Search “NVIDIA”

NVIDIA RTX Spark Laptop Update: Adds Nemotron 3.5 Transcribe and Local Agent Tools

NVIDIA announced multiple updates for RTX Spark thin-and-light laptops, including the open-weight voice model Nemotron 3.5 Transcribe, support for Liquid AI's LFM3 model family, and a control in the NVIDIA App that can allocate GPU resources between foreground applications and background agents. Nemotron 3.5 Transcribe supports transcription, speaker labeling, and translation in 40 languages, has approximately 4.3 billion active parameters, and is now available on Hugging Face. RTX Spark laptops are now shipping from ASUS, Dell, HP, Lenovo, Microsoft, and MSI, with a starting price of $2,599.

MiniMax-M3.1-Flash Opens Its Weights, NVIDIA NeMo Supports Day-0 Fine-Tuning

MiniMax officially thanked the NVIDIA NeMo team for providing day-0 fine-tuning support for MiniMax-M3.1-Flash. The cited post by Huiying Li states that before its public release, the model had processed 200T+ tokens through Space Bunny and has now opened its weights; NeMo-AutoModel's day-0 fine-tuning support includes FSDP2, expert parallelism, pipeline parallelism, and context parallelism, can scale up to 128 GPUs, and can directly load NVFP4 checkpoints.

NVIDIA DGX Spark supports eight-node clusters, unified memory expands to 1TB

At GTC Berlin, NVIDIA announced that DGX Spark officially supports eight-node clusters, with unified memory expanding to 1TB, enabling local execution of frontier models such as Z.ai's full GLM 5.3 (753B). Official benchmarks show that Qwen3.8-27B (NVFP4) decoding throughput rises from about 12 tokens/s on a single node to 52 tokens/s on eight nodes, and with the Spark-optimized RoCEnante collective communication backend, achieves a 4.43x decoding speedup on eight nodes. From-scratch pretraining of the nanochat model on eight nodes achieved a 7.79x speedup and 97.4% scaling efficiency. NVIDIA Sync's eight-node cluster support will launch before the end of the year.

NVIDIA launches Minerva AIOps platform, delivering end-to-end observability for AI factories

NVIDIA has launched Minerva, an AIOps platform for AI factories that collects telemetry from infrastructure and AI workloads, then uses purpose-built AI models to turn that data into actionable insights. The platform builds a dynamic knowledge graph that maps GPUs, SuperNICs, switches, cables, and the logic of the jobs running on them, analyzing billions of data points per hour to learn normal behavior, detect unseen anomalies, and predict failures. Minerva includes a built-in AI assistant and can connect to external agentic tools, supporting natural-language troubleshooting of cluster health with remediation recommendations. IREN, IBM, Jio, Lambda, Tesla, and others plan to deploy it, while Dell, Lenovo, and Supermicro will incorporate it into their AI factory operations portfolios.

NVIDIA Releases OpenShell Reference Architecture, Governing Agent Security with a Three-Layer Approach

NVIDIA has released the OpenShell reference architecture, arguing that Agents cannot govern themselves and that security and governance should be placed at the infrastructure layer, which Agents cannot reach. The approach has three layers: the open-source runtime OpenShell defines permission boundaries through Gateway, Supervisor, Sandbox, and a formally verified Policy Prover; Vera CPU provides 88 Olympus cores and 1.2 TB/s memory bandwidth, with CoreWeave measuring sandbox startup more than 3x faster and Agent tasks on Terminal Bench 1.8x faster; BlueField-4 DPU plus Sentry provides optional hardware-level monitoring that can isolate violating Agents within a few milliseconds.

Nvidia unveils truck-transportable Vera Rubin containerized AI factory DSX Pod

Nvidia has unveiled the DSX Pod AI Factory, which packs 7 Vera Rubin AI racks into a container roughly the size of a standard shipping container and weighing about 28,000 kilograms. Cabling and liquid cooling are assembled at the factory, and the unit can be transported by truck over ordinary roads. Nvidia says the solution can cut the time for a new AI data center to start outputting tokens from 6 months to 14 days. The product is a reference design, shipped by manufacturers including Foxconn, Gigabyte, Lenovo, HPE, Quanta, and Pegatron, with deliveries expected to begin in early 2027; no price has been announced.

Nvidia launches Open Agent Safety Platform, using BlueField-4 DPU to monitor agents

At GTC Berlin, Nvidia launched the Open Agent Safety Platform, adding hardware-level controls around the open-source runtime OpenShell; a BlueField-4 DPU works with Sentry software to independently monitor agent behavior and, when it crosses boundaries, isolate and stop it within milliseconds. OpenShell supports third-party agent frameworks such as Anthropic's Claude Code and OpenAI's Codex, and Nvidia says more than 100 organizations already use the platform's technology. Nvidia also introduced the Agent Knowledge Foundation and the Auto-Ontology library, designed to let agents access structured and unstructured data without enterprise data leaving local premises, and expanded its Nemotron 3.5 Super VL, Nemotron 3.5 Bolt 74B-A9B and Kumo series models.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

Nvidia unveils DSX POD containerized AI factory reference design

Nvidia announced the DSX POD reference design at GTC Berlin, enabling partners to install Vera Rubin racks, networking, and liquid cooling into containers at the factory before delivery. Each container holds 7 Vera Rubin NVL72 racks, totaling 504 Rubin GPUs, and comes pre-installed with about 20 kilometers of copper cable and 17 kilometers of scale-out networking, weighing about 28 tons; as long as the site already has power, cooling, and networking in place, it can start producing tokens in as little as 14 days. Nvidia says the system's cooling uses no water, Vera Rubin can run on 45 degrees Celsius coolant, and at a 2-gigawatt deployment, dry coolers can save about 54 million liters of water per day compared with cooling towers. Partners will begin global deliveries in early 2027.

