Physical AI is becoming the foundation of smart cities, facilities and industrial processes across the globe.

NVIDIA is working with companies including Accenture, Avathon, Belden, DeepHow, Milestone Systems and Telit Cinterion to enhance operations across the globe with physical AI-based perception and reasoning.

The continuous loop of simulating, training and deploying physical AI offers sophisticated industrial automation capabilities, making cities and infrastructure safer, smarter and more efficient.

For example, physical AI applications can automate potentially dangerous tasks for workers, such as working with heavy machinery. Physical AI can also improve transportation services and public safety, detect defective products in factories and more.

The need for this is greater than ever. The numbers tell the story:

Infrastructure that can perceive, reason and act relies on video sensors and the latest vision AI capabilities. Using the NVIDIA Metropolis platform — which simplifies the development, deployment and scaling of video analytics AI agents and services from the edge to the cloud — developers can build visual perception into their facilities faster to enhance productivity and improve safety across environments.

Below are five leading companies advancing physical AI — and five key NVIDIA Metropolis updates, announced today at the SIGGRAPH computer graphics conference, making such advancements possible.

Five Companies Advancing Physical AI

Global professional services company Accenture is collaborating with Belden, a leading provider of complete connection solutions, to enhance worker safety by creating smart virtual fences that factories can place around large robots to prevent accidents with human operators.

The smart virtual fence is a physical AI safety system that uses an OpenUSD-based digital twin and physics-grounded simulation to model complex industrial environments. Using computer vision-based mapping and 3D spatial intelligence, the system is adaptive to increased variability in the dynamic human-robot interactions that occur in a modern shopfloor environment.

Accenture taps into the NVIDIA Omniverse platform and Metropolis to build and simulate these smart fences. With Omniverse, Accenture created a digital twin of a robot arm and workers moving in a space. And with Metropolis, the company trained its AI models and deployed them at the edge with video ingestion and the NVIDIA DeepStream software development kit (SDK)’s real-time inference capabilities.

Avathon, an industrial automation platform provider, uses the NVIDIA Blueprint for video search and summarization (VSS), part of NVIDIA Metropolis, to provide manufacturing and energy facilities with real-time insights that improve operational efficiency and worker safety.

Reliance British Petroleum Mobility Limited, a leader in India’s fuel and mobility sector, used the Avathon video intelligence product during the construction of its gas stations to achieve higher standards of safety compliance, a reduction in safety noncompliance incidents and higher productivity by saving thousands of work hours.

DeepHow has developed a “Smart Know-How Companion” for employees in manufacturing and other industries. The companion uses the Metropolis VSS blueprint to transform key workflows into bite-sized, multilingual videos and digital instructions, improving onboarding, safety and floor operator efficiency.

Facing upskilling needs and retiring skilled workers, beverage company Anheuser-Busch InBev turned to the DeepHow platform to convert standard operating procedures into easy-to-understand visual guides. This has slashed onboarding time by 80%, boosted training consistency and improved long-term knowledge retention for employees.

Milestone Systems, which offers one of the world’s largest platforms for managing IP video sensor data in complex industrial and city deployments, is creating the world’s largest real-world computer vision data library through its platform, Project Hafnia. Among its capabilities, the platform provides physical AI developers with access to customized vision language models (VLMs).

Tapping NVIDIA NeMo Curator, Milestone Systems built a VLM fine-tuned for intelligent transportation systems for use within the VSS blueprint to help develop AI agents that better manage city roadways. Milestone Systems is also looking to use the new open, customizable NVIDIA Cosmos Reason VLM for physical AI.

Internet-of-things company Telit Cinterion has integrated NVIDIA TAO Toolkit 6 into its AI-powered visual inspection platform, which uses vision foundation models like FoundationPose, alongside other NVIDIA models, to support multimodal AI and deliver high-performance inferencing. TAO brings low-code AI capabilities to the Telit platform, enabling manufacturers to quickly develop and deploy accurate, custom AI models for defect detection and quality control.

Five NVIDIA Metropolis Updates for Physical AI

Key updates to NVIDIA Metropolis are enhancing developers’ capabilities to build physical AI applications more quickly and easily:

Cosmos Reason VLM

The latest version of Cosmos Reason — NVIDIA’s advanced open, customizable, 7-billion-parameter reasoning VLM for physical AI — enables contextual video understanding, temporal event reasoning for Metropolis use cases. Its compact size makes it easy to deploy from edge to cloud and ideal for automating traffic monitoring, public safety, visual inspection and intelligent decision-making. Cosmos Reason has topped the Physical Reasoning Leaderboard on Hugging Face.

VSS Blueprint 2.4

VSS 2.4 makes it easy to quickly augment existing vision AI applications with Cosmos Reason and deliver powerful new features to smart infrastructure. An expanded set of application programming interfaces in the blueprint offers users direct more flexibility in choosing specific VSS components and capabilities to augment computer vision pipelines with generative AI.

New Vision Foundation Models

The NVIDIA TAO Toolkit includes a new suite of vision foundation models, along with advanced fine-tuning methods, self-supervised learning and knowledge distillation capabilities, to optimize deployment of physical AI solutions across edge and cloud environments. The NVIDIA DeepStream SDK includes a new Inference Builder to enable seamless deployment of TAO 6 models.

Companies around the world — including Advex AI, Instrumental AI and Spingence — are experimenting with these new models and NVIDIA TAO to build intelligent solutions that optimize industrial operations and drive efficiency.

NVIDIA Isaac Sim Extensions

New extensions in the NVIDIA Isaac Sim reference application help solve common challenges in vision AI development — such as limited labeled data and rare edge-case scenarios. These tools simulate human and robot interactions, generate rich object-detection datasets, and create incident-based scenes and image-caption pairs to train VLMs, accelerating development and improving AI performance in real-world conditions.

Expanded Hardware Support

All of these Metropolis components can now run on NVIDIA RTX PRO 6000 Blackwell GPUs, the NVIDIA DGX Spark desktop supercomputer and the NVIDIA Jetson Thor platform for physical AI and humanoid robotics — so users can develop and deploy from the edge to the cloud.

Cosmos Reason 1 and NVIDIA TAO 6.0 are now available for download. Sign up to be alerted when VSS 2.4, the Cosmos Reason VLM fine-tuning update and NVIDIA DeepStream 8.0 become available.

Watch the NVIDIA Research special address at SIGGRAPH and learn more about how graphics and simulation innovations come together to drive industrial digitalization by joining NVIDIA at the conference, running through Thursday, Aug. 14.

See notice regarding software product information.

As one of the world’s largest emerging markets, Indonesia is making strides toward its “Golden 2045 Vision” — an initiative tapping digital technologies and bringing together government, enterprises, startups and higher education to enhance productivity, efficiency and innovation across industries.

Building out the nation’s AI infrastructure is a crucial part of this plan.

That’s why Indonesian telecommunications leader Indosat Ooredoo Hutchison, aka Indosat or IOH, has partnered with Cisco and NVIDIA to support the establishment of Indonesia’s AI Center of Excellence (CoE). Led by the Ministry of Communications and Digital Affairs, called Komdigi, the CoE aims to advance secure technologies, cultivate local talent and foster innovation through collaboration with startups.

Indosat Ooredoo Hutchison President Director and CEO Vikram Sinha, Cisco Chair and CEO Chuck Robbins and NVIDIA Senior Vice President of Telecom Ronnie Vasishta today detailed the purpose and potential of the CoE during a fireside chat at Indonesia AI Day, a conference focused on how artificial intelligence can fuel the nation’s digital independence and economic growth.

As part of the CoE, a new NVIDIA AI Technology Center will offer research support, NVIDIA Inception program benefits for eligible startups, and NVIDIA Deep Learning Institute training and certification to upskill local talent.

“With the support of global partners, we’re accelerating Indonesia’s path to economic growth by ensuring Indonesians are not just users of AI, but creators and innovators,” Sinha added.

“The AI era demands fundamental architectural shifts and a workforce with digital skills to thrive,” Robbins said. “Together with Indosat, NVIDIA and Komdigi, Cisco will securely power the AI Center of Excellence — enabling innovation and skills development, and accelerating Indonesia’s growth.”

“Democratizing AI is more important than ever,” Vasishta added. “Through the new NVIDIA AI Technology Center, we’re helping Indonesia build a sustainable AI ecosystem that can serve as a model for nations looking to harness AI for innovation and economic growth.”

Making AI More Accessible

The Indonesia AI CoE will comprise an AI factory that features full-stack NVIDIA AI infrastructure — including NVIDIA Blackwell GPUs, NVIDIA Cloud Partner reference architectures and NVIDIA AI Enterprise software — as well as an intelligent security system powered by Cisco.

Called the Sovereign Security Operations Center Cloud Platform, the Cisco-powered system combines AI-based threat detection, localized data control and managed security services for the AI factory.

Building on the sovereign AI initiatives Indonesia’s technology leaders announced with NVIDIA last year, the CoE will bolster the nation’s AI strategy through four core pillars:

Some 28 independent software vendors and startups are already using IOH’s NVIDIA-powered AI infrastructure to develop cutting-edge technologies that can speed and ease workflows across higher education and research, food security, bureaucratic reform, smart cities and mobility, and healthcare.

With Indosat’s coverage across the archipelago, the company can reach hundreds of millions of Bahasa Indonesian speakers with its large language model (LLM)-powered applications.

For example, using Indosat’s Sahabat-AI collection of Bahasa Indonesian LLMs, the Indonesia government and Hippocratic AI are collaborating to develop an AI agent system that provides preventative outreach capabilities, such as helping women subscribers over the age of 50 schedule a mammogram. This can help prevent or combat breast cancer and other health complications across the population.

Separately, Sahabat-AI also enables Indosat’s AI chatbot to answer queries in the Indonesian language for various citizen and resident services. A person could ask about processes for updating their national identification card, as well as about tax rates, payment procedures, deductions and more.

In addition, a government-led forum is developing trustworthy AI frameworks tailored to Indonesian values for the safe, responsible development of artificial intelligence and related policies.

Looking forward, Indosat and NVIDIA plan to deploy AI-RAN technologies that can reach even broader audiences using AI over wireless networks.

Learn more about NVIDIA-powered AI infrastructure for telcos.

In many parts of the world, including major technology hubs in the U.S., there’s a yearslong wait for AI factories to come online, pending the buildout of new energy infrastructure to power them.

Emerald AI, a startup based in Washington, D.C., is developing an AI solution that could enable the next generation of data centers to come online sooner by tapping existing energy resources in a more flexible and strategic way.

“Traditionally, the power grid has treated data centers as inflexible — energy system operators assume that a 500-megawatt AI factory will always require access to that full amount of power,” said Varun Sivaram, founder and CEO of Emerald AI. “But in moments of need, when demands on the grid peak and supply is short, the workloads that drive AI factory energy use can now be flexible.”

That flexibility is enabled by the startup’s Emerald Conductor platform, an AI-powered system that acts as a smart mediator between the grid and a data center. In a recent field test in Phoenix, Arizona, the company and its partners demonstrated that its software can reduce the power consumption of AI workloads running on a cluster of 256 NVIDIA GPUs by 25% over three hours during a grid stress event while preserving compute service quality.

