The keynote address delivered by NVIDIA CEO Jensen Huang at the CES 2025 event in Las Vegas underscored the significant advancements and emerging trends in artificial intelligence (AI) and computing industries. Over a span of 90 minutes, Huang elaborated on multiple groundbreaking developments in gaming, autonomous vehicles, robotics, and agentic AI, highlighting how NVIDIA’s latest products are fostering these innovations.
The Evolution of AI Capabilities
Artificial Intelligence (AI) has undergone significant transformation over recent years, evolving from basic algorithms to sophisticated systems capable of performing complex tasks. As AI technology continues to advance, its applications have expanded into various fields, including healthcare, finance, and customer service, revolutionizing how we approach problem-solving and decision-making.
From Perception AI to Physical AI
Huang traced the trajectory of AI from perception AI, which is adept at understanding images, words, and sounds, to generative AI, which creates text, images, and sounds. Now, he observed, we are entering the era of “physical AI”—AI that can perceive, reason, plan, and act. This new phase promises to significantly transform multiple sectors through enhanced NVIDIA GPUs and platforms. By integrating sophisticated algorithms and cutting-edge hardware, physical AI is set to not only enhance current technological applications but also pioneer groundbreaking innovations across various domains.
This evolution marks the dawn of AI systems that go beyond virtual environments to actively interact with the physical world in meaningful ways. For instance, in autonomous driving, physical AI can imbue vehicles with the capability to perceive their surroundings, anticipate potential hazards, and make real-time decisions that ensure safer travel. Similarly, robots equipped with physical AI can perform complex tasks in industrial settings, improving efficiency and reducing human error. Huang’s vision indicates a future where AI seamlessly merges with the physical world, amplifying the capabilities of machines and offering substantial advancements in technology-driven fields.
Impact on Various Sectors
The rapid evolution of AI capabilities is set to revolutionize industries such as gaming, autonomous vehicles, and robotics. Huang emphasized that the advancements in AI are not just theoretical but are being realized through NVIDIA’s cutting-edge technology, which is pushing the boundaries of what AI can achieve. In the gaming industry, for instance, the implementation of AI-driven graphics and real-time rendering can create more immersive and realistic experiences for players, elevating the standard of visual fidelity and interactivity.
In the automotive sector, autonomous vehicles equipped with advanced AI can navigate complex environments, recognize obstacles, and make informed decisions, ultimately leading to safer and more efficient transportation systems. Moreover, AI’s role in robotics is expanding, with robots becoming more adept at handling a broader range of tasks, from precision manufacturing to high-stakes search and rescue operations. NVIDIA’s technology enables these machines to learn and adapt, making them more intelligent and capable over time. The enhancements in AI not only promise to transform these industries but also pave the way for new applications that were previously unimaginable.
NVIDIA Cosmos Platform
Introduction of the Cosmos Platform
One of the most notable announcements Huang made was the unveiling of the NVIDIA Cosmos platform. This innovative platform advances physical AI by introducing new models and video data processing pipelines specifically designed for robots, autonomous vehicles, and vision AI. This integration is poised to revolutionize how AI systems interact with and navigate the physical world. By capturing and processing video data more efficiently, the Cosmos platform allows AI to understand and react to its surroundings with an unprecedented level of accuracy and speed.
The platform’s architecture is built to support high-performance computing and deep learning, ensuring that AI applications can run seamlessly and execute complex tasks in real-time. This breakthrough provides a foundational infrastructure for developing highly functional robots and self-driving cars, capable of performing tasks ranging from basic mobility to intricate operations that require precise environmental interaction. The introduction of the Cosmos platform signifies a leap forward in the deployment of physical AI, enabling a new era of intelligent machines that can operate autonomously in dynamic real-world settings.
