Over 6000 CES attendees were treated to a cascade of exciting announcements as NVIDIA founder and CEO Jensen Huang kicked off CES 2025 with his keynote on Monday at the Mandalay Bay Michelob ULTRA Arena. Demonstrating the power of ideas, technology, and conviction to drive innovation and impact in business and society, Huang shared NVIDIA’s advancements and vision for agentic AI, autonomous vehicles, robotics, and gaming.
Leading the lengthy list of announcements, Huang recalled the company’s humble beginnings up to the current and exciting innovations surrounding AI.
“It started with perception AI understanding images, words, and sounds. Then generative AI — creating text, images, and sound,” Huang said. “Every single layer of the technology stack has been completely changed…we can now understand the information of just about any modality. We can understand amino acids. We can understand physics. We understand them, we can translate them and generate them. The applications are just completely endless.”
“I am so thrilled to kick off this show with a keynote by one of the most consequential companies in the world,” said Gary Shapiro, CEO of CTA in his introduction of Huang. “Today, Nvidia is pioneering breakthroughs in AI and accelerated computing that touches nearly every person and every business.”
AI took center stage during Huang’s keynote, beginning with the announcement of NVIDIA’s most powerful GeForce GPU yet — the new NVIDIA GeForce RTX 5090 GPU, which features 92 billion transistors and delivers 3352 trillion AI operations per second.
“AI is the house that GeForce built. GeForce enabled AI to reach the masses, and now AI is coming home to GeForce,” said Huang as he proudly displayed the RTX 5090 to the attendees. The RTX 50 series utilizes Blackwell architecture, which harnesses AI to enable breakthrough graphics. Thanks to its AI capabilities, NVIDIA can slash the price for its RTX 5070 model to $549, allowing gamers and content creators to step up their graphics game without breaking the bank.
The GeForce RTX 5090 and GeForce RTX 5080 desktop GPUs are scheduled to be available by January 30, and the GeForce RTX 5070 Ti and GeForce RTX 5070 desktops are slated to be available starting in February 2025. Laptop GPUs are expected in March 2025.
Other groundbreaking announcements from NVIDIA included AI foundation models for RTX PCs featuring NVIDIA NIM microservices and AI Blueprints for crafting digital humans, podcasts, images, and videos.
“The ChatGPT moment for general robotics is just around the corner,” Huang enthused as he was joined virtually onstage by a lineup of robots in development with various partners that utilize NVIDIA technology to accelerate their development.
Huang also introduced the world foundation model platform, NVIDIA Cosmos, a game-changer for robotics and industrial AI, particularly for developers who do not have the expertise and resources to train their own. Cosmos integrates generative models, tokenizers, and a video processing pipeline to power physical AI systems like AVs and robots. Huang’s announcement that Cosmos is open license and available on GitHub was met with enthusiastic applause and is an encouraging step towards opening AI development to the world.
Huang concluded his keynote with the announcement of Project DIGITS — NVIDIA’s personal AI supercomputer that aims to put NVIDIA Grace-Blackwell on every desk and at every AI developer’s fingertips. “Every software engineer, every engineer, every creative artist, everybody who uses computers today as a tool will need an AI supercomputer,” Huang said. Project DIGITS runs the entire NVIDIA AI stack, and all of NVIDIA software, including DGX Cloud. Project DIGITS is expected to be available in May 2025.
