Topic: #nvidia

- Ryzen AI Halo is basically a Strix Halo PC built around AMD Ryzen AI Max+ 395, 16 Zen 5 cores, Radeon 8060S graphics, 128 GB unified memory and x86 compatibility. - NVIDIA DGX Spark uses the GB10 Grace Blackwell Superchip, 20 Arm cores, 128 GB coherent LPDDR5x memory, DGX OS and up to 1 PFLOP FP4 performance.

- NVIDIA frames Vera as a new CPU category for agentic AI: maximum single-thread speed at data-center scale, not just high core counts. - Vera uses the Olympus core, which NVIDIA says delivers 50% higher instructions per cycle than Grace, with 88 cores, up to 1.2 TB/s LPDDR5X bandwidth and 3.4 TB/s core-to-core bandwidth.

- NVIDIA and Hugging Face are bringing Isaac GR00T 1.7 and Isaac Teleop into LeRobot, Hugging Face's open source robotics library. Cosmos 3 is planned to follow later. - The integrations aim to standardize data collection, training, fine-tuning, evaluation and deployment for robotics models.

- NVIDIA says it had 74 papers accepted at ICML 2026. Around 2,000 accepted papers cite NVIDIA GPUs, and 145 cite NVIDIA Nemotron as a foundation for new research. - The post frames Nemotron as more than one model release: an open research stack with weights, datasets, and recipes for reasoning, tool use, safety, data curation, and efficient inference.

- NVIDIA is introducing a new partner model for AI clouds: providers can build large-scale, multi-tenant AI factories with NVIDIA infrastructure and sell capacity to startups, model builders, enterprises, research organizations and regional AI players. - The model combines standard product revenue with cloud revenue sharing and credit support.

- Amazon Bedrock now offers OpenAI GPT OSS and NVIDIA Nemotron in AWS GovCloud (US): gpt-oss-120b, gpt-oss-20b, plus Nemotron Nano 9B v2, Nano 12B v2, Nano 30B, and Super 120B. - The models run inside the GovCloud boundary. In-Region inference is available in us-gov-west-1, while Geo Cross-Region routing spans us-gov-west-1 and us-gov-east-1 without using commercial AWS Regions.

- MIT Technology Review frames Zurich as a dense R&D hub where Apple, Anthropic, Disney Research, Google, Meta, Microsoft, NVIDIA and OpenAI all maintain research teams or labs. - The point is concentration, not scale: just over 400,000 residents, proximity to ETH Zurich, and deep talent in AI, computer vision, graphics, robotics and systems engineering.

- Palantir is integrating NVIDIA Nemotron open models into a new AI engine aimed at US government agencies and critical infrastructure operators. - The models are meant to run in air-gapped environments on NVIDIA accelerated computing, keeping deployments separated from unsecured networks.

- OpenAI is working with Broadcom on Jalapeño, a custom inference chip, joining Google, Apple, and SpaceX in the push to reduce reliance on Nvidia. - The point is not a clean break from Nvidia. It is a hedge against single-supplier risk around pricing, supply, priority access, and product roadmaps.

- AWS explains how to tune Amazon SageMaker AI training jobs for NVIDIA Blackwell: batch size, sequence length, precision format and activation checkpointing are the main levers. - The examples use P6-B200 instances with 8 Blackwell GPUs and PyTorch FSDP, focused on transformer models from 1B to 64B parameters.

- NVIDIA and AWS are packaging several production AI infrastructure pieces: Amazon EC2 G7 with NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs, OpenSearch Serverless with NVIDIA cuVS, and validated GB300 training performance. - EC2 G7 is positioned as a step up from G6, with up to 4.6x AI inference performance, up to 2.1x graphics performance, and faster GPU analytics via cuDF on Amazon EMR.

- At DTW Ignite 2026, NVIDIA is presenting a telecom autonomy stack built from synthetic data, domain models, secure runtimes and simulation, aimed at moving operators beyond task automation into 24/7 agentic operations. - SoftBank is using NeMo Safe Synthesizer and NeMo Anonymizer to create privacy-preserving synthetic telecom datasets for fine-tuning large telecom models and specialized network agents.

- Groq confirmed a new $650 million funding round led by Disruptive and Infinitum. The company did not disclose a new valuation. - The raise follows Nvidia’s deal about six months ago, in which Nvidia licensed Groq technology and hired founder Jonathan Ross, president Sunny Madra, and other staff.