Applications in Robotics and Autonomous Vehicles
The Cosmos platform is expected to bring us closer to realizing highly functional robots and self-driving cars. By enhancing the ability of AI systems to perceive and interact with their environment, NVIDIA is paving the way for significant advancements in these fields, making them more reliable and efficient. In robotics, the platform’s robust processing capabilities allow for the development of machines that can perform complex tasks such as assembly line work, medical procedures, and even household chores with exceptional precision and dexterity.
For autonomous vehicles, the Cosmos platform provides the necessary computational power to process vast amounts of sensory data in real-time, allowing these vehicles to navigate through traffic, recognize road signs, and avoid obstacles with greater accuracy. This advancement not only enhances safety but also increases the efficiency of transportation networks, reducing congestion and lowering the risk of accidents. The integration of the Cosmos platform into autonomous systems heralds a new wave of technological innovation, offering practical solutions that address some of the most pressing challenges in today’s rapidly evolving landscape.
GeForce RTX 50 Series GPUs
Introduction of Blackwell-Based GPUs
Huang showcased the next generation of NVIDIA’s GPU technology with the introduction of the Blackwell-based GeForce RTX 50 Series GPUs. The flagship product, the GeForce RTX 5090 GPU, boasts a staggering 92 billion transistors and can perform 3,352 trillion AI operations per second. This level of performance enables unprecedented visual realism and significant performance boosts in gaming and other intensive applications. The enhanced computational power of these GPUs facilitates the rendering of ultra-high-definition graphics and complex simulations, making them ideal for both professional and entertainment purposes.
These revolutionary GPUs are designed to support the most demanding workloads, including real-time ray tracing and AI-driven content creation. The Blackwell architecture introduces new levels of efficiency and scalability, ensuring that applications can leverage the full potential of the hardware. Whether it’s creating lifelike animations or running extensive scientific simulations, the GeForce RTX 50 Series GPUs set a new standard in graphical and computational performance, pushing the boundaries of what is possible in digital content creation and consumption.
Key Features and Enhancements
The new GPUs, scheduled to be available starting from late January 2025, incorporate advancements such as DLSS 4, which includes Multi Frame Generation to increase performance by up to 8x, and NVIDIA Reflex 2, which reduces PC latency by up to 75%. Additionally, features like RTX Neural Shaders and RTX Mega Geometry are set to advance real-time rendering and significantly enhance the graphical detail and performance of gaming and simulation environments. DLSS 4 leverages advanced AI algorithms to upscale lower-resolution images in real-time, delivering crisp and detailed visuals without compromising performance.
On the other hand, NVIDIA Reflex 2 optimizes the responsiveness of gaming systems, enabling smoother gameplay and faster reaction times, which is crucial for competitive gaming. The integration of RTX Neural Shaders brings unprecedented efficiency in shader execution, allowing developers to create more complex and detailed scenes. RTX Mega Geometry further enhances the capability to render highly detailed virtual worlds, making the environments more immersive and realistic. Together, these features represent a significant leap forward in GPU technology, providing users with a superior graphical experience and setting new benchmarks for the industry.
AI Foundation Models for RTX PCs
NVIDIA has been at the forefront of developing AI models tailored for RTX PCs, offering users enhanced performance and capabilities.
Enhancing Everyday Computing
Further extending NVIDIA’s influence into everyday computing, Huang introduced AI foundation models for RTX PCs. These models, accelerated by the new GeForce RTX 50 Series GPUs, are designed to enhance digital human creation, content production, and various productivity tasks. These AI models are available on multiple cloud platforms and can be integrated into software to create AI agents tailored to specific needs. This integration empowers developers and creators to harness the power of AI for tasks such as image recognition, natural language processing, and predictive analytics, significantly enhancing the functionality and efficiency of everyday computing.
The AI foundation models are built to leverage the computational strengths of the RTX 50 Series GPUs, ensuring that AI-driven tasks can be executed with remarkable speed and precision. This advancement not only simplifies the development process but also broadens the scope of applications that can benefit from AI, ranging from automated customer service solutions to advanced data analysis tools. By making high-powered AI tools accessible and user-friendly, NVIDIA is enabling a new wave of innovation in personal computing, transforming how we interact with technology on a daily basis.
