journeys round nvidia world

 Agentic; Reasoning; Twin; Sovereignty; 25x energy; deep

AGENTIC . Agentic AI is being used to transform various industries, including customer service, cybersecurity, retail, automotive, financial services, and healthcare, with companies like Accenture, Meta, Microsoft, and ServiceNow discussing its applications and advancements.  6  15  16  The NVIDIA Agentic Eval is a tool that supports the evaluation of custom data using Ragas metrics, including topic_adherence, tool_call_accuracy, agent_goal_accuracy, and answer_accuracy, and can be run using NVIDIA's Eval Factory framework or the native CLI tool.  1  2  3  NVIDIA NeMo Microservices supports Agentic evaluation types, which assess the performance of agent-based models in scenarios requiring planning, tool use, and iterative reasoning, with options including Topic Adherence, Tool Call Accuracy, Agent Goal Accuracy, and Answer Accuracy.  2  3  5  Agentic AI systems can perceive, reason, and act autonomously, and are being leveraged by leading organizations to accelerate software patching and vulnerability management, with NVIDIA products such as the NVIDIA AI Blueprint, NVIDIA NIM, and NVIDIA Morpheus being used to advance agentic AI in cybersecurity.  14  16  19  The implementation of Agentic AI within an Enterprise AI Factory is being discussed, focusing on components such as the Enterprise Cloud Native Platform with Kubernetes, storage solutions, artifact repositories, and security, with NVIDIA products like NVIDIA AI Enterprise, NVIDIA NeMo, and NVIDIA NIM being used to create a secure, scalable, and high-performance AI environment.  17  18  19	REASONING AI reasoning is the ability of artificial intelligence systems to analyze and solve complex problems through multiple AI inference passes, mimicking human problem-solving, and is crucial for generative AI, bridging the gap between pattern recognition and decision-making.  1  NVIDIA provides various solutions and models to support AI reasoning, including the NVIDIA GB300 NVL72, NVIDIA Llama Nemotron, DeepSeek-R1, and OpenAI o1 and o3-mini, which can tackle complex problems more effectively and reliably.  1  2  11  The NVIDIA NeMo Microservices platform and NVIDIA NIM for Large Language Models (LLMs) provide tools and APIs for managing, deploying, and configuring AI models, including reasoning models, with features such as data generation, evaluation, inference, and guardrails.  3  4  12  Reasoning models can generate detailed, step-by-step thought processes, controlled by system prompts, and can be used in various applications, including RAG pipelines, AI agents, and collaborative multi-agent systems, to tackle complex tasks and provide transparent and auditable reasoning.  8  14  18  NVIDIA has introduced several reasoning models, including the Llama Nemotron Ultra Open Model, which delivers groundbreaking reasoning accuracy, and the Nemotron-H Reasoning Model Family, which offers significant inference-time speedups and supports 128K token contexts.  11  19  related "ai systems"  "pattern recognition" "ai health" "decision making" .	.TWINS Digital twins are virtual representations of products, processes, and facilities used to design, simulate, and operate their physical counterparts, with various use cases including product development, architectural design, and industrial facility management.  2  4  7  NVIDIA's Omniverse platform enables the creation of digital twins, which can be used to optimize industrial operations, enhance customer experiences, and develop adaptable robots, with companies like BMW Group, Delta Electronics, and Pegatron using NVIDIA's solutions to digitalize their physical world and improve efficiency.  2  7  11  Digital twins have various benefits, including cost savings, improved decision-making, and enhanced customer experience, and are being used by companies like Foxconn, Siemens, and Wistron to transform their facilities and products, with NVIDIA providing developer tools, learning labs, and reference architectures to help build and deploy industrial facility digital twin applications.  4  11  12  The use of digital twins is being expanded to various industries, including healthcare, where companies like Foxconn are using NVIDIA Omniverse and Isaac Sim to simulate robots assisting nurses and create disease digital twins, and preserving historical monuments, where a digital twin of the Garisenda Tower in Italy was built using NVIDIA Omniverse to monitor its stability and safety.  8  10  15  NVIDIA is continuously expanding its ecosystem and tools for digital twins, including the Omniverse Blueprint, which enables engineering teams to design, simulate, and optimize entire AI factories in virtual environments, and the Digital Twins for Physical AI learning path, which provides a free introduction to developing OpenUSD-based digital twin applications and workflows.  14  19  20
Sovereignty.   AI-generated summary of your search results Sovereign AI refers to a nation's domestic capability to produce artificial intelligence, which is critical for economic growth, national security, and cultural preservation. Many countries, including the UK, France, Italy, Denmark, Japan, India, and Brazil, are developing domestic AI capabilities to drive economic growth.  9  13  To achieve sovereign AI, countries are building AI factories, which are next-generation data centers that produce tokens for intelligent outcomes, and are transforming raw data into actionable insights. These AI factories are being powered by companies like NVIDIA, which is providing accelerated computing infrastructure and software stacks.  11  14  16  The development of sovereign AI is also being driven by the need for secure and private data processing, as well as the need for local control over AI models. Companies like NVIDIA are providing solutions such as NVIDIA AI Enterprise and NVIDIA NIM microservices to support the development of sovereign AI.  10  13  16  In addition to national governments, telcos are also playing a key role in the development of sovereign AI, by providing secure infrastructure for regional AI model development and deployment. Telcos such as Indosat Ooredoo Hutchison, Orange, and Telefónica are working with companies like NVIDIA to build AI factories and provide sovereign AI solutions.  7  11  The European Union is investing heavily in AI, with over $200 billion being invested, and companies like NVIDIA are unveiling new technologies to address key constraints in AI development, including limited compute availability, data-privacy needs, and safety priorities  .	.Deep	.Energy 25X