Emerald AI achieved this by orchestrating the host of different workloads that AI factories run. Some jobs can be paused or slowed, like the training or fine-tuning of a large language model for academic research. Others, like inference queries for an AI service used by thousands or even millions of people, can’t be rescheduled, but could be redirected to another data center where the local power grid is less stressed.

Emerald Conductor coordinates these AI workloads across a network of data centers to meet power grid demands, ensuring full performance of time-sensitive workloads while dynamically reducing the throughput of flexible workloads within acceptable limits.

Beyond helping AI factories come online using existing power systems, this ability to modulate power usage could help cities avoid rolling blackouts, protect communities from rising utility rates and make it easier for the grid to integrate clean energy.

“Renewable energy, which is intermittent and variable, is easier to add to a grid if that grid has lots of shock absorbers that can shift with changes in power supply,” said Ayse Coskun, Emerald AI’s chief scientist and a professor at Boston University. “Data centers can become some of those shock absorbers.”

A member of the NVIDIA Inception program for startups and an NVentures portfolio company, Emerald AI today announced more than $24 million in seed funding. Its Phoenix demonstration, part of EPRI’s DCFlex data center flexibility initiative, was executed in collaboration with NVIDIA, Oracle Cloud Infrastructure (OCI) and the regional power utility Salt River Project (SRP).

“The Phoenix technology trial validates the vast potential of an essential element in data center flexibility,” said Anuja Ratnayake, who leads EPRI’s DCFlex Consortium.

EPRI is also leading the Open Power AI Consortium, a group of energy companies, researchers and technology companies — including NVIDIA — working on AI applications for the energy sector.

Using the Grid to Its Full Potential

Electric grid capacity is typically underused except during peak events like hot summer days or cold winter storms, when there’s a high power demand for cooling and heating. That means, in many cases, there’s room on the existing grid for new data centers, as long as they can temporarily dial down energy usage during periods of peak demand.

A recent Duke University study estimates that if new AI data centers could flex their electricity consumption by just 25% for two hours at a time, less than 200 hours a year, they could unlock 100 gigawatts of new capacity to connect data centers — equivalent to over $2 trillion in data center investment.

Putting AI Factory Flexibility to the Test

Emerald AI’s recent trial was conducted in the Oracle Cloud Phoenix Region on NVIDIA GPUs spread across a multi-rack cluster managed through Databricks MosaicML.

“Rapid delivery of high-performance compute to AI customers is critical but is constrained by grid power availability,” said Pradeep Vincent, chief technical architect and senior vice president of Oracle Cloud Infrastructure, which supplied cluster power telemetry for the trial. “Compute infrastructure that is responsive to real-time grid conditions while meeting the performance demands unlocks a new model for scaling AI — faster, greener and more grid-aware.”

Jonathan Frankle, chief AI scientist at Databricks, guided the trial’s selection of AI workloads and their flexibility thresholds.

“There’s a certain level of latent flexibility in how AI workloads are typically run,” Frankle said. “Often, a small percentage of jobs are truly non-preemptible, whereas many jobs such as training, batch inference or fine-tuning have different priority levels depending on the user.”

Because Arizona is among the top states for data center growth, SRP set challenging flexibility targets for the AI compute cluster — a 25% power consumption reduction compared with baseline load — in an effort to demonstrate how new data centers can provide meaningful relief to Phoenix’s power grid constraints.

“This test was an opportunity to completely reimagine AI data centers as helpful resources to help us operate the power grid more effectively and reliably,” said David Rousseau, president of SRP.

On May 3, a hot day in Phoenix with high air-conditioning demand, SRP’s system experienced peak demand at 6 p.m. During the test, the data center cluster reduced consumption gradually with a 15-minute ramp down, maintained the 25% power reduction over three hours, then ramped back up without exceeding its original baseline consumption.

AI factory users can label their workloads to guide Emerald’s software on which jobs can be slowed, paused or rescheduled — or, Emerald’s AI agents can make these predictions automatically.

Orchestration decisions were guided by the Emerald Simulator, which accurately models system behavior to optimize trade-offs between energy usage and AI performance. Historical grid demand from data provider Amperon confirmed that the AI cluster performed correctly during the grid’s peak period.

Forging an Energy-Resilient Future

The International Energy Agency projects that electricity demand from data centers globally could more than double by 2030. In light of the anticipated demand on the grid, the state of Texas passed a law that requires data centers to ramp down consumption or disconnect from the grid at utilities’ requests during load shed events.

“In such situations, if data centers are able to dynamically reduce their energy consumption, they might be able to avoid getting kicked off the power supply entirely,” Sivaram said.

Looking ahead, Emerald AI is expanding its technology trials in Arizona and beyond — and it plans to continue working with NVIDIA to test its technology on AI factories.

“We can make data centers controllable while assuring acceptable AI performance,” Sivaram said. “AI factories can flex when the grid is tight — and sprint when users need them to.”

Learn more about NVIDIA Inception and explore AI platforms designed for power and utilities.

At NVIDIA GTC Paris, Europe-based healthcare and life sciences companies are showcasing themselves as leaders in global healthcare innovation, demonstrating their ability to deliver transformative, AI-driven impact at a time when the continent needs it most.

Aimed at addressing staffing shortages, aging populations and rising costs, their diverse efforts across biopharma, population health and digital medicine are supported by the NVIDIA Inception program for startups, which provides cutting-edge startups with benefits including access to the latest developer tools, technical training and exposure to venture capital firms.

Whether building the world’s largest biodiversity database or deploying intelligent agents and AI factories that accelerate discovery and patient care, here’s how European companies are tapping the NVIDIA BioNeMo platform, NVIDIA DGX Cloud and the NVIDIA Cloud Partner network to drive better health outcomes for the region and beyond.

Basecamp Research’s AI-Ready Genomics Database Breaks the Data Wall

London-based Inception startup Basecamp Research has unveiled BaseData, the world’s largest and most diverse biological dataset for generative AI in life sciences.

Built from samples collected at over 125 locations in 26 countries, BaseData contains more than 9.8 billion new biological sequences and over a million previously unknown species — making it 30x faster, and growing up to 1000x faster — than UniRef 50, a public database that’s been used to train more than 80% of all biological sequence models.

This resource is now being used to train next-generation foundation models using NVIDIA BioNeMo Framework on the NVIDIA DGX Cloud Lepton platform.

By collaborating with NVIDIA, Basecamp has overcome the bottlenecks of scale, diversity and data governance that have traditionally held back commercial biopharma research. Its approach — combining a new data supply chain, global partnerships and GPU-accelerated workflows — enables the retraining of new classes of biological foundation models with the goal of unlocking generalizable biological design and accelerating drug discovery.

With this milestone, Basecamp is setting a new industry benchmark for data-driven AI in biosciences and laying the groundwork for generative biology breakthroughs.

“Data-optimal scaling is the key to overcoming the real-world limitations of current biological models,” said Phil Lorenz, chief technology officer at Basecamp Research. “Combined with NVIDIA’s compute and AI stack, we’re training models that can understand and generate biology like never before.”

Intelligent Agents Transform Healthcare Delivery in UK

Guy’s and St. Thomas’ NHS Foundation Trust — the largest NHS Trust in the U.K., with over 2.8 million patient contacts a year — is launching the Proactive and Accessible Transformation of Healthcare initiative, aka PATH, in collaboration with global investment company General Catalyst and Inception startups Hippocratic AI and Sword Health.

PATH seeks to transform care delivery by integrating advanced AI agents, helping to reduce specialty care waitlists, improve pain management and streamline triage.

“PATH aims to deliver better, faster and fairer healthcare for all,” said Ian Abbs, CEO of Guy’s and St. Thomas’. “By combining cutting-edge technology, including AI, with clinical care, we can build a more proactive NHS.”

This initiative brings together Hippocratic AI’s conversational agents that automate tasks like patient outreach, history-taking and referral validation with Sword Health’s AI Care platform, which has treated over 500,000 patients globally across physical pain, pelvic health and other clinical areas, delivering 6.5 million AI sessions to date.

“Our safety-focused generative AI agents can enable healthcare abundance in the U.K.,” said Munjal Shah, founder and CEO of Hippocratic AI. “With more personalized care, patients can feel more supported and heard, improving outcomes.”

“Our AI Care platform transforms the way care is delivered, turning waiting lists into recovery journeys,” said Virgílio Bento, founder and CEO of Sword Health. “Together, we can reduce waste in healthcare and materially improve clinical outcomes.”

PATH will explore solutions to address the U.K.’s elective care crisis, with more than 53,000 patients waiting for a first appointment at Guy’s and St. Thomas’ alone. By prioritizing cases based on clinical need and enabling timely intervention, PATH aims to design a blueprint for national-scale, AI-driven healthcare transformation.

“The goal of PATH is to enable the NHS to work better for everyone,” said Chris Bischoff, managing director at General Catalyst. “By deploying applied AI to increase access, improve care, optimize resources and empower staff, we believe we can build an NHS fit for the future.”

Pangaea Data’s AI Platform Closes Care Gaps Across Hard-to-Diagnose Diseases

Pangaea Data, an Inception company based in London and San Francisco, is helping close care gaps by discovering patients who are untreated and under-treated despite available intelligence in their existing medical records.

Pangaea’s platform is powered by the NVIDIA NeMo Agent toolkit, an open-source library for profiling and optimizing connected teams of AI agents. Pangaea is also adopting NVIDIA NIM microservices to harness large language models  that can help identify patients with rare and prevalent hard-to-diagnose diseases.

These tools help calculate clinical scores required to discover patients with such diseases — involving non-specific features such as fever, nausea and headache — in a manner which emulates a clinician’s manual review process for higher accuracy.

“The NeMo Agent toolkit and NVIDIA NIM microservices enable us to reduce the time taken to configure our platform for each disease condition from weeks to a single day, helping accelerate our mission of improving patient outcomes globally,” said Vibhor Gupta, founder and CEO of Pangaea Data.

Pangaea helps global pharmaceutical providers, health systems and policymakers transform care for better outcomes and increased health equity.

Its platform is now being deployed across health systems in the U.S., U.K., Spain and Barbados as a key part of a population health initiative led by the Prime Minister of Barbados.

Sofinnova, Cure51 and Next-Generation Startups Tap NVIDIA DGX Cloud

Through NVIDIA DGX Cloud Lepton, top European healthcare and life sciences venture capital firms like Sofinnova are bolstering AI-native startups.

As part of this initiative, selected VCs can offer their top portfolio companies early access to DGX Cloud Lepton — including access to 25 NVIDIA H100 GPU nodes for three months, plus white-glove support — to help them scale into new markets requiring sovereign, localized compute.

DGX Cloud is already used by startups like Paris-based Cure51, U.K.-based Sensible Biotechnologies and U.K.-based Molecular Glue Labs (MGL) through an Inception program benefit.

Cure51, a Sofinnova portfolio company, achieved a 17x speedup in genomic analysis and 2x cost savings by shifting workloads to NVIDIA Parabricks running on DGX Cloud.