Nvidia adds Agent Knowledge Foundation to Agent Toolkit

At GTC Berlin, Nvidia added Agent Knowledge Foundation to Agent Toolkit, packaging the Nemotron Retriever models, the new Auto-Ontology library, the Kumo family of models for structured data, and CUDA-X data processing libraries; data remains in place and is managed under enterprises' existing permissions. Nvidia says Nemotron Retriever improved search relevance by 65% in tests in the open-source OpenSearch environment, SAP's fine-tuned reranker runs nearly three times faster, and ServiceNow's reranking delivers better quality with latency reduced by nearly 25%. Auto-Ontology provides a REST API and an MCP server for agents to query, and Nvidia says it does not compete with data platforms such as Snowflake but rather partners with them.

NVIDIA Releases SimReady Central, an Open Library of Simulation-Ready 3D Assets, on Hugging Face

NVIDIA has launched SimReady Central on Hugging Face, an open 3D asset library for robot and AI factory simulation. The assets are in OpenUSD format with simulation attributes such as mass, collision geometry, and friction. OEMs can use the SimReady Central submission skill to validate and submit CAD equipment models according to the SimReady configuration, while developers and AI agents can use NVIDIA USD Search to retrieve assets in natural language and assemble them into Isaac Sim scenes. The article uses a pick-and-place scenario with a FANUC robotic arm equipped with a Robotiq gripper and an AI factory layout comparison as examples to demonstrate the workflow of generating scene assembly and validation code with GPT-6 Astra.

NVIDIA Confidential Computing Adds Confidential Bare Metal Deployment Option

NVIDIA is expanding its confidential computing capabilities with the launch of a Confidential Bare Metal deployment option, enabling frontier model builders to retain control over their models and data when running sensitive workloads on third-party dedicated infrastructure. This offering, together with the existing Confidential Containers and Confidential Virtual Machines, forms three deployment forms, and adds controls such as data protection in transit, platform integrity measurement, and hardware-level access restrictions. Cohere plans to integrate it into Cohere Model Vault, Google Cloud will support the capability on A4X Max instances based on GB300 NVL72, and providers such as CoreWeave, Microsoft Azure Infrastructure, and Nebius also support related deployments.

English translation is being prepared.

NVIDIA Releases OpenShell Reference Architecture to Provide Security and Governance for Large-Scale AI Agent Workloads

NVIDIA has released the OpenShell reference architecture, using an open-source runtime to treat agent workloads as untrusted objects and grant access on demand. The architecture consists of four components: Gateway, Supervisor, Sandbox, and Policy Prover, which are responsible for sandbox lifecycle and policy management, credential access brokering, workload isolation, and formal verification of policy permissions, respectively. CoreWeave's measurements show that on Vera CPUs, sandbox startup is more than 3 times faster than on x86, agent loops complete 1.8 times faster, and a single rack can support more than 11,000 concurrent agent environments.

English translation is being prepared.

English translation is being prepared.

NVIDIA Releases Cosmos3-H-Dreams Surgical Generative Simulation Model

NVIDIA has released Cosmos3-H-Dreams, specialized from the 16-billion-parameter Cosmos3-H-Embodied-Nano for action-conditioned video generation of cholecystectomy. The model is trained with 28 action dimensions, 832 x 480 video, and 10 Hz action alignment, and distillation yields two student checkpoints, four-step and two-step; the two-step version is trained separately, offering faster inference but lower image quality. NVIDIA also provides the FlashDreams streaming runtime and methods for adapting to custom data and controllers, and states that the model is a research simulator, not a validated clinical system.

DeepSeek open-sources DeepInfer, the inference engine behind its API, with support for both Nvidia and Huawei Ascend

DeepSeek has open-sourced DeepInfer, its in-house inference framework used to serve its models in production. Released under the Apache 2.0 license, it lets developers run DeepSeek-V4.1-Flash on their own infrastructure, with support for subsequent models planned. The framework includes optimized compute kernels and communication libraries such as FlashMLA, DeepGEMM, and DeepEP, supports speculative decoding, asynchronous scheduling, and structured output, and also open-sources the deepseek-recipe library (available in Rust and Python versions) for handling request and response transformations. On the hardware side, it supports Nvidia SM100 (Blackwell) architecture and the Huawei Ascend 950 NPU series from day one; older architectures such as Hopper are not listed. Batch-invariant inference and deterministic sampling are available only on Nvidia hardware, and the published performance data is also based on Nvidia GPUs. The current version supports offline inference but lacks complete production serving components such as KV cache management, cross-machine inference stage separation, and replica load balancing.