x

• NVIDIA Agentic Eval is a factory-compatible container that supports the evaluation of custom data using Ragas metrics, including topic_adherence, tool_call_accuracy, agent_goal_accuracy, and answer_accuracy. The evaluation can be run using NVIDIA's Eval Factory framework or the native CLI tool, agentic_eval. The tool requires a judge model, which can be set up using OpenAI or NVIDIA NIM, and supports various configuration parameters, such as temperature, parallelism, and max_new_tokens. The results are saved in two files, <metric_name>.jsonl and scores.jsonl, which contain scores for each data entry and the final score for all the data, respectively. The tool also supports dynamic data templates and allows users to customize the data format. The Ragas API provides more information on the schema and metrics used in the evaluation.

  https://catalog.ngc.nvidia.com/orgs/nvidia/teams/eval-factory/containers/agentic_eval
• Agentic Evaluation Types — NVIDIA NeMo Microservices

  NVIDIA NeMo Microservices supports Agentic evaluation types, which assess the performance of agent-based models in scenarios requiring planning, tool use, and iterative reasoning. Agentic evaluation types include Topic Adherence, Tool Call Accuracy, Agent Goal Accuracy with and without Reference, and Answer Accuracy. Each type has its own configuration and data format requirements, and the judge model must be at least 70B parameters. The evaluation uses RAGAS metrics to score agent outputs, with metrics such as topic adherence, tool call accuracy, agent goal accuracy, and answer accuracy. Limitations include only working with cached_outputs targets and requiring different dataset formats for different metrics.

  https://docs.nvidia.com/nemo/microservices/25.7.0/evaluate/evaluation-types/agentic.html
• • Agentic AI Leaders to Showcase Latest Advancements at NVIDIA GTC | February 2025

    At the NVIDIA GTC conference in San Jose, California, industry experts from companies like Accenture, Meta, Microsoft, and ServiceNow will discuss the latest advancements in agentic AI, including its applications in customer service, cybersecurity, retail, automotive, financial services, and healthcare. NVIDIA founder and CEO Jensen Huang will deliver a keynote speech, and other notable speakers include David Reber, Adel El Hallak, and Bartley Richardson from NVIDIA, as well as executives from Deloitte, Yum! Brands, U.S. Bank, Mayo Clinic, Ford Motor Company, and Capital One. The conference will feature sessions on building custom reasoning models, deploying agentic AI systems, and implementing AI agents in various industries, with a focus on NVIDIA products like NVIDIA AI Blueprints and the NVIDIA AI Enterprise software platform.