Using NVIDIA’s sovereign AI infrastructure through the NVIDIA Cloud Partner network, Sensible reduced its optimization cycles for cell-based mRNA therapeutics design from 15 days to just one, while MGL validated new protein engineering approaches. This program demonstrates how regional innovation labs and cloud partners can empower early-stage biotech.

Learn more about NVIDIA Inception startups advancing AI for healthcare and life sciences in Europe at NVIDIA GTC Paris, taking place June 10-12 at VivaTech. 

Watch the NVIDIA GTC Paris keynote from NVIDIA founder and CEO Jensen Huang at VivaTech, and explore GTC Paris sessions.

See notice regarding software product information.

AI is already driving revenue increases for financial institutions — and with new investments in AI infrastructure and development across Europe, the region’s financial services industry is poised to mint even greater value from the technology.

With sovereign AI models and agents built using AI factories, financial institutions and digital payment companies can extract powerful insights from their vast data sources to protect investments, detect fraud and offer personalized services to customers.

At the NVIDIA GTC Paris at VivaTech, one of Europe’s largest finance companies announced that it’s building an NVIDIA-powered AI factory to deploy sovereign AI for wide-ranging financial services.

Banks and online payment companies operating across the continent are harnessing NVIDIA AI and data science libraries to speed up data analysis for fraud detection and other applications. And the region’s AI platforms and service providers are helping banks and fintech companies accelerate their workflows with AI agents and models built on NVIDIA software libraries, models and blueprints.

European Bank Builds AI Factory to Develop and Scale Financial Services Applications

Across Europe, banks are building regional AI factories to enable the deployment of AI models for customer service, fraud detection, risk modeling and the automation of regulatory compliance.

In Germany, Finanz Informatik, the digital technology provider of the Savings Banks Finance Group, is scaling its on-premises AI factory and using NVIDIA AI Enterprise software for applications including an AI assistant to help its employees automate routine tasks and efficiently process the institution’s banking data.

Financial Services Companies Speed Data Science and Processing

Leading online payment and banking providers in Europe are tapping NVIDIA CUDA-X AI and data science libraries to accelerate financial data processing and analysis.

Amsterdam-based neobank bunq, which serves over 17 million users in the European Union, uses NVIDIA-accelerated XGBoost to boost fraud detection workflows.

The company’s AI-powered monitoring system is used to flag suspicious transactions that present risk of fraud or money laundering. Using NVIDIA GPUs running XGBoost and NVIDIA cuDF, bunq accelerated its model training by 100x and data processing by 5x.

The company is also using NVIDIA NIM microservices to implement and scale large language model-powered applications like its personal AI assistant, dubbed Finn. The bank uses NVIDIA NeMo Retriever, a collection of NIM microservices for extracting, embedding and reranking enterprise data so it can be semantically searched, which can help further improve Finn’s accuracy.

The recently launched NVIDIA AI Blueprint for financial fraud detection also includes XGBoost to support anomaly detection from financial data. The NVIDIA AI Blueprint is available for customers to run on Amazon Web Services and Hewlett Packard Enterprise, with availability coming soon on Dell Technologies. Customers can also adopt the blueprint through NVIDIA partners including Cloudera, EXL, Infosys and SHI International.

Checkout.com is a London-based fintech company providing digital payment solutions to enterprises around the world. The company, which operates in more than 55 countries and supports over 180 currencies, is speeding up data analysis pipelines from minutes to under 10 seconds using the NVIDIA cuDF accelerator for pandas — the go-to Python library for data handling and analysis.

Checkout.com is also exploring the use of NVIDIA cuML and the RAPIDS Accelerator for Apache Spark to further boost analysis of the company’s terabyte-scale data lake.

PayPal, based in the U.S., is another popular digital payment platform for European customers. The company used the RAPIDS Accelerator for Apache Spark to achieve a 70% cost reduction for Spark-based data pipelines running on NVIDIA accelerated computing.

Investment management firms are adopting GPU-accelerated optimization for capital allocation in dynamic markets. The NVIDIA cuFOLIO module, built on the NVIDIA cuOpt optimization engine, enables rapid portfolio adjustments that balance risk, return and investor preferences — turning time-consuming, CPU-bound workflows into scalable, real-time simulation engines.

AI Platforms, Solution Providers Offer NVIDIA-Accelerated Financial Services

European software companies and solution providers are integrating NVIDIA AI software to accelerate financial services workflows for their customers.

Dataiku, an AI platform company founded in Paris and based in New York City, announced at GTC Paris a new blueprint to help banking and insurance institutions deploy agentic AI systems at scale. The company is also integrating the NVIDIA Enterprise AI Factory validated design to accelerate AI development, and offers native integration of its LLM Mesh platform with NVIDIA NIM microservices.

KX, a global data and analytics software company based in the U.K., launched an AI Banker Agent Blueprint at GTC Paris. Built with NVIDIA AI tools including the NVIDIA NeMo platform, Nemotron family of models and NIM microservices, the blueprint can be deployed by banks as an AI-powered research assistant, client relationship manager or personalized customer portfolio manager.

Temenos, a global provider of banking technology, uses NIM microservices to deploy its generative AI models to banks. The company’s generative AI solutions can be applied to use cases including credit scoring, fraud detection and customer service.

Watch the NVIDIA GTC Paris keynote from NVIDIA founder and CEO Jensen Huang at VivaTech, and explore GTC Paris sessions.

AI’s in fashion in France — as it is across the globe — with the technology already helping solve some of the country’s greatest challenges across research and innovation, transportation, manufacturing and many other industries. And this fashion’s here to stay.

France’s National Strategy for AI, part of the broader France 2030 investment plan, includes more than €109 billion in investments for the country’s AI infrastructure projects.

Such projects include a collaboration between NVIDIA and Mistral AI, an independent generative AI pioneer headquartered in France, to build a cutting-edge, end-to-end compute platform that answers the comprehensive compute infrastructure needs of enterprise customers.

Plus, a slew of the nation’s AI-native companies, startups and research centers are innovating with NVIDIA AI infrastructure.

These leading innovators are using the latest agentic and industrial AI technologies to bolster and accelerate work in areas ranging from advertising for skincare and beauty, spearheaded by L’Oréal and Accenture, to transportation and the electric grid.

With its decarbonized, abundant electricity supply, expanding high-voltage electric grid and more than 30 ready-to-use, low-carbon AI sites throughout the country, France is poised to become one of the world’s greenest leaders in artificial intelligence.

Below are some of the key players making AI development the nation’s hottest trend.

AI Infrastructure Development Across Industries

Mistral AI’s new compute platform will feature the latest-generation NVIDIA Grace Blackwell systems, with 18,000 Blackwell and Blackwell Ultra GPUs planned for deployment in the initial phase and additional plans to expand across multiple sites in 2026. The infrastructure will host Mistral AI’s cloud application service, which customers can use to develop and run AI applications with Mistral AI’s and other providers’ open-source models.

Mistral AI and NVIDIA are optimizing inference performance for several Mistral models with NVIDIA NIM microservices, including the new Mistral Nemotron model, exclusively available with the NVIDIA AI Enterprise software platform.

“We are forging Europe’s AI future in partnership with NVIDIA, combining strategic autonomy with our expertise in AI and NVIDIA’s most advanced technology,” said Arthur Mensch, CEO of Mistral AI. “This new infrastructure will provide enterprises and the public sector with Mistral’s AI expertise in building the best compute for AI, ensuring full control to businesses.”

In addition, Mistral AI and NVIDIA are collaborating with Bpifrance, the French national investment bank, and MGX, the UAE’s investment fund focused on AI and advanced technology, to establish Europe’s largest AI campus — to be located in the Paris region and expected to reach a capacity of 1.4 gigawatts. The campus will feature advanced NVIDIA compute infrastructure to support the full AI lifecycle, from model training and inference to deployment of generative and applied AI systems.

France-founded European cloud service provider Scaleway offers the European cloud’s largest compute capacity, powered by more than a thousand NVIDIA Hopper GPUs, with plans to offer NVIDIA Blackwell GPUs — which enable building and running real-time generative AI on trillion-parameter large language models at up to 25x less cost and energy consumption than its predecessor. As a European provider, Scaleway provides domestic infrastructure that ensures access and compliance with EU data protection laws — critical to businesses with a European footprint.

Mistral AI and Scaleway plan to participate in the DGX Cloud Lepton marketplace to provide startups and developers access to compute infrastructure.

Orange Business, the enterprise division of Orange, one of Europe’s leading telco operators, has joined the NVIDIA Cloud Partner program to accelerate the development of enterprise-grade agentic AI, including its innovative Live Intelligence platform, which empowers companies of all sizes to securely deploy generative AI at scale. Those AI solutions tap into the Orange Business Cloud Avenue platform, built on high-performance NVIDIA infrastructure.

AI Deployments, From Beauty to Transportation

Paris-based beauty company L’Oréal Groupe’s generative AI content platform CREAITECH uses the NVIDIA AI Enterprise platform to develop and deploy 3D digital renderings of L’Oréal’s products for faster, more creative development of marketing and advertising campaigns. Eighty percent of L’Oréal’s production in France is exported globally, helping make cosmetics the third-largest contributor to national economic growth.

Learn more about how L’Oréal and other leading retailers are using NVIDIA technologies to redefine their operations.

The France public sector uses NVIDIA technologies for use cases ranging from transportation and public safety in cities to cybersecurity in schools and better fraud detection at the French Ministry of the Economy, Finance, and Industrial and Digital Sovereignty, which oversees national funds and the economic system. Local governments have deployed solutions in generative and vision AI, document analytics and more through NVIDIA partners Dell Technologies, Hewlett Packard Enterprise, LightOn, SCC, ThinkDeep, XXII and others.

France’s national rail operator SNCF Gares&Connexions, which operates internationally and has a network of 3,000 train stations across France and Monaco, is developing digital twins to simulate railway scenarios.

Powered by NVIDIA Omniverse, Metropolis and ecosystem partners Akila and XXII, SNCF Gares&Connexions’ AI deployment, including at the Monaco-Monte-Carlo station, has helped SNCF Gares&Connexions achieve a 100% on-time preventive maintenance completion rate, a 50% reduction in downtime and issue response time, as well as a 20% reduction in energy consumption.

Schneider Electric — a French multinational company driving the digital transformation of energy management and automation — has introduced publicly available engineered reference designs for optimizing performance, scalability and energy efficiency of NVIDIA-powered AI data centers. In addition, ETAP, a subsidiary of Schneider Electric, is connecting its digital twin platform to NVIDIA Omniverse to deliver a unified virtual simulation and collaboration environment for designing and deploying optimized data centers.

Electricité de France, commonly known as EDF, the French national electricity company, has partnered with NVIDIA to transition its open-source code_saturne computational fluid dynamics (CFD) application, developed by EDF R&D, onto accelerated computing platforms for improved performance in power and industrial applications. This collaboration, which also involves NVIDIA developer partner ANEO, taps into NVIDIA Nsight tools to iteratively adapt the CFD code for optimized GPU operation.