NIM-Ready RTX AI PCs
The launch of NIM-ready RTX AI PCs marks a significant step towards bringing advanced AI capabilities into households and personal workflows. By making high-powered AI accessible to consumers, NVIDIA is enabling a new level of productivity and creativity in personal computing. These PCs are equipped with the latest NVIDIA hardware, optimized to run intensive AI workloads smoothly and efficiently. This allows users to perform complex tasks such as real-time video editing, 3D rendering, and sophisticated data analysis without the need for expensive and cumbersome setups.
NIM-ready RTX AI PCs come with pre-installed software that leverages NVIDIA’s AI foundation models, providing an out-of-the-box experience that caters to both novice users and seasoned professionals. This user-friendly approach democratizes access to cutting-edge AI technology, encouraging innovation and enabling individuals to explore new possibilities in their creative and professional endeavors. The introduction of these PCs signifies a transformative shift in personal computing, where powerful AI-driven applications become an integral part of everyday life, enhancing both productivity and creative expression.
Transforming Industries with AI
Autonomous Vehicles and Robotics
Huang’s keynote emphasized the pivotal role of AI in revolutionizing the automotive industry. The new NVIDIA DRIVE Hyperion AV platform, incorporating the NVIDIA AGX Thor SoC, aims to elevate generative AI models and functional safety in autonomous driving. This comprehensive platform aligns with the needs of automotive giants like Mercedes-Benz and Toyota, which are adopting NVIDIA’s technology for next-gen vehicle development. By providing the necessary tools and computational power, NVIDIA is facilitating the creation of autonomous vehicles that are not only safer but also more efficient and reliable.
The DRIVE Hyperion platform supports advanced sensor fusion, enabling vehicles to perceive their environment with a higher degree of accuracy. This, coupled with AI-driven decision-making algorithms, ensures that autonomous vehicles can navigate complex traffic scenarios, make informed decisions, and react promptly to unforeseen obstacles. The platform’s emphasis on functional safety is particularly crucial, as it ensures that these vehicles meet stringent safety standards, reducing the risk of accidents and enhancing overall road safety. NVIDIA’s contributions are driving the automotive industry towards a future where autonomous vehicles are a common sight on the roads, offering unparalleled convenience and safety.
Synthetic Data and Training
Huang highlighted the importance of synthetic data, produced using Omniverse and Cosmos simulations, in training autonomous vehicles. This approach allows for the creation of extensive datasets, significantly enhancing the reliability and safety of self-driving systems. Synthetic data generation enables the simulation of a wide range of driving scenarios, including rare and hazardous conditions that a vehicle may not frequently encounter in real life. This comprehensive training ensures that autonomous systems are well-prepared to handle diverse and unpredictable environments, improving their overall performance and safety.
The use of synthetic data also accelerates the development cycle of autonomous vehicles, as it reduces the dependency on real-world data collection, which can be time-consuming and expensive. Omniverse and Cosmos simulations provide a controlled and scalable environment where various parameters can be adjusted to test the vehicle’s response to different situations. This not only improves the robustness of the AI models but also allows for continuous enhancement and fine-tuning. By leveraging synthetic data, NVIDIA is enabling the development of highly reliable and safe autonomous vehicles, paving the way for widespread adoption and integration into everyday life.
Agentic AI and Digital Manufacturing
AI Blueprints for Workflow Automation
In the realm of enterprise and industrial applications, NVIDIA is providing tools for creating AI Blueprints, which facilitate the development of custom AI agents for automating workflows. These blueprint models, including innovative tools like PDF-to-podcast conversion and video search and summarization, are designed to integrate seamlessly with leading platforms, thereby improving efficiency in various professional settings. AI Blueprints offer a flexible and scalable solution for businesses to optimize their operations, reducing manual workloads and enhancing overall productivity and accuracy.