    https://blogs.nvidia.com/blog/agentic-ai-gtc-2025/
  • [7] 
    Agentic Cloud: Using Agents to Build Tomorrow’s Cloud (Presented by DigitalOcean) | GTC 2025

    DigitalOcean's Bratin Saha discussed using agents to improve cloud experiences with generative AI. Agents allowed customization for customers and simplified AI use. The company deployed agents at scale and operationalized NVIDIA H100 and H200 GPUs for model training, inference, and natural language interfaces.

    https://www.nvidia.com/en-us/on-demand/session/gtc25-exs74250/
  • [8] 
    Fundamentals of Agentic AI: A Brief Introduction | GTC 2025

    Vadim Kudlay, Sr. Content Developer at NVIDIA's Deep Learning Institute, discussed the fundamentals of agentic AI, breaking down the agent abstraction and highlighting key insights driving the space forward, referencing the course on Building Agentic AI Applications with LLMs.

    https://www.nvidia.com/en-us/on-demand/session/gtc25-dlit74813/
  • [9] 
    Agentic Architecture and Digital Human Capabilities for Complex Customer Support | GTC 2025

    Jagan Mohan Kaliamurthy, Srinivasan Saravanamuthu, and Yarmon Kennedy discussed agentic architecture and digital human capabilities for complex customer support. They demonstrated a business use case with a retirement plan provider, showcasing a virtual advisor and AI agent built jointly by Accenture and NVIDIA teams to improve the participant experience.

    https://www.nvidia.com/en-us/on-demand/session/gtc25-s72752/
  • [10] 
    How to Build an Agentic AI System Using the Best Tools and Frameworks | GTC Paris 2025

    NVIDIA's GTC Paris 2025 session discussed building agentic AI systems using tools and frameworks, connecting AI agents to reusable tools to unlock data and drive efficiency gains, helping enterprises thrive in the AI era.

    https://www.nvidia.com/en-us/on-demand/session/gtcparis25-gp1075/

  • Try These Search Refinements

    multi-agent systems

    artificial intelligence

    transparent ai

    ai reasoning

    physical ai

    1-10 of 100

    ---

    • [1] 
      What are Multi-Agent Systems?

      Multi-agent systems are collections of specialized AI agents that work together to solve complex problems, with each agent having a specific role and autonomy. These systems can be used in various applications, such as end-to-end software development, customer service, and supply chain management. NVIDIA provides AI software solutions, including the NVIDIA NeMo Agent toolkit, to help build and deploy multi-agent systems. The company also offers AI Blueprints, which provide a starting point for developing agents to address specific use cases, such as retrieval-augmented generation (RAG).

      https://www.nvidia.com/en-us/glossary/multi-agent-systems/
    • [2] 
      What is Artificial Intelligence and Why Does it Matter?

      Artificial Intelligence (AI) is transforming the global economy, with potential contributions of nearly $17 trillion in ten years, according to PwC. AI systems, powered by graphical processing units (GPUs) like those from NVIDIA, can perform tasks autonomously, learning from vast amounts of data. NVIDIA's GPUs, including the T4 and V100, are used in various industries such as healthcare, retail, telecommunications, and financial services, to improve operations, predict demand, and detect anomalies. NVIDIA's platforms, including DGX-2 and DGX A100, enable accelerated computing for AI workloads, while the EGX platform brings AI performance to the edge.

      https://www.nvidia.com/en-us/glossary/artificial-intelligence/
    • [3] 
      Designing Transparent AI Systems for Scientific Discovery: Patterns and Practices for Human-AI Collaboration | GTC 2025

