NVIDIA has officially unveiled the GeForce RTX 50 Series Desktop and Laptop GPUs at CES 2025. The new GPU series is powered by the NVIDIA Blackwell architecture, fifth-generation Tensor Cores and fourth-generation RT Cores. The GeForce RTX 50 Series delivers breakthroughs in 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 5090 GPU is the fastest GeForce RTX GPU available right now in the market. It has 92 billion transistors and can handle over 3,352 trillion AI operations every second. Thanks to new architecture and technology like DLSS 4, it is up to twice as fast as the GeForce RTX 4090 GPU.
The GeForce Blackwell also comes to laptops, offering the same great features as desktop versions. This makes portable computing much better, with impressive graphics and energy efficiency. The new NVIDIA Max-Q technology in Blackwell increases battery life by up to 40%, allowing for slim and light laptops that look good while still delivering strong performance.
NVIDIA DLSS 4 Boosts Performance by Up to 8x
DLSS 4 introduces Multi Frame Generation, a new feature that uses AI to create up to three extra frames for each frame that is rendered. This technology works with other DLSS features to improve performance by as much as 8 times compared to traditional rendering, while ensuring quick responses with NVIDIA Reflex technology.
Additionally, DLSS 4 brings the first real-time use of transformer model architecture in gaming. The new DLSS Ray Reconstruction and Super Resolution models use double the parameters and four times the computing power, resulting in better stability, less ghosting, more detail, and improved anti-aliasing in game graphics. DLSS 4 will be available on GeForce RTX 50 Series GPUs for over 75 games and applications on launch day.
NVIDIA Reflex 2 features Frame Warp, a clever method that decreases latency in games by updating a rendered frame with the latest mouse input just before it is shown on the screen. Reflex 2 can cut latency by up to 75%, giving gamers an advantage in multiplayer games and making single-player games feel more responsive.
Blackwell Brings AI to Shaders
Twenty-five years ago, NVIDIA launched the GeForce 3 and programmable shaders, starting a long journey of graphics advancements. These innovations include pixel shading, compute shading, and real-time ray tracing. Now, with the GeForce RTX 50 Series GPUs, NVIDIA is introducing RTX Neural Shaders. This new feature uses small AI networks in shaders to create high-quality textures, lighting, and more for real-time games.
Rendering game characters is difficult because small mistakes can easily be spotted. RTX Neural Faces takes a basic digital face and pose data and uses AI to produce a smooth, high-quality digital face in real time.
RTX Neural Faces works together with new RTX technologies for ray-traced hair and skin. In addition, the new RTX Mega Geometry allows for up to 100 times more ray-traced triangles in a scene. These developments will greatly enhance the realism of game characters and environments.
The capabilities of neural rendering, DLSS 4, and the new DLSS transformer model are demonstrated on GeForce RTX 50 Series GPUs in Zorah, a cutting-edge technology demo from NVIDIA.
Autonomous Game Characters
GeForce RTX 50 Series GPUs provide top-level AI performance to help create game characters that can act independently while the game is running.
NVIDIA has introduced new technologies called NVIDIA ACE that allow game characters to see, think, and act like human players. ACE-powered characters are being included in KRAFTON’s PUBG: BATTLEGROUNDS and the upcoming life simulation game InZOI, as well as Wemade Next’s MIR5.
In PUBG, companions using NVIDIA ACE make plans and take action, working with human players to survive. In InZOI, Smart Zoi characters change their behavior based on goals and events in the game. In MIR5, raid bosses driven by a large language model adapt their tactics based on how players act, making the game more exciting and challenging.
AI Foundation Models for RTX AI PCs
NVIDIA will offer a series of NIM microservices and AI Blueprints for RTX AI PCs, created by leading model developers like Black Forest Labs, Meta, Mistral, and Stability AI. These services help RTX enthusiasts and developers build AI agents and assistants.
The uses include large language models (LLMs), vision language models, generating images, speech recognition, models for improved information retrieval, PDF extraction, and computer vision. The NIM microservices provide everything needed to run AI on PCs and are optimized for all NVIDIA GPUs.
To show how these tools can be used, NVIDIA presented Project R2X, a PC avatar with vision capabilities that can help users access information, assist with applications and video calls, read documents, and summarize them.
AI-Powered Tools for Creators
The GeForce RTX 50 Series GPUs make creative tasks faster and easier. They are the first consumer GPUs to use FP4 precision, which helps improve AI image generation speed for tools like FLUX by 2 times. This means generative AI models can now run on your computer with less memory needed than older models.
The NVIDIA Broadcast app now has two new features for livestreamers. Studio Voice enhances microphone audio, while Virtual Key Light improves lighting on people’s faces for better streams. Additionally, Streamlabs is introducing the Intelligent Streaming Assistant, powered by NVIDIA ACE and Inworld AI, which acts as a cohost and technical helper to make livestreams even better.
Availability
For desktop users, the GeForce RTX 5090 GPU with 3,352 AI TOPS and the GeForce RTX 5080 GPU with 1,801 AI TOPS will be available on Jan. 30 at $1,999 and $999, respectively. The GeForce RTX 5070 Ti GPU with 1,406 AI TOPS and GeForce RTX 5070 GPU with 988 AI TOPS will be available starting in February at $749 and $549, respectively.
The NVIDIA Founders Editions of the GeForce RTX 5090, RTX 5080 and RTX 5070 GPUs will be available directly from nvidia.com and select retailers worldwide.
Stock-clocked and factory-overclocked models will be available from top add-in card providers such as ASUS, Colorful, Gainward, GALAX, GIGABYTE, INNO3D, KFA2, MSI, Palit, PNY and ZOTAC, and in desktops from system builders including Falcon Northwest, Infiniarc, MAINGEAR, Mifcom, ORIGIN PC, PC Specialist and Scan Computers.
Laptops with GeForce RTX 5090, RTX 5080 and RTX 5070 Ti Laptop GPUs will be available starting in March, and RTX 5070 Laptop GPUs will be available starting in April from the world’s top manufacturers, including Acer, ASUS, Dell, GIGABYTE, HP, Lenovo, MECHREVO, MSI and Razer.