AI-Native Companies Build Models, Cloud Services to Accelerate Next Industrial Revolution

To accelerate France’s AI-driven transformation, NVIDIA is partnering with the country’s leading model builders and AI-native companies to support large language models in various languages including Arabic, French, English, Italian, German, Polish, Spanish and Swedish.

H Company and LightOn are tailoring and optimizing their models with NVIDIA Nemotron techniques to maximize cost efficiency and accuracy for enterprise AI workloads including agentic AI.

Plus, a new Hugging Face integration with DGX Cloud Lepton will let companies fine-tune their AI models on local NCP infrastructure.

Startups Develop Breakthroughs With NVIDIA AI Infrastructure

France has a rich ecosystem of more than 1,000 AI startups pursuing breakthroughs in healthcare, quantum computing and more.

Alice & Bob, a member of the NVIDIA Inception program for cutting-edge startups, is building quantum computing hardware and has integrated the NVIDIA CUDA-Q hybrid computing platform into its quantum simulation library, called Dynamiqs. This allows the company to accelerate its qubit design process with GPUs. Adding NVIDIA acceleration on top of Dynamiqs’ advanced optimization capabilities can increase the efficiency of these challenging qubit-design simulations by up to 75x.

Quandela, a leader in full-stack photonic quantum computing, has announced MerLin, a quantum machine learning programming framework that uses NVIDIA CUDA-Q to deliver high-performance simulations for photonic quantum circuits. This enables developers to build new models and assess the performance of candidate algorithms on simulations of larger quantum processors.

Moon Surgical, a robotic surgery company and also an NVIDIA Inception member, is using the NVIDIA Holoscan and IGX platforms to power its Maestro System for minimally invasive surgery, a technique where surgeons operate through small incisions with an internal camera and instruments. Moon Surgical and NVIDIA are also collaborating to bring generative AI features to the operating room using Maestro and Holoscan.

Research Centers Bring Future of Technology Closer to Reality

As the country with the world’s third largest number of AI researchers, France supports a vast spectrum of projects and centers advancing supercomputing, AI education and other initiatives to make the future of technology possible.

The Jean Zay supercomputer, operated by IDRIS, a national computing centre for the CNRS (France’s National Centre for Scientific Research), is a French AI flagship serving research academia and startups users. Built by Eviden and powered by NVIDIA, the supercomputer accelerates the work of university and public sector researchers, developers and data scientists across France.

Acquired by the French government through intermediary French civil company GENCI, the supercomputer integrates NVIDIA accelerated computing, including more than a thousand NVIDIA Hopper GPUs.

It supports more than 150 startups — including Hugging Face, Mistral AI, H Company and LightOn — and powered 1,400+ AI projects in 2024. Jean Zay is among the most eco-efficient machines in Europe, thanks to the accelerated technologies and core warm-water cooling of the computing servers. In addition, the supercomputer’s waste heat is reused to help heat more than 1,500 homes in the Saclay area, to the southwest of Paris.

Learn more about the latest AI advancements in France and other countries at NVIDIA GTC Paris, running through Thursday, June 12, at VivaTech. Watch the keynote from NVIDIA founder and CEO Jensen Huang, and explore GTC Paris sessions.

See notice regarding software product information.

U.K. Prime Minister Keir Starmer’s ambition for Britain to be an “AI maker, not an AI taker,” is becoming a reality at London Tech Week.

With NVIDIA’s support, the U.K. is building sovereign compute infrastructure, investing in cutting-edge research and skills, and fostering AI leadership across sectors.

As London Tech Week kicks off today, NVIDIA and some of Britain’s best companies are convening and hosting the first U.K. Sovereign AI Industry Forum.

The initiative unites leading U.K. businesses — including founding members Babcock, BAE Systems, Barclays, BT, National Grid and Standard Chartered — to strengthen the nation’s economic security by advancing sovereign AI infrastructure and accelerating the growth of the U.K. AI startup ecosystem.

“We have big plans when it comes to developing the next wave of AI innovations here in the U.K. — not only so we can deliver the economic growth needed for our Plan for Change, but maintain our position as a global leader,” U.K. Secretary of State for Science, Innovation and Technology Peter Kyle said. “Central to that is making sure we have the infrastructure to power AI, so I welcome NVIDIA setting up the U.K. Sovereign AI Industry Forum — bringing together leading British businesses to develop and deploy this across the U.K. so we can drive growth and opportunity.”

The U.K. is a global AI hub, leading Europe in newly funded AI startups and total private AI investment through 2024. And the sector is growing fast, backed by over $28 billion in private investment since 2013.

And AI investment benefits the whole of the U.K.

According to an analysis released today by Public First, regions with more AI and data center infrastructure consistently show stronger economic growth. Even a modest increase in AI data center capacity could add nearly £5 billion to national economic output, while a more significant increase, for example, doubling access, could raise the annual benefit to £36.5 billion.

Responding to this opportunity, cloud provider Nscale announced at London Tech Week its commitment to deploy U.K. AI infrastructure with 10,000 NVIDIA Blackwell GPUs by the end of 2026. This facility will help position the U.K. as a global leader in AI, supporting innovation, job creation and the development of a thriving domestic AI ecosystem.

And cloud provider Nebius is continuing the region’s momentum with the launch of its first AI factory in the U.K. It announced it’s bringing 4,000 NVIDIA Blackwell GPUs online, making available scalable, high-performance AI capacity at home in the U.K. — to power U.K. research, academia and public services, including the NHS.

Mind the (Skills) Gap

AI developers are the engine of this new industrial revolution. That’s why NVIDIA is supporting the U.K. government’s national skills drive by training developers in AI.

To support this goal, a new NVIDIA AI Technology Center in the U.K. will provide hands-on training in AI, data science and accelerated computing with support from the NVIDIA Deep Learning Institute, focusing on foundation model builders, embodied AI, materials science and earth systems modeling.

Beyond training, this collaboration drives cutting-edge AI applications and research.

For example, the U.K.’s world-leading financial services industry gets a boost from a new AI-powered digital sandbox. This sandbox, a digital testing environment for safe AI innovation in financial services, will be provided by the Financial Conduct Authority, with infrastructure provided by NayaOne and supported by NVIDIA’s platform.

At the same time, Barclays Eagle Labs’ launch of an Innovation Hub in London will help AI and deep tech startups grow to the next level. NVIDIA is supporting the program by offering startups a pathway to the NVIDIA Inception program with access to advanced tools and training.

Furthermore, the Department for Science, Innovation and Technology announced a collaboration with NVIDIA to promote the nation’s goals for AI development in telecoms. Leading U.K. universities will gain access to a suite of powerful AI tools, 6G research platforms and training resources to bolster research and development on AI-native wireless networks.

The Research Engine

Universities are central to the U.K.’s strategy.

Led by Oxford University, the JADE consortium, comprising 20 universities and the Turing Institute, uses NVIDIA technologies to advance AI development and safety. At University College London, researchers are developing a digital twin of the human body enabled by NVIDIA technology. At the University of Bristol, the Isambard-AI supercomputer, built on NVIDIA Grace Hopper Superchips, is powering progress in AI safety, climate modeling and next-generation science. And at the University of Manchester, the NVIDIA Earth-2 platform is being deployed to develop pollution-flow models.

Meanwhile, U.K. tech leaders use NVIDIA’s foundational technologies to innovate across diverse sectors.

It’s how Wayve trains AI for autonomous vehicles. How JBA Risk Management helps organizations anticipate and mitigate climate risks with new precision. And how Stability AI is unleashing creativity with open-source generative AI that turns ideas into images, text and more — instantly.

NVIDIA also champions the U.K.’s most ambitious AI startups through NVIDIA Inception, providing specialized resources and support for startups building new products and services.

Basecamp Research is revolutionizing drug discovery with AI trained on the planet’s biodiversity. Humanoid advances automation and brings commercially scalable, reliable and safe humanoid robots closer to real-world deployment. Relation is accelerating the discovery of tomorrow’s medicines. And Synthesia turns text into studio-quality, multilingual videos with lifelike avatars.

Industry in Motion

The U.K.’s biggest companies are moving fast, too.

Companies like BT, LSEG and NatWest are transforming industries with AI. BT is powering agentic AI-based autonomous operations; LSEG is empowering customers with highly accurate, AI-driven data and insights; and NatWest is streamlining operations and safeguarding customers.

With government vision, talent and cutting-edge tech converging, the U.K. is taking its place among those making AI advances at home and worldwide.

Watch NVIDIA founder and CEO Jensen Huang’s keynote NVIDIA GTC Paris at VivaTech.

Whether at sea, land or in the sky — even outer space — NVIDIA technology is helping research scientists and developers alike explore and understand oceans, wildlife, the climate and far out existential risks like asteroids.

These increasingly intelligent developments are helping to analyze environmental pollutants, damage to habitats and natural disaster risks at an accelerated pace. This, in turn, enables partnerships with local governments to take climate mitigation steps like pollution prevention and proactive planting.

Sailing the Seas of AI

Amphitrite, based in France, uses satellite data with AI to simulate and predict ocean currents and weather. Its AI models, driven by the NVIDIA AI and Earth-2 platforms, offer insights for positioning vessels to best harness the power of ocean currents. This helps determine when it’s best to travel, as well as the optimal course, reducing travel times, fuel consumption and carbon emissions. Amphitrite is a member of the NVIDIA Inception program for cutting-edge startups.

Watching Over Wildlife With AI

München, Germany-based OroraTech monitors animal poaching and wildfires with NVIDIA CUDA and Jetson. The NVIDIA Inception program member uses the EarthRanger platform to offer a wildfire detection and monitoring service that uses satellite imagery and AI to safeguard the environment and prevent poaching.

Keeping AI on the Weather

Weather agencies and climate scientists worldwide are using NVIDIA CorrDiff, a generative AI weather model enabling kilometer-scale forecasts of wind, temperature and precipitation type and amount. CorrDiff is part of the NVIDIA Earth-2 platform for simulating weather and climate conditions. It’s available as an easy-to-deploy NVIDIA NIM microservice.

In another climate effort, NVIDIA Research announced a new generative AI model, called StormCast, for reliable weather prediction at a scale larger than storms.

The model, outlined in a paper, can help with disaster and mitigation planning, saving lives.

Avoiding Mass Extinction Events

Researchers reported in Nature how a new method was able to spot 10-meter asteroids within the main asteroid belt located between Jupiter and Mars. Such space rocks can range from bus-sized to several Costco stores in width and deliver destruction to cities. It used NASA’s James Webb Space Telescope (JWST), which was tapped for views of these asteroids from previous research and enabled by NVIDIA accelerated computing.

Boosting Energy Efficiency With Liquid-Cooled Blackwell

NVIDIA GB200 NVL72 rack-scale, liquid-cooled systems, built on the Blackwell platform, offer exceptional performance while balancing energy costs and heat. It delivers 40x higher revenue potential, 30x higher throughput, 25x more energy efficiency and 300x more water efficiency than air-cooled architectures. NVIDIA GB300 NVL72 systems built on the Blackwell Ultra platform offer a 50x higher revenue potential, 35x higher throughput with 30x more energy efficiency.