      Sutanay Choudhury discussed designing transparent AI systems for scientific discovery, exploring architectural patterns for human-AI collaboration, and enabling scientists to evaluate AI's analytical process using NVIDIA's technology was not mentioned, instead the talk focused on general AI systems.

      https://www.nvidia.com/en-us/on-demand/session/gtc25-s72049/
    • [4] 
      NVIDIA Unveils AI-Q Blueprint to Connect AI Agents for the Future of Work | April 2025

      NVIDIA has unveiled the AI-Q Blueprint, a framework for developing agentic systems that enables AI agents to collaborate and solve complex problems. The blueprint integrates NVIDIA accelerated computing, partner storage platforms, and software tools, including the NVIDIA Llama Nemotron reasoning models and NVIDIA NeMo Retriever. The AI-Q Blueprint is powered by the NVIDIA Agent Intelligence (AIQ) toolkit, an open-source software library that provides unified monitoring and optimization of agentic systems. The toolkit is available on GitHub and can be easily integrated with existing multi-agent systems, including Salesforce's Agentforce, Atlassian Rovo, and ServiceNow AI platform. Companies like Visa are using AI agents to streamline tasks, such as automating phishing email analysis, and the AIQ toolkit to optimize agent performance. Developers can use the Agent Intelligence toolkit to build hands-on skills and advance agentic systems, and NVIDIA partners are delivering customized AI data platforms to support these systems.

      https://blogs.nvidia.com/blog/ai-agents-blueprint/
    • [5] 
      NVIDIA’s Bartley Richardson on How Teams of AI Agents Provide Next-Level Automation | May 2025

      Bartley Richardson, senior director of engineering and AI infrastructure at NVIDIA, discusses how enterprises can deploy agentic AI systems, which provide next-level automation. He explains that AI reasoning models, like NVIDIA's Llama Nemotron models, play a critical role in these systems by enabling better planning capabilities. Richardson also highlights the importance of acknowledging the multi-vendor reality of modern environments and developing systems that can work together seamlessly. To address this challenge, NVIDIA developed the AI-Q Blueprint, which uses the open-source NVIDIA Agent Intelligence (AIQ) toolkit to evaluate and profile agent workflows. Richardson emphasizes the importance of maintaining realistic expectations and notes that agentic systems will make mistakes, but can still provide significant business value, citing customers who have achieved 15x speedups using AI-Q. Other NVIDIA experts, such as Rama Akkiraju and Jacob Liberman, also discuss the role of AI in enterprises, including the importance of AI platform architects and agentic AI in bridging the gap between powerful AI models and practical applications.

      https://blogs.nvidia.com/blog/ai-agents-next-level-automation/
    • [6] 
      What is Physical AI?

      Physical AI enables autonomous systems like robots, self-driving cars, and smart spaces to perceive, understand, and perform complex actions in the real world. It extends generative AI with an understanding of spatial relationships and physical behavior through simulations and reinforcement learning. NVIDIA provides tools such as Omniverse, DGX, and Isaac Sim to support physical AI development, training, and deployment. Physical AI has various applications, including robots, autonomous vehicles, and smart spaces, and is transforming industries with its ability to generate insights and actions.

      https://www.nvidia.com/en-us/glossary/generative-physical-ai/
    • [7] 
      What are World Foundation Models?

      World Foundation Models (WFMs) are neural networks that simulate real-world environments, predicting accurate outcomes based on input data. They are used in autonomous vehicles, robotics, and video analytics, and are built using extensive real-world data and neural networks with billions of parameters. WFMs can be fine-tuned for specific applications and can be used with 3D simulators to safely train autonomous machines. NVIDIA offers platforms such as NVIDIA Cosmos and NVIDIA Isaac GR00T to accelerate the development of physical AI systems using WFMs.

      https://www.nvidia.com/en-us/glossary/world-modelsTransparency AI
    • Transparency AI