These AI-driven solutions enable companies to automate repetitive tasks, allowing employees to focus on more strategic and creative aspects of their work. For example, the PDF-to-podcast conversion tool can automatically convert written content into audio formats, making information more accessible and convenient for users. Similarly, video search and summarization tools can streamline content management, providing quick and accurate retrieval of relevant information. By integrating these AI Blueprints into their workflows, businesses can achieve significant time and cost savings, driving innovation and improving operational efficiency.
Isaac GR00T Blueprint
Huang also presented the NVIDIA Isaac GR00T Blueprint for synthetic motion generation, a tool that helps developers produce large synthetic motion datasets for training humanoids. By leveraging NVIDIA’s Omniverse, companies can simulate vast robotic fleets and optimize automation tasks, especially critical in sectors like warehousing and manufacturing. The Isaac GR00T Blueprint enables the creation of realistic motion patterns and behaviors, allowing robots to perform complex tasks with human-like precision and adaptability.
This tool facilitates the development and training of humanoid robots, enhancing their ability to interact with dynamic environments and perform tasks that require fine motor skills. By generating large synthetic motion datasets, developers can create more robust and versatile robots capable of handling a wide range of applications. The integration of these robots into industrial processes can lead to significant improvements in efficiency and accuracy, reducing the reliance on manual labor and minimizing errors. NVIDIA’s Isaac GR00T Blueprint represents a major advancement in robotic technology, pushing the boundaries of what is achievable in automation and digital manufacturing.
Project DIGITS: A Revolutionary Personal AI Supercomputer
Arguably one of the most groundbreaking aspects of Huang’s presentation was the introduction of Project DIGITS, an advanced AI supercomputer aimed at individual developers, engineers, and creatives. Powered by the GB10 Grace Blackwell Superchip, Project DIGITS delivers the full capabilities of NVIDIA’s AI stack in a compact form factor. Huang emphasized that this powerful device, available in May, represents a democratization of AI supercomputing, providing unprecedented computational power directly to end-users’ desks. This innovation makes high-performance AI tools accessible to a broader audience, enabling individual users to perform tasks that previously required large-scale computing resources.
Project DIGITS is designed to support a wide range of applications, from complex scientific research to creative content production. Its powerful processing capabilities allow for real-time data analysis, advanced machine learning, and high-fidelity simulations, making it an invaluable tool for professionals across various fields. By bringing AI supercomputing to the desktop, NVIDIA is empowering individuals to explore new possibilities, drive innovation, and push the boundaries of their respective disciplines. The introduction of Project DIGITS signifies a significant leap forward in personal computing, offering a level of performance and accessibility that was previously unattainable.
Conclusion: A Year of Breakthroughs
At the CES 2025 event held in Las Vegas, NVIDIA CEO Jensen Huang delivered an impactful keynote address, shedding light on the rapid advancements and emerging trends in artificial intelligence (AI) and the computing sector. The presentation, which spanned 90 minutes, was packed with exciting updates and innovations.
Huang delved into various groundbreaking developments across several domains, notably gaming, autonomous vehicles, robotics, and agentic AI. He emphasized how NVIDIA’s newest products are driving these innovations forward, making remarkable strides in each field.
In the gaming industry, Huang highlighted NVIDIA’s cutting-edge graphic technologies that are enhancing gaming experiences with more realistic and immersive visuals. The advancements in autonomous vehicles showcased NVIDIA’s commitment to transforming transportation, making it safer and more efficient with AI-driven technologies.
The presentation also touched upon significant progress in robotics, where NVIDIA’s innovations are enabling smarter and more capable robots that can adapt to complex environments. Moreover, Huang discussed agentic AI, illustrating how NVIDIA’s developments are pushing the boundaries of AI, allowing for more intuitive and responsive artificial agents.
Overall, Huang’s keynote at CES 2025 captured the essence of NVIDIA’s role in fostering technological innovations, demonstrating how their latest products are shaping the future of AI and computing, with far-reaching implications across various industries.