Enroll in the free new NVIDIA Deep Learning Institute course Applying AI Weather Models With NVIDIA Earth-2. Learn more about NVIDIA Earth-2 and NVIDIA Blackwell.

From Seattle, Washington, to Cape Town, South Africa — and everywhere around and between — AI is helping conserve the wild plants and animals that make up the intricate web of life on Earth.

It’s critical work that sustains ecosystems and supports biodiversity at a time when the United Nations estimates over 1 million species are threatened with extinction.

World Wildlife Day, a UN initiative, is celebrated every March 3 to recognize the unique contributions wild animals and plants have on people and the planet — and vice versa.

“Our own survival depends on wildlife,” the above video on this year’s celebration says, “just as much as their survival depends on us.”

Learn more about some of the leading nonprofits and startups using NVIDIA AI and accelerated computing to protect wildlife and natural habitats, today and every day:

Ai2’s EarthRanger Offers World’s Largest Elephant Database

Seattle-based nonprofit AI research institute Ai2 offers EarthRanger, a software platform that helps protected-area managers, ecologists and wildlife biologists make more informed operational decisions for wildlife conservation in real time, whether preventing poaching, spotting ill or injured animals, or studying animal behavior.

Among Ai2’s efforts with EarthRanger is the planned development of a machine learning model — trained using NVIDIA Hopper GPUs in the cloud — that predicts the movement of elephants in areas close to human-wildlife boundaries where elephants could raid crops and potentially prompt humans to retaliate.

With access to the world’s largest repository of elephant movement data, made possible by EarthRanger users who’ve shared their data, the AI model could help predict elephant behaviors, then alert area managers to safely guide the elephants away from risky situations that could arise for them or for people in the vicinity. Area managers or rangers typically use helicopters, other vehicles and chili bombs to safely reroute elephants.

Beyond elephants, EarthRanger collects, integrates and displays data on a slew of wildlife — aggregated from over 100 data sources, including camera traps, acoustic sensors, satellites, radios and more. Then, the platform combines the data with field reports to provide a unified view of collared wildlife, rangers, enforcement assets and infrastructure within a protected area.

“Name a country, species or an environmental cause and we’re probably supporting a field organization’s conservation efforts there,” said Jes Lefcourt, director of EarthRanger at Ai2.

It’s deployed by governments and conservation organizations in 76 countries and 650 protected areas, including nearly every national park in Africa, about a dozen state fishing and wildlife departments in the U.S., as well as many other users across Latin America and Asia.

Four of these partners — Rouxcel Technology, OroraTech, Wildlife Protection Services and Conservation X Labs — are highlighted below.

Rouxcel Technology Saves Rhinos With AI

South African startup Rouxcel Technology’s AI-based RhinoWatches, tapping into EarthRanger, learn endangered black and white rhinos’ behaviors, then alert authorities in real time of any detected abnormalities. These abnormalities can include straying from typical habitats, territorial fighting with other animals and other potentially life-threatening situations.

It’s critical work, as there are just about 28,000 rhinos left in the world, from 500,000 at the beginning of the 20th century.

Rouxcel, based in Cape Town, has deployed over 1,200 RhinoWatches — trained and optimized using NVIDIA accelerated computing — across more than 40 South African reserves. The startup, which uses the Ai2 EarthRanger platform, protects more than 1.2 million acres of rhino habitats, and has recently expanded to help conservation efforts in Kenya and Namibia.

Looking forward, Rouxcel is developing AI models to help prevent poaching and human-wildlife conflict for more species, including pangolins, a critically endangered species.

OroraTech Monitors Wildfires and Poaching With NVIDIA CUDA, Jetson

OroraTech — a member of the NVIDIA Inception program for cutting-edge startups — uses the EarthRanger platform to protect wildlife in a different way, offering a wildfire detection and monitoring service that fuses satellite imagery and AI to safeguard the environment and prevent poaching.

Combining data from satellites, ground-based cameras, aerial observations and local weather information, OroraTech detects threats to natural habitats and alerts users in real time. The company’s technologies monitor more than 30 million hectares of land that directly impact wildlife in Africa and Australia. That’s nearly the size of the Great Barrier Reef.

OroraTech flies an NVIDIA Jetson module for edge AI and data processing onboard all of its satellite payloads — the instruments, equipment and systems on a satellite designed for performing specific tasks. Through GPU-accelerated image processing, OroraTech achieves exceptional latency, delivering fire notifications to users on the ground as fast as five minutes after image acquisition.

The AI-based fire-detection pipeline uses the NVIDIA cuDNN library of deep neural network primitives and the NVIDIA TensorRT software development kit for thermal anomaly detection and cloud masking in space, leading to high-precision fire detections.

Wildlife Protection Solutions Help Preserve Endangered Species

International nonprofit Wildlife Protection Solutions (WPS) supports more than 250 conservation projects in 50+ countries. Its remote cameras — about 3,000 deployed across the globe — using AI models provide real-time monitoring of animals and poachers, alerting rangers to intercede before wildlife is harmed.

WPS — which also taps into the EarthRanger platform — harnesses NVIDIA accelerated computing to optimize training and inference of its AI models, which process and analyze 65,000 photos per day.

The WPS tool is free and available on any mobile, tablet or desktop browser, enabling remote monitoring, early alerting and proactive, automated deterrence of wildlife or humans in sensitive areas.

Conservation X Labs Identifies Species From Crowdsourced Images

Washington D.C.-based Conservation X Labs — which is on a mission to prevent the sixth mass extinction, or the dying out of a high percentage of the world’s biodiversity due to natural phenomena and human activity — also uses EarthRanger, including for its Wild Me solution: open-source AI software for the conservation research community.

Wild Me supports over 2,000 researchers across the globe running AI-enabled wildlife population studies for marine and terrestrial species.

For example, Wild Me helps researchers classify whale sharks using computer vision:

Above video courtesy of Wild Me by Conservation X Labs.

The crowdsourced database — which currently comprises 14 million photos — lets anyone upload imagery of species. Then, AI foundation models trained using NVIDIA accelerated computing help identify species to ease and accelerate animal population assessments and other research that supports the fight against species extinction.

In addition, Conservation X Labs’s Sentinel technology transforms traditional wildlife monitoring tools — like trail cameras and acoustic recorders — with AI, processing environmental data as it’s collected and providing conservationists with real-time, data-driven insights through satellite and cellular networks.

To date, Sentinel devices have delivered about 100,000 actionable insights for 80 different species. For example, see how the technology flags a limping panther, so wildlife protectors could rapidly step in to offer aid:

Above video courtesy of Sentinel by Conservation X Labs.

Learn more about how NVIDIA technologies bolster conservation and environmental initiatives at NVIDIA GTC, a global AI conference running March 17-21 in San Jose, California, including at sessions on how AI is supercharging Antarctic flora monitoring, enhancing a digital twin of the Great Barrier Reef and helping mitigate urban climate change.

Featured video — depicting Persian leopards in Turkmenistan — courtesy of Sentinel by Conservation X Labs.

Editor’s note: This article, originally published on March 13, 2023, has been updated.

The mics were live and tape was rolling in the studio where the Miles Davis Quintet was recording dozens of tunes in 1956 for Prestige Records.

When an engineer asked for the next song’s title, Davis shot back, “I’ll play it, and tell you what it is later.”

Like the prolific jazz trumpeter and composer, researchers have been generating AI models at a feverish pace, exploring new architectures and use cases. According to the 2024 AI Index report from the Stanford Institute for Human-Centered Artificial Intelligence, 149 foundation models were published in 2023, more than double the number released in 2022.

They said transformer models, large language models (LLMs), vision language models (VLMs) and other neural networks still being built are part of an important new category they dubbed foundation models.

Foundation Models Defined

A foundation model is an AI neural network — trained on mountains of raw data, generally with unsupervised learning — that can be adapted to accomplish a broad range of tasks.

Two important concepts help define this umbrella category: Data gathering is easier, and opportunities are as wide as the horizon.

No Labels, Lots of Opportunity

Foundation models generally learn from unlabeled datasets, saving the time and expense of manually describing each item in massive collections.

Earlier neural networks were narrowly tuned for specific tasks. With a little fine-tuning, foundation models can handle jobs from translating text to analyzing medical images to performing agent-based behaviors.

“I think we’ve uncovered a very small fraction of the capabilities of existing foundation models, let alone future ones,” said Percy Liang, the center’s director, in the opening talk of the first workshop on foundation models.

AI’s Emergence and Homogenization

In that talk, Liang coined two terms to describe foundation models:

Emergence refers to AI features still being discovered, such as the many nascent skills in foundation models. He calls the blending of AI algorithms and model architectures homogenization, a trend that helped form foundation models. (See chart below.)

The field continues to move fast.

A year after the group defined foundation models, other tech watchers coined a related term — generative AI. It’s an umbrella term for transformers, large language models, diffusion models and other neural networks capturing people’s imaginations because they can create text, images, music, software, videos and more.

Generative AI has the potential to yield trillions of dollars of economic value, said executives from the venture firm Sequoia Capital who shared their views in a recent AI Podcast.

A Brief History of Foundation Models

“We are in a time where simple methods like neural networks are giving us an explosion of new capabilities,” said Ashish Vaswani, an entrepreneur and former senior staff research scientist at Google Brain who led work on the seminal 2017 paper on transformers.

That work inspired researchers who created BERT and other large language models, making 2018 “a watershed moment” for natural language processing, a report on AI said at the end of that year.

Google released BERT as open-source software, spawning a family of follow-ons and setting off a race to build ever larger, more powerful LLMs. Then it applied the technology to its search engine so users could ask questions in simple sentences.

In 2020, researchers at OpenAI announced another landmark transformer, GPT-3. Within weeks, people were using it to create poems, programs, songs, websites and more.

“Language models have a wide range of beneficial applications for society,” the researchers wrote.

Their work also showed how large and compute-intensive these models can be. GPT-3 was trained on a dataset with nearly a trillion words, and it sports a whopping 175 billion parameters, a key measure of the power and complexity of neural networks. In 2024, Google released Gemini Ultra, a state-of-the-art foundation model that requires 50 billion petaflops.

“I just remember being kind of blown away by the things that it could do,” said Liang, speaking of GPT-3 in a podcast.

The latest iteration, ChatGPT — trained on 10,000 NVIDIA GPUs — is even more engaging, attracting over 100 million users in just two months. Its release has been called the iPhone moment for AI because it helped so many people see how they could use the technology.

Going Multimodal

Foundation models have also expanded to process and generate multiple data types, or modalities, such as text, images, audio and video. VLMs are one type of multimodal models that can understand video, image and text inputs while producing text or visual output.

Trained on 355,000 videos and 2.8 million images,

Cosmos Nemotron 34B is a leading VLM that enables the ability to query and summarize images and video from the physical or virtual world.

From Text to Images

About the same time ChatGPT debuted, another class of neural networks, called diffusion models, made a splash. Their ability to turn text descriptions into artistic images attracted casual users to create amazing images that went viral on social media.