    • NVIDIA is committed to developing Trustworthy AI, adhering to guiding principles of privacy, safety and security, transparency, and nondiscrimination, and has introduced various technologies and initiatives to promote AI transparency, including Model Card++ and NVIDIA Exemplar Clouds  1  4  5  6 
    • Techniques such as Chain of Thought (CoT) prompting and Explainable AI (XAI) are being explored to enhance AI transparency and trust, with CoT prompting improving problem-solving, transparency, and versatility, and XAI providing practical methods to bring interpretability into developing and deploying AI systems  3  7 
    • NVIDIA has also introduced the Llama Nemotron reasoning model family, which delivers exceptional reasoning capabilities and compute efficiency, and provides transparent and auditable reasoning, making it a valuable tool for enterprises looking to build advanced agentic AI systems  13  16 
    • Additionally, NVIDIA AI Enterprise provides a range of tools and services to support AI transparency, including the RAPIDS Accelerator for Apache Spark, which leverages GPUs to accelerate processing via RAPIDS libraries, and NVIDIA NIM microservices, which enable enterprises to deploy generative AI models securely on their own terms  17  18  19  15 

    AI-generated content may be incomplete. Verify important info. Learn more

    Built with Llama (Llama 3.3 & Llama 3.2 Community License Agreements).

    Try These Search Refinements

    explainable ai

    model transparency

    ai safety

    human ai collaboration

    cloud infrastructure

    1-10 of 64

    ---

    • [1] 
      How to Automate AI Transparency: Enhancing Model Transparency With Minimal Effort | GTC 2025

      Arun George Zachariah, Michael Boone, and Pratyusha Maiti from NVIDIA, along with Paul Taylor from IBM, discussed automating AI transparency. They explained how the model card generator streamlined model card creation and management, capturing metadata during training and from source code, ensuring comprehensive documentation for AI models.

      https://www.nvidia.com/en-us/on-demand/session/gtc25-s72399/
    • [2] 
      Designing Transparent AI Systems for Scientific Discovery: Patterns and Practices for Human-AI Collaboration | GTC 2025

      Sutanay Choudhury discussed designing transparent AI systems for scientific discovery, exploring architectural patterns for human-AI collaboration, and enabling scientists to evaluate AI's analytical process using NVIDIA's technology was not mentioned, instead the talk focused on general AI systems.

      https://www.nvidia.com/en-us/on-demand/session/gtc25-s72049/
    • [3] 
      Crack the AI Black Box: Practical Techniques for Explainable AI (Presented by DigitalOcean) | GTC 2025

      David vonThenen, Sr. Developer Advocate II, AI/ML at DigitalOcean, presented techniques for Explainable AI (XAI) to bridge the trust and transparency gap in Artificial Intelligence. He explored foundational XAI concepts and provided practical methods to bring interpretability into developing and deploying AI systems. A live demonstration showed how to apply these methods to real-world problems, enabling diagnosis, debugging, and optimization of models. The presentation aimed to provide a roadmap for integrating XAI practices into workflows to build trust and confidence in AI-powered solutions, including using feature attributions and decision-path analyses.

      https://www.nvidia.com/en-us/on-demand/session/gtc25-s74147/
    • [4] 
      Conclusion — NVIDIA AI Enterprise Security White Paper

      NVIDIA's enterprise security strategy is built on understanding the complexities of accelerated AI software stacks, providing security and reliability through best-in-class vulnerability scanning, patch management, and transparency for AI enterprise container images, as outlined in the NVIDIA AI Enterprise Security White Paper.

      https://docs.nvidia.com/ai-enterprise/planning-resource/ai-enterprise-security-white-paper/latest/conclusion.html
    • [5] 
      Announcing NVIDIA Exemplar Clouds for Benchmarking AI Cloud Infrastructure | May 2025