- The NSF NAIRR pilot has run for two years and, according to NVIDIA, now supports more than 700 U. research projects, from protein prediction to infectious-disease outbreak management. - NVIDIA contributes cloud-based DGX capacity: researchers get access to at least four DGX nodes for at least one month, plus technical support for onboarding and project work.

- NVIDIA is using Cannes Lions to frame advertising and marketing as a GPU infrastructure market. The partner list includes Alembic, AWS, Criteo, Higgsfield, KERV. - Alembic plans to use DGX Vera Rubin SuperPODs for Causal AI, aiming to separate true marketing impact from correlation across channels, markets and audiences.

- NVIDIA frames France as a growing European AI hub: one year after GTC Paris, AI factories, national compute capacity and industrial AI platforms are moving from announcements into deployment. - Mistral is already running 18,000 GB200 systems, according to NVIDIA, and is building a 44 MW data center in Bruyères-le-Châtel as part of a roadmap toward 200 MW of European compute by 2027.

Accelerated computing has transformed industrial engineering, cutting simulation times from weeks to hours. The remaining bottlenecks now sit around the simulation itself: CAD design, meshing, simulation setup and debugging, plus post-processing and report generation.

This post walks through how to build a multi-agent campaign review system step by step: NVIDIA NIM provides GPU-accelerated inference, Amazon Bedrock AgentCore brings managed runtime, shared memory and observability, and Strands Agents handle serverless multi-agent orchestration. The same architecture transfers to digital assistants, review automation, and RAG pipelines.

At NVIDIA GTC Taipei during COMPUTEX, developers, researchers, and industry leaders converge to cover the latest in AI factories, scaling infrastructure, agentic AI, and physical AI. NVIDIA traditionally uses this stage for major announcements — anyone building or buying into the AI stack should watch the livestream. It's not just hardware news; it's the roadmap for the next 12 months.

After open-source hardware accelerated robotics development, the software running robot brains is increasingly going open too. Hugging Face, Nvidia, and Alibaba have all made significant open-source robotics bets in the past two years, releasing foundation models for robot control. This lowers the cost of building capable robots for research labs and startups, with the industry hoping for an LLM-style breakthrough.

The first NVIDIA Vera CPUs arrived at three of the world's leading AI labs — Anthropic in San Francisco, OpenAI in Mission Bay, and SpaceXAI in Palo Alto — followed by a delivery to Oracle Cloud Infrastructure in Santa Clara. NVIDIA VP of Hyperscale and HPC Ian Buck hand-delivered them.

NVIDIA’s latest AI model Nemotron 3 Nano Omnia, featuring an impressive 30 billion parameters, is designed to excel in multimodal processing, handling images, video and audio with remarkable efficiency. Highlighted by Two Minute Papers, this system achieves exceptional throughput, processing nearly 10 hours of video per hour, a speed 10 times faster than real-time playback.

The rise of electricity-guzzling data centers has forced the AI industry to get creative about finding power. Nvidia is teaming up with InfraPartners, Prologis, and nonprofit EPRI to build about 25 micro data centers (5–20 MW each) next to utility substations at five US utilities.

Nvidia founder and CEO Jensen Huang told graduates at Carnegie Mellon University in Pittsburgh yesterday that demand for AI infrastructure is creating a once-in-a-generation opportunity to reindustrialize America and restore the nation's capacity to build. With many college grads fearing AI could obliterate their career dreams, Huang pointed to boundless opportunity as a new industry is being born and a new era of science and discovery begins.

The SpaceX–Anthropic deal marks a pivotal moment in the AI market and underscores the industry's brutal compute hunger. SpaceX is leasing Anthropic the Colossus 1 supercomputer — over 220,000 Nvidia GPUs and 300 MW of power — so the lab can keep scaling its models. Notable twist: Elon Musk's SpaceX is supplying compute to a direct rival of his own xAI.

The Pentagon has struck deals with OpenAI, Google, Microsoft, Amazon, Nvidia, Elon Musk's xAI, and the startup Reflection, allowing the agency to use their AI tools in classified settings. The Defense Department has left out Anthropic — which it previously used for classified information — after declaring it a supply-chain risk.

Samsung Electronics reported record quarterly profit driven by a 49-fold jump in chip income, warning that the global memory shortage will deepen into 2027. The AI datacenter buildout is pushing Samsung and rivals to allocate capacity to high-end chips used in Nvidia's AI accelerators. The squeeze hits conventional chips hard — pricing pressure is mounting across SSDs and DRAM.