The first paper to describe a diffusion model arrived with little fanfare in 2015. But like transformers, the new technique soon caught fire.

In a tweet, Midjourney CEO David Holz revealed that his diffusion-based, text-to-image service has more than 4.4 million users. Serving them requires more than 10,000 NVIDIA GPUs mainly for AI inference, he said in an interview (subscription required).

Toward Models That Understand the Physical World

The next frontier of artificial intelligence is physical AI, which enables autonomous machines like robots and self-driving cars to interact with the real world.

AI performance for autonomous vehicles or robots requires extensive training and testing. To ensure physical AI systems are safe, developers need to train and test their systems on massive amounts of data, which can be costly and time-consuming.

World foundation models, which can simulate real-world environments and predict accurate outcomes based on text, image, or video input, offer a promising solution.

Physical AI development teams are using NVIDIA Cosmos world foundation models, a suite of pre-trained autoregressive and diffusion models trained on 20 million hours of driving and robotics data, with the NVIDIA Omniverse platform to generate massive amounts of controllable, physics-based synthetic data for physical AI. Awarded the Best AI And Best Overall Awards at CES 2025, Cosmos world foundation models are open models that can be customized for downstream use cases or improve precision on a specific task using use case-specific data.

Dozens of Models in Use

Hundreds of foundation models are now available. One paper catalogs and classifies more than 50 major transformer models alone (see chart below).

The Stanford group benchmarked 30 foundation models, noting the field is moving so fast they did not review some new and prominent ones.

Startup NLP Cloud, a member of the NVIDIA Inception program that nurtures cutting-edge startups, says it uses about 25 large language models in a commercial offering that serves airlines, pharmacies and other users. Experts expect that a growing share of the models will be made open source on sites like Hugging Face’s model hub.

Foundation models keep getting larger and more complex, too.

That’s why — rather than building new models from scratch — many businesses are already customizing pretrained foundation models to turbocharge their journeys into AI, using online services like NVIDIA AI Foundation Models.

The accuracy and reliability of generative AI is increasing thanks to techniques like retrieval-augmented generation, aka RAG, that lets foundation models tap into external resources like a corporate knowledge base.

AI Foundations for Business

Another new framework, the NVIDIA NeMo framework, aims to let any business create its own billion- or trillion-parameter transformers to power custom chatbots, personal assistants and other AI applications.

It created the 530-billion parameter Megatron-Turing Natural Language Generation model (MT-NLG) that powers TJ, the Toy Jensen avatar that gave part of the keynote at NVIDIA GTC last year.

Foundation models — connected to 3D platforms like NVIDIA Omniverse — will be key to simplifying development of the metaverse, the 3D evolution of the internet. These models will power applications and assets for entertainment and industrial users.

Factories and warehouses are already applying foundation models inside digital twins, realistic simulations that help find more efficient ways to work.

Foundation models can ease the job of training autonomous vehicles and robots that assist humans on factory floors and logistics centers. They also help train autonomous vehicles by creating realistic environments like the one below.

New uses for foundation models are emerging daily, as are challenges in applying them.

Several papers on foundation and generative AI models describing risks such as:

• amplifying bias implicit in the massive datasets used to train models,
• introducing inaccurate or misleading information in images or videos, and
• violating intellectual property rights of existing works.

“Given that future AI systems will likely rely heavily on foundation models, it is imperative that we, as a community, come together to develop more rigorous principles for foundation models and guidance for their responsible development and deployment,” said the Stanford paper on foundation models.

Current ideas for safeguards include filtering prompts and their outputs, recalibrating models on the fly and scrubbing massive datasets.

“These are issues we’re working on as a research community,” said Bryan Catanzaro, vice president of applied deep learning research at NVIDIA. “For these models to be truly widely deployed, we have to invest a lot in safety.”

It’s one more field AI researchers and developers are plowing as they create the future.

NVIDIA’s contributions to accelerating medical imaging, genomics, computational chemistry and AI-powered robotics were honored Friday at the Precision Medicine World Conference in Santa Clara, California, where NVIDIA founder and CEO Jensen Huang received a Luminary award.

The Precision Medicine World Conference brings together healthcare leaders, top global researchers and innovators across biotechnology. Its Luminary award recognizes people transforming healthcare by advancing precision medicine in the clinic.

For nearly two decades, NVIDIA has advanced computing in healthcare — working with researchers and industry leaders to build instruments that enable scientists to better understand life sciences, medical imaging and genomics.

“We built, if you will, a computational instrument. Not a gene sequencer and all the incredible scientific instruments that you all talk about here — in our case, it was a programmable scientific instrument,” Huang said in his acceptance speech. “We built it in service of researchers and scientists as you strive to better understand life in our universe.”

The first use of accelerated computing in life sciences was in the 2000s — and the introduction of the NVIDIA CUDA parallel computing platform in 2006 paved the path for researchers to demonstrate how NVIDIA GPUs could be used in medical imaging applications like CT reconstruction.

“NVIDIA developed and continues to develop GPUs that are at the heart of AI and machine learning that are changing the world, including precision medicine,” said Dr. Gad Getz, an internationally acclaimed leader in cancer genomics and the director of bioinformatics at the Massachusetts General Hospital, as he presented the award.

Today, NVIDIA AI and accelerated computing is “impacting analysis, interpretation and translation of sequencing data, new sequencing technologies, imaging data, spatial technologies, single-cell genomics, proteomics, molecular dynamics and drug development, as well as the large language models that can be used by doctors, patients, students and teachers to learn this field,” Getz said.

Advancing Precision Medicine With Accelerated Computing 

Huang spoke about the ways AI will support the work of doctors, scientists and researchers advancing medicine. By investing in AI, he explained, research organizations and businesses can set up a powerful flywheel that continuously improves in accuracy, efficiency and insights by integrating additional data and feedback from every expert who interacts with it over time.

“Even though people say you want humans in the loop with AI, in fact, the opposite is true. You want AI in the loop with humans,” Huang said. “The reason for that is because when the AI is in the loop with humans, it codifies our life experience. If there’s an AI in the loop with every single researcher, scientist, engineer and marketer — every single employee in your company — that AI in the loop codifies that life experience and keeps it in the company.”

Looking ahead, Huang said that “in the coming years, AI will advance with incredible speed and revolutionize the healthcare industry. AI will help doctors predict, diagnose and treat disease in ways we never thought possible. It will scan a patient’s genome in seconds, identifying risks before symptoms even appear. AI will build a digital twin of us and model how a tumor evolves, predicting which treatments will work best.”

“I wouldn’t be surprised if before 2030, within this decade, we’re representing basically all cells,” said Huang. “We have a representation of it, we understand the language of it, and we can predict what happens.”

Huang predicts that surgical robots will perform minimally invasive procedures with unparalleled precision, robotic caregivers will assist nurses and other healthcare professionals, and robotic labs will run experiments around the clock, accelerating drug discovery. AI assistants, he said, will let doctors focus on what matters most to them: patients.

In his talk, Huang also thanked the medical research community and highlighted how great breakthroughs come from partnerships between technology companies, researchers, biotech firms and healthcare leaders. Over 4,000 healthcare companies are part of the NVIDIA Inception program designed to help startups evolve faster.

Learn more about accelerated computing in healthcare at NVIDIA GTC, a global AI conference taking place March 17-21 in San Jose, California.

Named after Greek mythology’s goddess of the sea, France-based startup Amphitrite is fusing satellite data and AI to simulate and predict oceanic currents and weather.

It’s work that’s making waves in maritime-shipping and oceanic litter-collection operations.

Amphitrite’s AI models — powered by the NVIDIA AI and Earth-2 platforms — provide insights on positioning vessels to best harness the power of ocean currents, helping ships know when best to travel, as well as the optimal course. This helps users reduce travel times, fuel consumption and, ultimately, carbon emissions.

“We’re at a turning point on the modernization of oceanic atmospheric forecasting,” said Alexandre Stegner, cofounder and CEO of Amphitrite. “There’s a wide portfolio of applications that can use these domain-specific oceanographic AI models — first and foremost, we’re using them to help foster the energy transition and alleviate environmental issues.”

Optimizing Routes Based on Currents and Weather

Founded by expert oceanographers, Amphitrite — a member of the NVIDIA Inception program for cutting-edge startups — distinguishes itself from other weather modeling companies with its domain-specific expertise.

Amphitrite’s fine-tuned, three-kilometer-scale AI models focus on analyzing one parameter at a time, making them more accurate than global numerical modeling methods for the variable of interest. Read more in this paper showcasing the AI method, dubbed ORCAst, trained on NVIDIA GPUs.

Depending on the user’s needs, such variables include the current of the ocean within the first 10 meters of the surface — critical in helping ships optimize their travel and minimize fuel consumption — as well as the impacts of extreme waves and wind.

“It’s only with NVIDIA accelerated computing that we can achieve optimal performance and parallelization when analyzing data on the whole ocean,” said Evangelos Moschos, cofounder and chief technology officer of Amphitrite.

Using the latest NVIDIA AI technologies to predict ocean currents and weather in detail, ships can ride or avoid waves, optimize routes and enhance safety while saving energy and fuel.

“The amount of public satellite data that’s available is still much larger than the number of ways people are using this information,” Moschos said. “Fusing AI and satellite imagery, Amphitrite can improve the accuracy of global ocean current analyses by up to 2x compared with traditional methods.”

Fine-Tuned to Handle Oceans of Data

The startup’s AI models, tuned to handle seas of data on the ocean, are based on public data from NASA and the European Space Agency — including its Sentinel-3 satellite.

Plus, Amphitrite offers the world’s first forecast model incorporating data from the Surface Water and Ocean Topography (SWOT) mission — a satellite jointly developed and operated by NASA and French space agency CNES, in collaboration with the Canadian Space Agency and UK Space Agency.

“SWOT provides an unprecedented resolution of the ocean surface,” Moschos said.

While weather forecasting technologies have traditionally relied on numerical modeling and computational fluid dynamics, these approaches are harder to apply to the ocean, Moschos explained. This is because oceanic currents often deal with nonlinear physics. There’s also simply less observational data available on the ocean than on atmospheric weather.

Computer vision and AI, working with real-time satellite data, offer higher reliability for oceanic current and weather modeling than traditional methods.

Amphitrite trains and runs its AI models using NVIDIA H100 GPUs on premises and in the cloud — and is building on the FourCastNet model, part of Earth-2, to develop its computer vision models for wave prediction.

According to a case study along the Mediterranean Sea, the NVIDIA-powered Amphitrite fine-scale routing solution helped reduce one shipping line’s carbon emissions by 10%.

Through NVIDIA Inception, Amphitrite gained technical support when building its on-premises infrastructure, free cloud credits for NVIDIA GPU instances on Amazon Web Services, as well as opportunities to collaborate with NVIDIA experts on using the latest simulation technologies, like Earth-2 and FourCastNet.

Customers Set Sail With Amphitrite’s Models

Enterprises and organizations across the globe are using Amphitrite’s AI models to optimize their operations and make them more sustainable.

CMA-CGM, Genavir, Louis Dreyfus Armateurs and Orange Marine are among the shipping and oceanographic companies analyzing currents using the startup’s solutions.

