CES 2025: NVIDIA Had A Field Day On Media Day

It was a field day for NVIDIA as people queued up to witness the company’s Founder and CEO Jensen Huang deliver the keynote speech at CES 2025, which is being held at Las Vegas, USA, from January 7 to January 10.

Huang’s 90-minute keynote speech was delivered to over 6,000 attendees, and included new products to advance gaming, autonomous vehicles, robotics, and agentic AI.

While the annual consumer technology trade show garners the attention of tech conglomerates and tech enthusiasts alike, this year all eyes are on NVIDIA, as people eagerly awaited the newest and hottest updates from the undisputed GPU leader of the artificial intelligence (AI) era—and the AI giant didn’t disappoint.

However, among the various updates provided by NVIDIA, the one that stood out the most was the launch of the company’s new GPU (Graphics Processing Unit) series—the GeForce RTX 50 series.

NVIDIA GeForce RTX 50 Series

NVIDIA unveiled and provided a demo of its newest and most advanced consumer GPUs for gamers, creators and developers—the GeForce RTX 50 series of GPUs, which will be available for desktops and laptops.

The GPUs are powered by NVIDIA’s flagship Blackwell architecture, and can deliver advanced AI-driven rendering, including neural shaders, digital human technologies, geometry and lighting.

“Blackwell, the engine of AI, has arrived for PC gamers, developers and creatives,” said Jensen Huang, founder and CEO of NVIDIA. “Fusing AI-driven neural rendering and ray tracing, Blackwell is the most significant computer graphics innovation since we introduced programmable shading 25 years ago.”

The GeForce RTX 50 series consists of 4 options: RTX 5090, RTX 5080, RTX 5070 Ti and RTX 5070.

As per the company, the GeForce RTX 5090 is said to outperform its predecessor, the RTX 4090, by up to 2x better performance, and offers 125 watts higher total graphics power (575 watts) than the RTX 4090. Coming in with 92 billion transistors and 3,352 AI TOPS (trillion AI operations per second), it’s the fastest GeForce RTX GPU to date. It costs a whopping $1,999 and will be available on January 30.

The RTX 5080 (1,801 AI TOPS) will cost $999 and will also be available on January 30.

However, the RTX 5070 Ti (1,406 AI TOPS) and RTX 5070 (988 AI TOPS) will be available in February and cost $749 and $549, respectively. The GPUs are also smaller in size than their respective predecessors.

Laptops with the new GPUs will be available from March 2025, other than those powered by the RTX 5070, which will come in April.

Ahead of this, NVIDIA said that its DLSS 4 (Deep Learning Super Sampling) enabled up to 8x performance boosts over traditional rendering while also introducing the graphics industry’s first real-time application of the transformer model architecture. This will be supported by the new series of GPUs in over 75 games and applications when launched.

NVIDIA Project DIGITS

NVIDIA also announced Project DIGITS, a personal AI supercomputer that provides AI researchers, data scientists and students worldwide with access to the power of the NVIDIA Grace Blackwell platform.

The device sports a GB10 Grace Blackwell Superchip, which allows users to seamlessly develop and run inference on models using their own desktop system, as well as deploy them on accelerated cloud or data center infrastructure. This offers users a petaflop of AI computing performance for prototyping, fine-tuning and running large AI models.

These supercomputers, which start at $3,000, will be available from May 2025, and as per NVIDIA, each unit can run models up to 200 billion parameters in size.

“AI will be mainstream in every application for every industry. With Project DIGITS, the Grace Blackwell Superchip comes to millions of developers,” said Jensen Huang, founder and CEO of NVIDIA.

“Placing an AI supercomputer on the desks of every data scientist, AI researcher and student empowers them to engage and shape the age of AI."

Cosmos World Foundation Models

Jensen Huang also revealed that NVIDIA was diving into world models, which are inspired by the mental models of the world that humans develop naturally.

NVIDIA is calling their platform Cosmos World Foundation Models or Cosmos WFM, which is described as a game-changer for robotics and industrial AI. NVIDIA is making the first wave of Cosmos WFMs available to researchers and developers from all business sizes, freely.

The idea is to help developers build next-generation robots and autonomous vehicles (AVs) using advanced AI technology.

According to Jensen Huang, “The ChatGPT moment for general robotics is just around the corner.”

The access to models didn’t end here, as NVIDIA also revealed that businesses planning to build AI agents can also use the new open NVIDIA Llama Nemotron (built on Meta’s Llama model) and Cosmos Nemotron models.

The extended to the NVIDIA DRIVE Hyperion AV platform, which is built on the new NVIDIA AGX Thor system-on-a-chip (SoC) and is designed for generative AI models and delivering advanced functional safety and autonomous driving capabilities.

“The autonomous vehicle revolution is here,” Huang said. “Building autonomous vehicles, like all robots, requires three computers: NVIDIA DGX to train AI models, Omniverse to test drive and generate synthetic data, and DRIVE AGX, a supercomputer in the car.”

As such, NVIDIA announced that Toyota, Aurora and Continental have joined the list of global mobility leaders that will infuse their next-generation vehicles with NVIDIA’s Drive AGX Orin supercomputer and DriveOS, a safety-focused operating system.

NVIDIA Mega

NVIDIA also unveiled Mega, an Omniverse Blueprint designed to develop, test, and optimize physical AI and robot fleets at scale within digital twins before real-world deployment.

The move comes as NVIDIA recognizes that modern factories and warehouses operate with fleets of autonomous robots, robotic arms, and humanoids alongside human workers and such environments demand sophisticated coordination and simulation to enhance safety, prevent disruptions, and optimize operations.

Essentially, Mega leverages digital twins featuring a world simulator to orchestrate robot activities and sensor data. This enables continuous updates to robotic intelligence for smarter task execution and operational efficiency. With Mega, businesses can refine AI-driven robotics, ensuring seamless integration into real-world facilities.

This can be enhanced further by using Apple Vision Pros to teach robots how to execute specific tasks by demonstrating actions.

Generative Physical AI

Another major announcement included NVIDIA’s new generative AI models and blueprints that expand NVIDIA Omniverse integration further into physical AI applications such as robotics, autonomous vehicles and vision AI. As per NVIDIA, it’s currently being used to develop new products and services that will accelerate the next era of industrial AI.

So far, Accenture, Altair, Ansys, Cadence, Foretellix, Microsoft and Neural Concept revealed their plans to integrate Omniverse into their next-generation software products and professional services.

“Physical AI will revolutionize the $50 trillion manufacturing and logistics industries. Everything that moves—from cars and trucks to factories and warehouses—will be robotic and embodied by AI,” said Jensen Huang. “NVIDIA’s Omniverse digital twin operating system and Cosmos physical AI serve as the foundational libraries for digitalizing the world’s physical industries.”

What do you think about NVIDIA’s announcements?

Let us know in the comments below!