Altman trial, exhibits are surfacing piece by piece — emails, photos, and corporate documents from OpenAI's earliest days, some predating even the lab's name. Highlights so far: Nvidia CEO Jensen Huang gifted OpenAI an in-demand supercomputer, Musk largely drafted OpenAI's mission and shaped its early structure, Sam Altman apparently wanted to lean heavily on Y Combinator for early support, and Greg Brockman and Ilya Sutskever were already worried about…

NVIDIA is launching a 12 GB version of the RTX 5070 GPU for laptops, alongside the existing 8 GB model. The new SKU uses 24 Gb G7 modules to ease pressure on the constrained 16 Gb G7 supply currently shipping with most GPUs. First laptops with the variant arrive in June from ASUS, Lenovo and MSI.

IT budgets are blowing up as some companies spend more on AI than on employee salaries. Nvidia's Bryan Catanzaro says compute costs already exceed the people costs of his team, and Uber's CTO has already burned through his full 2026 AI budget. Gartner projects worldwide IT spending of $6.31 trillion in 2026, up 13.5% from 2025, driven by sustained momentum across AI infrastructure, software and cloud services.

Google Cloud has launched two new AI accelerator chips designed to compete with Nvidia's dominance in the AI infrastructure market. The new TPUs are faster and cheaper than previous generations, offering cloud customers a Google-native alternative for AI workloads. Despite the push, Google Cloud still supports Nvidia hardware—signaling a pragmatic dual-track strategy for the near term.

Nvidia CEO Jensen Huang suggested that AI will function as a constant, ever-present manager that monitors and nudges workers continuously. The vision positions AI less as a helpful assistant and more as an inescapable digital overseer. Huang's comments spark debate about how much AI-driven supervision is actually desirable in the workplace—and where the line between productivity tool and surveillance lies.

Amazon announced the availability of G7e instances powered by NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs on Amazon SageMaker AI. You can provision nodes with 1, 2, 4, and 8 GPU instances, with each GPU providing 96 GB of GDDR7 memory. This enables cost-effective hosting of large foundation models like GPT-OSS-120B and Qwen3.5-35B-A3B on a single node.

The Intel Arc Pro B70 is a professional-grade GPU designed to meet the needs of AI professionals and computational workloads, offering a balance between affordability and performance. With 32 GB of VRAM and a price under $1,000, it provides a cost-effective alternative to higher-priced competitors like Nvidia’s RTX Pro 4000. However, as Alex Ziskind highlights, […] The post Why the Intel Arc Pro B70 Might Be the Ultimate Budget GPU for Local AI appeared…

- Iran's Islamic Revolutionary Guard Corps (IRGC) published a video on April 3rd directly threatening OpenAI's planned data center in Abu Dhabi. - The video appeared on an Iranian state-backed outlet's X account and vows the 'complete and utter annihilation' of US-linked energy and tech companies in the region. - It shows footage of OpenAI's $30 billion Stargate facility in the UAE, currently under construction.

- OpenAI has closed a $122 billion funding round, achieving a valuation of $852 billion. - Amazon, Nvidia, and SoftBank are among the lead investors – SoftBank alone reportedly committed $110 billion, according to the Wall Street Journal. - A select group of individual investors was also allowed to contribute around $3 billion.

- NVIDIA and Emerald AI announced a collaboration at CERAWeek to treat AI data centers as dynamic, grid-responsive assets rather than fixed power drains. - The approach lets AI factories ramp consumption up or down in real time based on grid conditions – absorbing surplus or shedding load during stress events.

The rapid expansion of AI data centers is fueling global conflicts over their impact on power grids, energy costs, and local communities. From Senate hearings demanding electricity usage transparency to legal battles over pollution, the stakes are growing.

- Mark Zuckerberg (Meta), Larry Ellison (Oracle), Jensen Huang (Nvidia), and Sergey Brin (Google) will be the first four members of Trump's revived PCAST advisory panel. - The council will 'weigh in on AI policy' and launches with 13 members, expandable to 24. - AI and crypto czar David Sacks and White House tech advisor Michael Kratsios will co-chair the panel.

- Arm is launching its first ever self-produced chip, the Arm AGI CPU, purpose-built for AI inference workloads in cloud data centers. - Meta is both the lead partner and co-developer, and is first in line to deploy the chip — with plans to collaborate on 'multiple generations' of data center CPUs.

- NVIDIA donates a Dynamic Resource Allocation (DRA) driver for GPUs to the Kubernetes community as an open-source contribution for better GPU management in AI workloads. - DRA allows GPU resources in Kubernetes to be allocated more flexibly and granularly, moving away from an all-or-nothing model. - The driver is intended to help developers run high-performance AI infrastructure more transparently and efficiently.

