work. The excerpt is truncated.
Fuente: NVIDIA BlogNVIDIA An NVIDIA blog post reports that EMEA cities and local partners are using NVIDIA accelerated computing to build physical AI agents and digital twins. The tools help teams decide on road repairs and energy efficiency.
Fuente: NVIDIA BlogNVIDIA NVIDIA Developer Blog describes how to rewrite Apache Spark UDFs for GPU acceleration with AI agents, citing growing enterprise data demands driven by agentic AI.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA is making the full Run:ai platform available as open-source software, as part of the NVIDIA DSX AI Factory Platform.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA announces RTX Spark, a PC platform described as combining gaming PC, creative studio, and personal AI supercomputer capabilities, with up to 128GB unified memory and 1 petaflop of AI compute. The post also covers global PC app ecosystem readiness.
Fuente: NVIDIA BlogNVIDIA NVIDIA's blog reports that companies in France are deploying agentic computing across space, finance, the public sector and research. It credits the 2025 AI Action Summit and the 2025 and 2026 Choose France Summits with anchoring major compute commitments.
Fuente: NVIDIA BlogNVIDIA NVIDIA Developer Blog discusses ReplaySSM, a state-storage approach for faster decoding in hybrid models that mix state-space model (SSM) layers with a few attention layers. The excerpt names Nemotron 3, Qwen3.5 through Qwen3.8, Kimi-K3, and GLM-5.3-Flash.
Fuente: NVIDIA Developer BlogNVIDIA At the OCP Global Summit in San Jose, NVIDIA announced partnerships and contributions to open AI infrastructure standards and opened its Vera Rubin architecture to help scale AI factories. The excerpt says the aim is to build them faster, more efficiently, and more reliably.
Fuente: NVIDIA BlogNVIDIA An NVIDIA blog post describes how developers combine frontier AI models with Omniverse libraries to build simulation applications, directing AI agents to explore scenarios and investigate failures.
Fuente: NVIDIA BlogNVIDIA An NVIDIA developer blog outlines five steps for creating SimReady robotics assets from CAD files with frontier AI models, including configuring and validating materials and collision.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA's KGMON team placed second in the KDD Cup 2026 Data Agents competition with a system built around a smaller agent harness.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA announces that Gears of War: E-Day is launching on GeForce NOW, and Fire TV users will soon be able to buy GeForce NOW memberships directly.
Fuente: NVIDIA BlogNVIDIA At a Microsoft event in San Francisco, Jensen Huang and Satya Nadella described how NVIDIA and Microsoft are co-engineering hardware and software so AI agents can run on Windows PCs. Huang framed the move as AI agents coming to Windows.
Fuente: NVIDIA BlogNVIDIA NVIDIA shares insights from automating the assembly of GB300 tester trays using robotics, discussing lessons about robot learning, mechanical intelligence, and engineering practices.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA explores using digital twins and AI agents to validate changes in complex AI factory operations, which combine GPUs, CPUs, network switches, DPUs, and orchestration software.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA's cuOpt with mPDLP scaling enables decision optimization for supply chain and energy grid problems with up to 100 million variables, addressing increasingly complex logistics and distributed energy management.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA's cuPhoton accelerates scientific image analysis for high-throughput instruments like telescopes and X-ray sources, enabling processing speeds that exceed traditional CPU-based approaches.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA Developer Blog describes using NVIDIA Megatron Core to scale bitwise-deterministic pretraining, which makes large-scale training easier to debug, validate and resume reproducibly.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA introduces GPUNetIO to enable GPU applications to control networking and data movement directly as native GPU operations, rather than relying on host-driven services.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA releases AICR v1.0, an open standard for configuring GPU-accelerated Kubernetes clusters with verified compatibility across host kernels and GPU drivers.
Fuente: NVIDIA Developer BlogNVIDIA NVIDIA introduces Green Contexts technology to manage GPU workload sharing for applications with multiple independent components running simultaneously within a single process.
Fuente: NVIDIA Developer BlogNVIDIA Telecom operators increasingly adopt open-source AI models for their strategies, prioritizing control and customization capabilities for critical workloads like autonomous networks and customer care, per NVIDIA's telecommunications industry report.
Fuente: NVIDIA BlogTodavía no hay contenido.