In addition, Amphitrite is working with a nongovernmental organization to help track and remove pollution in the Pacific Ocean. The initiative uses Amphitrite’s models to analyze currents and follow plastics that drift from a garbage patch off the coast of California.

Moschos noted that another way the startup sets itself apart is by having an AI team — led by computer vision scientist Hannah Bull — that comprises majority women, some of whom are featured in the image above.

“This is still rare in the industry, but it’s something we’re really proud of on the technical front, especially since we founded the company in honor of Amphitrite, a powerful but often overlooked female figure in history,” Moschos said.

Learn more about NVIDIA Earth-2.

More than 90 million new vehicles are introduced to roads across the globe every year, leading to an annual 12% increase in traffic congestion — according to NoTraffic, a member of the NVIDIA Inception program for cutting-edge startups and the NVIDIA Metropolis vision AI ecosystem.

Still, 99% of the world’s traffic signals run on fixed timing plans, leading to unnecessary congestion and delays.

To reduce such inefficiencies, mitigate car accidents and reduce carbon emissions from vehicles, NoTraffic’s AI Mobility platform predicts road scenarios, helps ensure continuous traffic flow, minimizes stops and optimizes safety at intersections across the U.S., Canada and elsewhere.

The platform — which enables road infrastructure management at both local-intersection and city-grid scale — integrates NVIDIA-powered software and hardware at the edge, under a cloud-based operating system.

It’s built using the NVIDIA Jetson edge AI platform, NVIDIA accelerated computing and the NVIDIA Metropolis vision AI developer stack.

“With NVIDIA accelerated computing, we achieved a 3x speedup in AI training and doubled AI Mobility’s energy efficiency,” said Uriel Katz, cofounder and chief technology officer of NoTraffic. “These optimizations in time, money and energy efficiency are all bolstered by NVIDIA Jetson, which sped our image preprocessing tasks by 40x compared with a CPU-only workflow. Plus, GPU-accelerated NVIDIA CUDA libraries increased our model throughput by 30x.”

These libraries include the NVIDIA TensorRT ecosystem of application programming interfaces for high-performance deep learning inference and the NVIDIA cuDNN library of primitives for deep neural networks.

Taming Traffic in Tuscon, Vancouver and Beyond

In Tuscon, Arizona, more than 80 intersections are tapping into the NoTraffic AI Mobility platform, which has enabled up to a 46% reduction in road delays during rush hours — and a half-mile reduction in peak queue length.

The work is an expansion of NoTraffic’s initial deployment on Tuscon’s West Ajo Way. That effort led to an average delay reduction of 23% for drivers.

Since installation, NoTraffic technology has helped free Tucson drivers from over 1.25 million hours stuck in traffic, the company estimates, representing an economic benefit of over $24.3 million. The company has also tracked a nearly 80% reduction in red-light runners since its platform was deployed, helping improve safety at Tucson intersections.

By reducing travel times, drivers have also saved over $1.6 million in gas, cutting emissions and improving air quality to make the equivalent impact of planting 650,000 trees.

In Vancouver, Canada, the University of British Columbia (UBC) is using the NoTraffic platform and Rogers Communications’ 5G-connected, AI-enabled smart-traffic platform to reduce both pedestrian delays and greenhouse gas emissions.

Rogers Communications’ 5G networks provide robust and stable connectivity to the sensors embedded on the traffic poles.

This advanced network infrastructure enhances the NoTraffic platform’s efficacy and scalability, as the improved speed and reduced latency of 5G networks means traffic data can be processed in real time. This is critical for predicting numerous potential traffic scenarios, adjusting signal timings and prioritizing road users accordingly.

With AI Mobility deployed at seven intersections across the campus, the university experienced an up to 40% reduction in pedestrian delays and significant decreases in vehicle wait time.

In addition, UBC reduces 74 tons of carbon dioxide emissions each year thanks to the NoTraffic and Rogers solution, which is powered by NVIDIA edge AI and accelerated computing.

The platform is also in action on the roads of Phoenix, Arizona; Baltimore, Maryland; and in 35 states through 200+ agencies across the U.S. and Canada.

Honk If You Love Reducing Congestion, Carbon Emissions

The NoTraffic AI Mobility platform offers local AI-based predictions that, based on sensor inputs at multiple intersections, analyze numerous traffic scenarios up to two minutes in advance.

It can adapt to real-time changes in traffic patterns and volumes, send messages between intersections and run optimization algorithms that control traffic signals to improve overall transportation efficiency and safety through cloud connectivity.

Speedups in the AI Mobility platform mean quicker optimizations of traffic signals — and reduced congestion on the roads means reduced carbon emissions from vehicles.

NoTraffic estimates that for every city optimized with this platform, eight hours of traffic time could be saved per driver. Plus, with over 300,000 signalized intersections in the U.S., the company says this could result in a total of $14 billion in economic savings per year.

Learn more about the NVIDIA Metropolis platform and how it’s used in smart cities and spaces.

Editor’s note: As of June 6, 2025, NVIDIA Edify is no longer available as an NVIDIA NIM microservice preview. To explore available visual AI models, visit build.nvidia.com.

From creating the GPU, RTX real-time ray tracing and neural rendering to now reinventing computing for AI, NVIDIA has for decades been at the forefront of computer graphics — pushing the boundaries of what’s possible in media and entertainment.

NVIDIA Media2 is the latest AI-powered initiative transforming content creation, streaming and live media experiences.

Built on technologies like NVIDIA NIM microservices and AI Blueprints — and breakthrough AI applications from startups and software partners — Media2 uses AI to drive the creation of smarter, more tailored and more impactful content that can adapt to individual viewer preferences.

Amid this rapid creative transformation, companies embracing NVIDIA Media2 can stay on the $3 trillion media and entertainment industry’s cutting edge, reshaping how audiences consume and engage with content.

NVIDIA Technologies at the Heart of Media2

As the media and entertainment industry embraces generative AI and accelerated computing, NVIDIA technologies are transforming how content is created, delivered and experienced.

NVIDIA Holoscan for Media is a software-defined, AI-enabled platform that allows companies in broadcast, streaming and live sports to run live video pipelines on the same infrastructure as AI. The platform delivers applications from vendors across the industry on NVIDIA-accelerated infrastructure.

Delivering the power needed to drive the next wave of data-enhanced intelligent content creation and hyper-personalized media is the NVIDIA Blackwell architecture, built to handle data-center-scale generative AI workflows with up to 25x more energy efficiency over the NVIDIA Hopper generation. Blackwell integrates six types of chips: GPUs, CPUs, DPUs, NVIDIA NVLink Switch chips, NVIDIA InfiniBand switches and Ethernet switches.

Blackwell is supported by NVIDIA AI Enterprise, an end-to-end software platform for production-grade AI. NVIDIA AI Enterprise comprises NVIDIA NIM microservices, AI frameworks, libraries and tools that media companies can deploy on NVIDIA-accelerated clouds, data centers and workstations. Of the expanding list, these include:

• The Llama 3.1-405B-Instruct NIM microservice, which enables synthetic data generation, distillation and inference for chatbots, coding and domain-specific tasks.

• The Mistral-NeMo-12B-Instruct NIM microservice, which enables multilingual information retrieval — the ability to search, process and retrieve knowledge across languages. This is key in enhancing an AI model’s outputs with greater accuracy and global relevancy.

• The NVIDIA Omniverse Blueprint for 3D conditioning for precise visual generative AI, which can help advertisers easily build personalized, on-brand and product-accurate marketing content at scale using real-time rendering and generative AI without affecting a hero product asset.

• NVIDIA NeMo Retriever embedding and reranking NIM microservices, which can vectorize text documents, transcripts, news articles and other written content. Media companies can use these to expand their generative AI efforts and build accurate, multilingual systems.

• The NVIDIA Cosmos Nemotron vision language model NIM microservice, which is a multimodal VLM that can understand the meaning and context of text, images and video. With the microservice, media companies can query images and videos with natural language and receive informative responses.

• The NVIDIA AI Blueprint for video search and summarization (VSS), which integrates VLMs and LLMs and provides cloud-native building blocks to build video analytics, search and summarization applications.

• The NVIDIA Edify multimodal generative AI architecture, which can generate visual assets — like images, 3D models and HDRi environments — from text or image prompts. It offers advanced editing tools and efficient training for developers. With NVIDIA AI Foundry, service providers can customize Edify models for commercial visual services using NVIDIA NIM microservices.

Partners in the Media2 Ecosystem

Partners across the industry are adopting NVIDIA technology to reshape the next chapter of storytelling.

Getty Images and Shutterstock are intelligent content creation services built with NVIDIA Edify. The AI models have also been optimized and packaged for maximum performance with NVIDIA NIM microservices.

Bria is a commercial-first visual generative AI platform designed for developers. It’s trained on 100% licensed data and built on responsible AI principles. The platform offers tools for custom pipelines, seamless integration and flexible deployment, ensuring enterprise-grade compliance and scalable, predictable content generation. Optimized with NVIDIA NIM microservices, Bria delivers faster, safer and scalable production-ready solutions.

Runway is an AI platform that provides advanced creative tools for artists and filmmakers. The company’s Gen-3 Alpha Turbo model excels in video generation and includes a new Camera Control feature that allows for precise camera movements like pan, tilt and zoom. Runway’s integration of the NVIDIA CV-CUDA open-source library combined with NVIDIA GPUs accelerates preprocessing for high-resolution videos in its segmentation model.

Wonder Dynamics, an Autodesk company, recently launched the beta version of Wonder Animation, featuring powerful new video-to-3D scene technology that can turn any video sequence into a 3D-animated scene for animated film production. Accelerated by NVIDIA GPU technology, Wonder Animation provides visual effects artists and animators with an easy-to-use, flexible tool that significantly reduces the time, complexity and efforts traditionally associated with 3D animation and visual effects workflows — while allowing the artist to maintain full creative control.

Comcast’s Sky innovation team is collaborating with NVIDIA on lab testing NVIDIA NIM microservices and partner models for its global platforms. The integration could lead to greater interactivity and accessibility for customers around the world, such as enabling the use of voice commands to request summaries during live sports and access other contextual information.

Vū, a creative technology company and home to the largest network of virtual studios, is broadening access to the creation of virtual environments and immersive content with NVIDIA-accelerated generative AI technologies.

Twelve Labs, a member of the NVIDIA Inception program for startups, is developing advanced multimodal foundation models that can understand videos like humans, enabling precise semantic search, content analysis and video-to-text generation. Twelve Labs uses NVIDIA H100 GPUs to significantly improve the models’ inference performance, achieving up to a 7x improvement in requests served per second.

S4 Capital’s Monks is using cutting-edge AI technologies to enhance live broadcasts with real-time content segmentation and personalized fan experiences. Powered by NVIDIA Holoscan for Media, the company’s solution is integrated with tools like NVIDIA VILA to generate contextual metadata for injection within a time-addressible media store framework — enabling precise, action-based searching within video content.