- NVIDIA introduces OpenShell, a framework designed to make autonomous AI agents 'Secure by Design' – baking security in from the start rather than patching it on later. - Modern agents can read files, write and execute code, use tools, and orchestrate workflows across enterprise systems. - Application-layer risk scales exponentially once agents can expand their own capabilities autonomously.

- NVIDIA RTX PRO 6000 Blackwell Workstation Edition is NVIDIA's new high-end GPU targeting enterprise data science workstations. - The article is sponsored content by PNY Technologies, an NVIDIA board partner that manufactures and sells GPU cards. - Core problems cited: CPU bottlenecks in data prep, exploding dataset sizes, and forced downsampling as a workaround.

- Nvidia's AI beautification feature, designed to enhance in-game character faces, produced a grotesque oversized nostril on a Starfield character – dubbed the 'Giga-Nostril'. - The tool is meant to 'yassify' – i. idealize – NPC faces using AI upscaling or facial enhancement.

- NVIDIA released NemoClaw, an open-source framework designed to secure autonomous AI agents through declarative security policies and real-time monitoring. - It builds on its predecessor OpenClaw with added sandboxing, stricter access controls, and operational safety features for multi-agent workflows.

- Three individuals have been charged by the US Attorney's Office for the Southern District of New York with illegally exporting NVIDIA GPUs to China, violating the Export Control Reform Act. - The accused are Yih-Shyan 'Wally' Liaw, Ruei-Tsang 'Steven' Chang, and Ting-Wei 'Willy' Sun – two employees and one contractor at US IT firm Super Micro Computer.

- NVIDIA announced DLSS 5 at its GTC conference, triggering immediate backlash across gaming communities online. - Unlike DLSS 3 and 4, which focused on AI upscaling and frame generation, DLSS 5 uses 'neural processing' to deliver photorealistic lighting and materials. - Analyst Anshel Sag from Moor Insights & Strategy joins the Engadget Podcast to break down his hands-on experience with NVIDIA's DLSS 5 demos.

- NVIDIA extends the OpenClaw framework with NemoClaw – an enterprise layer introducing privacy controls and security guardrails for autonomous AI agents. - NemoClaw targets organizations deploying AI agents at scale while meeting compliance and data protection requirements. - The new security features are designed to ensure data integrity and operational reliability in production agent deployments.

- Nvidia unveiled DLSS 5, a '3D guided neural rendering model' that alters a game's lighting and materials in real time using AI. - The community response was overwhelmingly negative: memes flooded social media, with players accusing Nvidia of 'yassifying' Resident Evil Requiem characters in demo footage. - Jensen Huang dismissed the backlash bluntly: 'They're completely wrong.

- Nvidia unveiled a new AI feature that enhances video game visuals in real time, marketed internally as a major breakthrough. - CEO Jensen Huang called it the 'GPT moment' for graphics, drawing a direct parallel to the rise of large language models. - The community responded with mockery, coining the term 'Sloptracing' as a counter-narrative to Nvidia's own hype.

- NVIDIA announced DLSS 5, an AI feature that adds 'photorealistic' lighting and materials to in-game models and environments. - Social media and Reddit reactions were overwhelmingly negative, with virtually no genuine enthusiasm found. - NVIDIA markets DLSS 5 as the 'biggest breakthrough in computer graphics' since RTX ray tracing launched in 2018.

- Chinese AI labs are cut off from NVIDIA's latest hardware, including Blackwell chips, Groq LPUs, and Rubin NVL72 modules. - US export controls block shipment of these systems to China, significantly widening the training compute gap. - While US labs like OpenAI and Anthropic deploy tens of thousands of high-end GPUs, Chinese players are left with older hardware or domestic alternatives.

- At GTC 2026, NVIDIA is pushing local AI hardware to the forefront: RTX PCs and the DGX Spark desktop supercomputer are being positioned as 'agent computers' — a new device category. - The DGX Spark is a compact desktop AI supercomputer capable of running powerful open-source models fully locally, no cloud required.

- Nvidia CEO Jensen Huang expects at least $1 trillion in revenue from its newest chips through 2027, backed by record sales and surging orders from Big Tech data center operators. - Nvidia's cumulative AI chip market share dropped from 100% in Q1 2022 to 65% in Q4 2024, per SemiAnalysis – but the company still dominates decisively.

- At GTC in San Jose (30,000+ attendees), Nvidia CEO Jensen Huang unveiled the Vera Rubin chip line — Nvidia's first chip designed specifically for AI inference. - The Nvidia Groq 3 LPU (language processing unit) incorporates IP licensed from startup Groq for US $20 billion, a deal struck on Christmas Eve 2024.