      NVIDIA has launched NVIDIA Exemplar Clouds, an initiative designed to bring transparency and rigor to AI cloud infrastructure benchmarking, starting with the NVIDIA Cloud Partner ecosystem. This initiative introduces frameworks that evaluate cloud providers' actual performance and resiliency, not just theoretical specs. NVIDIA Exemplar Clouds provides a comprehensive evaluation process, open benchmarking recipes, and transparent results through NVIDIA DGX Cloud Benchmarking, enabling customers to make informed decisions. Yotta is the first APAC cloud provider to join the initiative, and NVIDIA will work with them to demonstrate high performance and resiliency. NVIDIA founder and CEO Jensen Huang will discuss this further at the COMPUTEX 2025 keynote, and attendees can learn more at GTC Taipei sessions. The initiative aims to democratize access to world-class AI infrastructure and help users build and deploy AI with confidence, with abhishsinha as a key speaker.

      https://developer.nvidia.com/blog/announcing-nvidia-exemplar-clouds-for-benchmarking-ai-cloud-infrastructure/
    • [6] 
      Trustworthy AI For A Better World

      NVIDIA is committed to developing Trustworthy AI, adhering to guiding principles of privacy, safety and security, transparency, and nondiscrimination. The company is building technologies such as data factories for generative AI, tools for unbiased datasets, and security features like confidential computing. NVIDIA's Trustworthy AI Solutions include Model Card++, NVIDIA Halos for AI safety, and NeMo Guardrails for large language models. The company is partnering with organizations like Te Hiku Media, Google DeepMind, and Adobe to advance Trustworthy AI, and is focused on AI research, including publications on AI risk assessment, debiasing, and social bias testing. NVIDIA's commitment to Trustworthy AI is in line with the White House Voluntary Commitments and other global AI Safety initiatives, and will be further discussed at the GTC AI Conference.

      https://www.nvidia.com/en-us/trust-center/trustworthy-ai/
    • [7] 
      What is Chain of Thought (CoT) Prompting? - What Is Chain of Thought Prompting?

      Chain of Thought (CoT) prompting is a technique that enhances large language models' reasoning capabilities by asking them to show their work, similar to human reasoning. It works by providing an initial question and answer, establishing a pattern, and then using this pattern to generate answers to subsequent questions. CoT prompting is important as it improves problem-solving, transparency, and versatility, and has been applied to various tasks, including mathematics, common-sense reasoning, and complex decision-making. NVIDIA provides information and resources on CoT prompting, including its application in AI reasoning models, such as long-thinking or test-time scaling models, and its potential to drive advancements in healthcare, robotics, and finance.

      https://www.nvidia.com/en-us/glossary/cot-prompting/
    • [8] 
      Conclusion — NVIDIA AI Enterprise Security for Azure AI Foundry

      NVIDIA's enterprise security strategy focuses on addressing the complexities of accelerated AI software stacks, ensuring security and reliability of AI enterprise container images through vulnerability scanning, patch management, and transparency. The company collaborates with Microsoft to provide secure and stable software within Azure AI Foundry, offering production-grade solutions as part of NVIDIA AI Enterprise Security for Azure AI Foundry.

      https://docs.nvidia.com/ai-enterprise/planning-resource/security-for-azure-ai-foundry/latest/conclusion.html
    • [9] 
      European Broadcasting Union and NVIDIA Partner on Sovereign AI to Support Public Broadcasters | June 2025

      NVIDIA and the European Broadcasting Union (EBU) are partnering to advance sovereign AI for European public service media, aiming to create a more equitable and resilient media landscape. The collaboration, announced at NVIDIA GTC Paris, will focus on building sovereign AI and cloud frameworks, driving workforce development, and cultivating an AI ecosystem. The EBU's chief technology officer, Michael Eberhard, and chief technology and innovation officer, Antonio Arcidiacono, expressed their support for the partnership, which will utilize NVIDIA products such as NVIDIA AI Enterprise, NVIDIA GB200 Grace Blackwell Superchip, NVIDIA RTX PRO Servers, NVIDIA DGX Cloud, and NVIDIA Holoscan for Media. The partnership will also provide public service media companies with tools for personalized services and content recommendation systems, with a focus on transparency and cultural identity. NVIDIA founder and CEO Jensen Huang discussed the collaboration during the NVIDIA GTC Paris keynote at VivaTech.