Additionally, Monks uses NVIDIA NeMo Curator to help process data to build tailored AI models for sports leagues and IP holders, unlocking new monetization opportunities through licensing. By combining these technologies, broadcasters can seamlessly deliver hyper-relevant content to fans as events unfold, while adapting to the evolving demands of modern audiences.

Media companies manage vast amounts of video content, which can be challenging and time-consuming to locate, catalog and compile into finished assets. Leading media-focused consultant and system integrator Qvest has developed an AI video discovery engine, built on NIM microservices, that accelerates this process by automating the data capture of video files. This streamlines a user’s ability to both discover and contextualize how videos can fit in their intended story.

Verizon is transforming global enterprise operations, as well as live media and sports content, by integrating its reliable, secure private 5G network with NVIDIA’s full-stack AI platform, including NVIDIA AI Enterprise and NIM microservices, to deliver the latest AI solutions at the edge.

Using this solution, streamers, sports leagues and rights holders can enhance fan experiences with greater interactivity and immersion by deploying high-performance 5G connectivity along with generative AI, agentic AI, extended reality and streaming applications that enable personalized content delivery. These technologies also help elevate player performance and viewer engagement by offering real-time data analytics to coaches, players, referees and fans. It can also enable private 5G-powered enterprise AI use cases to drive automation and productivity.

Welcome to NVIDIA Media2

The NVIDIA Media2 initiative empowers companies to redefine the future of media and entertainment through intelligent, data-driven and immersive technologies — giving them a competitive edge while equipping them to drive innovation across the industry.

NIM microservices from NVIDIA and model developers are now available to try, with additional models added regularly.

Get started with NVIDIA NIM and AI Blueprints, and watch the CES opening keynote delivered by NVIDIA founder and CEO Jensen Huang to hear the latest advancements in AI.

See notice regarding software product information.

Vay, a Berlin-based provider of automotive-grade remote driving (teledriving) technology, is offering an alternative approach to autonomous driving.

Through the company’s app, a user can hail a car, and a professionally trained teledriver will remotely drive the vehicle to the customer’s location. Once the car arrives, the user manually drives it.

After completing their trip, the user can end the rental in the app and pull over to a safe location to exit the car, away from traffic flow. There’s no need to park the vehicle, as the teledriver will handle the parking or drive the car to the next customer.

This system offers sustainable, door-to-door mobility, with the unique advantage of having a human driver remotely controlling the vehicle in real time.

Vay’s technology is built on the NVIDIA DRIVE AGX centralized compute platform, running the NVIDIA DriveOS operating system for safe, AI-defined autonomous vehicles.

These technologies enable Vay’s fleets to process large volumes of camera and other vehicle data over the air. DRIVE AGX’s real-time, low-latency video streaming capabilities provide enhanced situational awareness for teledrivers, while its automotive-grade design ensures reliability in any driving condition.

“By combining Vay’s innovative remote driving capabilities with the advanced AI and computing power of NVIDIA DRIVE AGX, we’re setting a new standard for remotely driven vehicles,” said Justin Spratt, chief business officer at Vay. “This collaboration helps us bring safe, reliable and accessible driverless options to the market and provides an adaptable solution that can be deployed in real-world environments now — not years from now.”

High-Quality Video Stream

Vay’s advanced technology stack includes NVIDIA DRIVE AGX software that’s optimized for latency and processing power. By harnessing NVIDIA GPUs specifically designed for autonomous driving, the company’s teledriving system can process and transmit high-definition video feeds in real time, delivering critical situational awareness to the teledriver, even in complex environments. In the event of an emergency, the vehicle can safely bring itself to a complete stop.

“Working with NVIDIA, Vay is setting a new standard in driverless technology,” said Bogdan Djukic, cofounder and vice president of engineering, teledrive experience and autonomy at Vay. “We are proud to not only accelerate the deployment of remotely driven and autonomous vehicles but also to expand the boundaries of what’s possible in urban transportation, logistics and beyond — transforming mobility for both businesses and communities.”

Reshaping Mobility With Teledriving

Vay’s technology enables professionally trained teledrivers to remotely drive vehicles from specialized teledrive stations equipped with industry-standard controls, such as a steering wheel and pedals.

The company’s teledrivers are totally immersed in the drive — road traffic sounds, such as those from emergency vehicles and other warning signals, are transmitted via microphones to the operator’s headphones. Camera sensors reproduce the car’s surroundings and transmit them to the screens of the teledrive station with minimum latency. The vehicles can operate at speeds of up to 26 mph.

Vay’s technology effectively addresses complex edge cases with human supervision, enhancing safety while significantly reducing costs and development challenges.

Vay is a member of NVIDIA Inception, a program that nurtures AI startups with go-to-market support, expertise and technology. Last year, Vay became the first and only company in Europe to teledrive a vehicle on public streets without a safety driver.

Since January, Vay has been operating its commercial services in Las Vegas. The startup recently secured a partnership with Bayanat, a provider of AI-powered geospatial solutions, and is working with Ush and Poppy, Belgium-based car-sharing companies, as well as Peugeot, a French automaker.

In October, Vay announced a $34 million investment from the European Investment Bank, which will help it roll out its technology across Europe and expand its development team.

Learn more about the NVIDIA DRIVE platform.

Accelerated computing is supercharging AI and data processing workloads, helping enterprises across industries achieve greater efficiency with reduced time and costs.

For over a decade, NVIDIA has worked with Amazon Web Services (AWS) to bring accelerated computing to businesses and developers around the world.

At AWS re:Invent 2024, taking place Dec. 2-6 in Las Vegas, NVIDIA’s full-stack offerings will be on display. Attendees can take a deep dive into the broad range of NVIDIA hardware and software platforms available on AWS and learn how partners and customers use them to accelerate their most compute-intensive workloads.

Highlights from the session catalog include:

• “NVIDIA Accelerated Computing Platform on AWS” with Dave Salvator, director of accelerated computing products at NVIDIA (AIM110-S)
• “Build, Customize and Deploy Generative AI With NVIDIA on AWS” with Abhishek Sawarkar, product manager at NVIDIA, and Charlie Huang, senior product marketing at NVIDIA (AIM241-S)
• “Advancing Physical AI: NVIDIA Isaac Lab and AWS for Next-Gen Robotics” with Rishabh Chadha, technical marketing engineer at NVIDIA; Abhishek Srivastav, senior solutions architect at AWS; and Shaun Kirby, principal enterprise architect at AWS (AIM113-S)
• “NVIDIA AI Startups: Innovations in Action” with Jen Hoskins, global head of Inception cloud partnerships and go-to-market at NVIDIA, and speakers from Inception startups, including Bria, Contextual AI, Hippocratic AI, Mendel AI, Twelve Labs and Writer (AIM121-S)
• “AI-Driven Value: Capital One’s Path to Better Customer Experience” with Joey Conway, senior director of product management for large language model software at NVIDIA, and Prem Natarajan, chief scientist and head of enterprise AI at Capital One (AIM130-S)
• “Accelerate Apache Spark Up to 5 Times on AWS With RAPIDS” with Sameer Raheja, senior director of engineering at NVIDIA (ANT208-S)

For a more hands-on experience, join an AWS Jam session and workshops:

• AWS Jam: Building a RAG Chat Agent With NVIDIA NIM (GHJ305)
• Robotic Simulation With NVIDIA Isaac Lab on AWS Batch (MFG319)
• Unleash Edge Computing With AWS IoT Greengrass on NVIDIA Jetson (IOT316)
• Building Scalable Drug Discovery Applications With NVIDIA BioNeMo (HLS205)
• Creating Immersive 3D Digital Twins From Photos, Videos and LiDAR With NVIDIA Omniverse (CMP315)

NVIDIA booth 1620 will feature a variety of demos, including a full NVIDIA GB200 NVL72 rack, coming soon to Amazon Elastic Compute Cloud (Amazon EC2) and NVIDIA DGX Cloud, as well as Spot, an agile mobile robot from Boston Dynamics.

Other demos showcasing the NVIDIA platform on AWS include:

• Powering Digital Twins and Physical AI With NVIDIA Omniverse
• Deploying Generative AI Faster With NVIDIA NIM
• Speed Deployment of AI With NVIDIA AI Blueprints, Including Generative Virtual Screening for Accelerated Drug Discovery
• The NVIDIA Accelerated Computing Platform on AWS, Hardware Show-and-Tell
• Fraud Prevention Reference Architecture on RAPIDS With AWS

NVIDIA will also feature demos from select partners and customers, including startups Balbix, Bria, Mendel AI, Standard Bots, Union and Writer.

Attendees can learn more about NVIDIA’s full-stack accelerated computing platform on AWS, including three new Amazon EC2 instance types released this year: Amazon EC2 P5e instances (NVIDIA H200 Tensor Core GPUs) for large-scale AI training and inference, G6e instances (NVIDIA L40S GPUs) for AI and graphics, and G6 instances (NVIDIA L4 Tensor Core GPUs) for small model deployments.

Plus, discover how NVIDIA’s GPU-optimized software stack delivers high performance across AWS services, making it easy for developers to accelerate their applications in the cloud. Some examples include:

• Generative AI: NVIDIA NeMo and NIM microservices, part of the NVIDIA AI Enterprise software platform and available in AWS Marketplace, for building, customizing and deploying generative AI on Amazon EC2, Amazon SageMaker and Amazon Elastic Kubernetes Service (EKS).
• Digital Twins and Physical AI: NVIDIA Earth-2 weather simulation on AWS and NVIDIA Isaac Sim and Isaac Lab for robot simulation and learning using AWS Batch.
• Healthcare: NVIDIA BioNeMo for drug discovery applications, including an NVIDIA AI Blueprint for virtual screening and NIM microservices for protein prediction and small molecule generation, supported on AWS Healthomics.
• Data Processing: RAPIDS Accelerator for Apache Spark for accelerating analytic and machine learning workloads, available on Amazon EMR, Amazon EC2 and Amazon EMR on EKS.
• Edge AI: NVIDIA Jetson with AWS IoT Greengrass for streamlined experiences to deploy generative AI models and manage fleets of connected edge devices and robots.

Members of the NVIDIA Inception program for cutting-edge startups are already testing, developing and deploying their most challenging workloads using the NVIDIA platform on AWS:

• Twelve Labs achieved up to a 7x inference improvement in requests served per second when upgrading to NVIDIA H100 Tensor Core GPUs. Its Marengo and Pegasus models, soon available as NIM microservices, power video Al solutions that enable semantic search on embedded enterprise video archives.
• Wiz doubled inference throughput speed for DSPM data classification using NIM microservices over alternatives.
• Writer achieved 3x faster model iteration cycles using SageMaker HyperPod with NVIDIA H100 GPUs. With NVIDIA accelerated computing and AWS, Writer optimized the training and inference of its Palmyra models, significantly reducing time to market for its customers.

Inception helps startups evolve faster by offering the latest NVIDIA technologies, opportunities to connect with venture capitalists, and access to technical resources and experts.

Register for AWS re:Invent to see how businesses can speed up their generative AI and data processing workloads with NVIDIA accelerated computing on AWS.

Send an email to schedule a meeting with NVIDIA experts at the show.