- NVIDIA unveiled DSX Air at GTC 2026 in San Jose, introduced by CEO Jensen Huang as a core part of the DSX platform. - DSX Air simulates AI factory setups before physical deployment, cutting rollout time from months to days. - It sits within NVIDIA DSX Sim and targets faster planning, testing, and optimization of AI infrastructure.

- Nvidia introduces DLSS 5, using generative AI and structured graphics data to push video game visuals closer to photorealism. - Unlike previous DLSS versions focused mainly on upscaling, DLSS 5 actively generates new image content based on scene geometry and motion data. - CEO Jensen Huang suggests the technology could eventually expand beyond gaming into other industries.

- Jay built Zeus: a custom supercomputer for $8,500 featuring an AMD Ryzen 9 CPU, 128 GB RAM, and an Nvidia RTX 5090 GPU. - Zeus replaces cloud services for tasks like data scraping, email verification, and AI model training — all running locally with no recurring subscription fees. - The system runs Unraid OS and uses a modular design, making it easy to swap or upgrade individual components over time.

- ByteDance is partnering with a firm called Aolani Cloud to build Blackwell computing systems in Malaysia, sidestepping US export restrictions. - The plan involves acquiring roughly 36,000 NVIDIA B200 chips — NVIDIA's most powerful AI processor currently available. - The hardware buildout will reportedly cost more than $2.5 billion, according to the Wall Street Journal.

- Nvidia is holding its annual GTC (GPU Technology Conference) – the chipmaker's flagship event for product announcements and partnerships. - CEO Jensen Huang will deliver a keynote focused on Nvidia's vision for the future of computing and AI. - GTC is considered essential viewing for anyone tracking the direction of AI hardware and data centers.

- AI usage is cheaper today than it has ever been – but that window may be closing. - Writer CEO May Habib told Axios that LLM companies will be forced to raise prices around their IPOs. - New models from OpenAI, Google, and Anthropic are faster and cheaper, driven by massive efficiency gains in inference.

- NVIDIA launched Nemotron 3 Super, an open model with 120 billion total parameters but only 12 billion active ones, using a mixture-of-experts architecture. - NVIDIA claims 5x higher throughput compared to dense models of similar scale, specifically targeting agentic AI workloads. - Perplexity is among the first AI-native companies to offer users direct access to the model.

- Rabbit is developing 'Project Cyberdeck', a compact PC designed for vibe coding, inspired by late-2000s netbooks. - CEO Jesse Lyu was motivated by watching his engineers heavily use Claude Code – then searching online for a fitting device and finding nothing satisfactory. - The goal is not a high-end AI workstation like NVIDIA's $3,999 DGX Spark, but a lean CLI-focused machine for on-the-go use.

A Guardian investigation has put the UK government’s AI plans under the microscope. Here are the key details Revealed: UK’s multibillion AI drive is built on ‘phantom investments’ The Essex ‘supercomputer’ that’s still a scaffolding yard A Guardian investigation has examined a series of massive AI investments announced by the government over the past two years, comparing what was promised with what has so far been delivered.

London-based startup, which is vital to the government’s artificial intelligence ambitions, is now valued at $14.6bn UK’s multibillion AI drive is built on ‘phantom investments’ The Essex supercomputer that’s still a scaffolding yard Business live – latest updates Nscale, a UK company vital to the government’s AI ambitions, has raised $2bn (£1.5bn) in a funding round and appointed the former Meta executives Sheryl Sandberg and Nick Clegg to its board of…

- The Trump administration is weighing rules that would require foreign buyers to obtain U. government licenses before purchasing American AI chips. - Nvidia and AMD – the dominant players in AI training hardware – would be directly affected.

• OpenAI is reportedly finalizing a $100 billion funding round with closing expected imminently. • Participating investors include Amazon, Nvidia, SoftBank, and Microsoft. • The deal would value the ChatGPT-maker at over $850 billion, among the highest private company valuations on record.

Nvidia delays its RTX 50 Super refresh indefinitely – the cards were expected at CES 2026 in January, but managers decided against it in December. Reason: Nvidia prioritizes AI chips due to limited RAM supply and is cutting production of the current RTX 50-series, which is already sold out everywhere.

The $100bn deal between Nvidia and OpenAI, announced in September 2024, appears to have collapsed. The arrangement was circular: Nvidia would fund OpenAI heavily, with most funds flowing back to purchase Nvidia chips. The apparent failure raises questions about who will bear AI infrastructure costs and whether the current funding model is sustainable.