      https://blogs.nvidia.com/blog/european-broadcasting-union-sovereign-ai/
    • • [1] 
        Why Does My AI Strategy Need to Include Data Sovereignty? (Presented by HPE) | GTC Paris 2025

        HPE presented a session at GTC Paris 2025, discussing the importance of data sovereignty in AI strategy, covering control over data and operations, regulatory compliance, and building a secure AI-accelerated cloud platform, with a customer sharing their experience using a data flywheel to achieve sovereignty.

        https://www.nvidia.com/en-us/on-demand/session/gtcparis25-gps1077/
      • [2] 
        Innovation While Addressing Sovereignty Needs: Security and Resilience at Scale (Presented by Amazon Web Services) | GTC Paris 2025

        Bruno Raillard, Sasha Rubel, and Eiso Kant discussed how NVIDIA AI and AWS cloud technologies empowered local large language model development, supported secure data practices, and fueled agile AI experimentation at scale. They explored practical examples from across Europe, illustrating how democratized cloud access contributed to the continent's digital autonomy, economic competitiveness, and strategic resilience. The panel highlighted the importance of security, compliance, sovereignty, and flexibility in realizing the full potential of AI and driving innovation responsibly.

        https://www.nvidia.com/en-us/on-demand/session/gtcparis25-gps1070/
      • [3] 
        Driving European Sovereign AI Priorities: From Vision to Leadership | GTC Paris 2025

        Karl Havard, Thomas Jarzombek, Bea Longworth, and Audrey Herblin-Stoop discussed Europe's vision to become a leading AI continent, exploring best practices to deliver AI at scale. They examined critical enablers of sovereignty, including computing infrastructure and regulatory clarity, grounded in the European Commission's AI Continent Action Plan. The panel assessed Europe's strengths, such as public investment in AI Factories and the open-source ecosystem, and addressed gaps in private sector uptake and talent scalability, discussing how to accelerate Europe's leadership in trustworthy AI with NVIDIA and other companies like Mistral AI and NScale.

        https://www.nvidia.com/en-us/on-demand/session/gtcparis25-gp1043/
      • [4] 
        AI Sovereign – Design, Build, and Operate AI Factories (Presented by Vertiv) | GTC Paris 2025

        Industry leaders from Vertiv, NVIDIA, and Domyn discussed their collaboration on a large sovereign AI data center. Kresimir Krpan, Robby van Roosmalen, and Antoine Perchet explored how NVIDIA's accelerated computing, Vertiv's infrastructure, and NVIDIA Omniverse technologies are transforming AI deployment. They covered the modular design of NVIDIA DGX SuperPOD, digital twin technology, and balancing computational power with data sovereignty and security. The discussion also included the role of NVIDIA Omniverse in accelerating design-to-deploy cycles and scaling AI infrastructure for regulated industries.

        https://www.nvidia.com/en-us/on-demand/session/gtcparis25-gps1052/
      • [5] 
        Building Enterprise Sovereign AI: The Future of Data Sovereignty and Secure AI Infrastructure — A Testimonial Journey (Presented by PNY Technologies) | GTC 2025

        David Chanial, Nicholas Dimotakis, and Taher Behbehani discussed designing sovereign AI data centers, emphasizing national security and data governance, with PNY Technologies sharing their experience in optimizing AI workloads with NVIDIA technology and investing in energy infrastructure.

        https://www.nvidia.com/en-us/on-demand/session/gtc25-s74196/
      • [6] 
        Europe’s AI Race: The Business and Tech Behind Sovereign AI (Presented by Nscale) | GTC Paris 2025

        Josh Payne and David Power of Nscale discussed Europe's opportunity to lead in building sovereign AI infrastructure, shaped by its regulatory, cultural, and economic context. They explored the importance of sovereign control over data, compute, and governance, and highlighted how organizations are rethinking AI procurement and infrastructure strategies. The session also examined the building of sovereign AI superclusters, including data center design and secure deployment, and how Nscale is delivering high-performance infrastructure grounded in European values of transparency, sustainability, and digital trust, leveraging NVIDIA technology.

        https://www.nvidia.com/en-us/on-demand/session/gtcparis25-gps1114/
      • [7] 
        How Indonesia Delivered a Telco-led Sovereign AI Platform for 270M Users | GTC 2025

        Anissh Pandey, Lilach Ilan, Munjal Shah, Senthil Ramani, and Vikram Sinha discussed Indonesia's delivery of a telco-led sovereign AI platform for 270M users. Indosat Ooredoo Hutchison deployed a Sovereign AI Factory and large language models, including Sahabat.ai, leveraging the NVIDIA Cloud Partner program, NVIDIA NEMO, and NIMs. Accenture and Hippocratic AI accelerated AI use-case deployment in critical sectors like banking and healthcare using the NVIDIA AI Platform. The platform enabled Indonesia's universities, startups, enterprises, and government agencies to create AI, participating in the new global economy.

        https://www.nvidia.com/en-us/on-demand/session/gtc25-s73440/
      • [8] 
        Accelerating Sovereign AI: Strategies for Local AI Development and Sustainable Transformation (Presented by Capgemini) | GTC Paris 2025

        Experts Naghmeh Gorgin, Thordur Arnason, Hans Bendik Jahren, and Etienne Grass discussed sovereign AI, its impact, and strategies for local development. They presented a pilot with the Red Cross, using Capgemini's BabelSpeak and Telenor's AI Factory, and provided guidance on infrastructure design and model engineering using NVIDIA technology was not mentioned but Capgemini's AI-driven translation service was.

        https://www.nvidia.com/en-us/on-demand/session/gtcparis25-gps1050/
      • [9] 
        Sovereign AI Summit | GTC 2025

        The Sovereign AI Summit featured Calista Redmond, Landon Davidson, and Shilpa Kolhatkar from NVIDIA, along with other industry leaders, including Nadia Carlsten from the Danish Centre for AI Innovation, Nagendra Bandaru from Wipro, and Nelson Leoni from WideLabs. The summit explored how nations like the UK, France, Italy, Denmark, Japan, India, Korea, Thailand, and Brazil are developing domestic AI capabilities to drive economic growth. Peter Kyle, UK's Secretary of State for Science, Innovation and Technology, discussed the UK's AI Opportunities Action Plan. The event included panels on unleashing sovereign AI's potential, empowering ecosystems with sovereign foundation models, and lessons learned from deploying Gen AI applications, with participation from companies like NAVER Cloud, iGenius, and ThinkDeep AI.

        https://www.nvidia.com/en-us/on-demand/session/gtc25-se73317/
      • [10] 
        Indonesia on Track to Achieve Sovereign AI Goals With NVIDIA, Cisco and IOH | July 2025

        Indonesia is making strides towards its Golden 2045 Vision with the help of NVIDIA, Cisco, and Indosat Ooredoo Hutchison (IOH), aiming to establish a sovereign AI ecosystem. The partnership includes the creation of an AI Center of Excellence (CoE) led by the Ministry of Communications and Digital Affairs (Komdigi), which will feature an NVIDIA AI Technology Center offering research support, NVIDIA Inception program benefits, and NVIDIA Deep Learning Institute training. The CoE will comprise an AI factory with NVIDIA Blackwell GPUs, NVIDIA Cloud Partner reference architectures, and NVIDIA AI Enterprise software, as well as a Cisco-powered Sovereign Security Operations Center Cloud Platform. Key figures involved include Vikram Sinha, President Director and CEO of IOH, Chuck Robbins, Cisco Chair and CEO, and Ronnie Vasishta, NVIDIA Senior Vice President of Telecom. The initiative aims to democratize AI, establish sovereign infrastructure, secure AI workloads, make AI accessible to all, and develop a talent and development ecosystem, with plans to equip 1 million people with digital skills by 2027.

        https://blogs.nvidia.com/blog/indonesia-ai-center-o

    • Try These Search Refinements

      ai strategy

      data security

      cloud platform

      regulatory compliance

      digital autonomy

      1-10 of 100