enfrdepluk
Search find 4120  disqus socia  tg2 f2 lin2 in2 X icon 3 y2  p2 tik steam2

NVIDIA Blackwell-powered GeForce RTX 50 Series graphics cards open up a new world of AI-powered graphics

Author Maximum Games, . Published in IT market

Go to comments

Today, NVIDIA announced the world’s most advanced graphics cards for gamers, creators, and developers—the GeForce RTX 50 series desktop and mobile GPUs. Powered by the NVIDIA Blackwell architecture with fifth-generation Tensor Cores and fourth-generation RT Cores, GeForce RTX 50 series GPUs deliver breakthrough AI rendering technologies like neural shading, digital character technology, geometry processing, and lighting.

“The Blackwell architecture, the powerhouse of AI, is now available to PC gamers, developers, and content creators,” said Jensen Huang, founder and CEO of NVIDIA. “By combining AI-powered neural rendering and ray tracing, the Blackwell architecture is the most significant innovation in computer graphics since we first introduced programmable shading 25 years ago.”

RTX 5090 Front

The GeForce RTX 5090, the fastest GeForce RTX graphics card ever, features 92 billion transistors and delivers over 3 trillion AI operations per second (TOPS). Powered by Blackwell architecture and DLSS 352, the GeForce RTX 4 delivers up to 5090X the performance of the GeForce RTX 4090.

RTX 5000 Specs: Detailed Description

GeForce RTX 5090

The flagship RTX 5090 model will go on sale for $1. This graphics card is equipped with:

  • 21 CUDA cores,

  • 32GB of GDDR7 memory,

  • 512-bit memory bus,

  • Throughput up to 1,792 GB/s.

The RTX 5090 has a maximum power of 600W, making it the perfect choice for enthusiasts who want maximum performance. The latest DLSS 4.0 and RTX Neural Shaders technologies double the performance of the RTX 4090.

GeForce RTX 5080

The RTX 5080 is available for $999 and features:

  • 10 CUDA cores,

  • 16GB GDDR7 memory,

  • 256-bit memory bus,

  • Bandwidth up to 960 GB/s,

  • Power consumption: 360 W.

The RTX 5080 is ideal for premium gamers and content creators, delivering high performance and advanced ray tracing capabilities.

GeForce RTX 5070 Ti

For the mass market, NVIDIA offers the RTX 5070 Ti for $749. Its specifications include:

  • 8 CUDA cores,

  • 16GB GDDR7 memory,

  • Bandwidth 896 GB/s,

  • Power consumption 300 W.

This model provides an excellent balance between price and performance, making it a popular choice for experienced gamers.

GeForce RTX 5070

The most affordable RTX 5070 model costs $549 and comes with:

  • 6 CUDA cores,

  • 12GB GDDR7 memory,

  • Bandwidth 672 GB/s,

  • Power consumption 250 W.

The RTX 5070 delivers solid performance and supports all the latest ray tracing and AI acceleration technologies.

All cards support DLSS 4.0, which delivers improved ray-traced gaming and accelerated AI performance. With official availability in January 2025, the RTX 5000 series promises to be the number one choice for gamers and professionals.

GeForce RTX for notebooks powered by Blackwell architecture deliver all the capabilities of desktop models, delivering significant performance gains on mobile platforms, while delivering outstanding graphics capabilities and exceptional power efficiency. Blackwell-generation NVIDIA Max-Q technologies deliver up to 40% longer battery life and enable thin and light notebooks with sleek designs without sacrificing power or performance. NVIDIA DLSS 4 delivers up to 8x faster performance

GeForce RTX 5000 Mobile Graphics Cards

At CES 2025, NVIDIA officially announced prices for laptops with GeForce RTX 5000 series mobile graphics cards, which will go on sale in March 2025. New GPU use innovative GDDR7 memory, delivering high bandwidth and performance for gamers and professionals.

MSI NB Vector A18

GeForce RTX 5090 for laptops

Flagship laptops with RTX 5090 are priced at $2. Key features:

  • 24GB GDDR7 memory,

  • 1850 AI TOPS of performance,

  • 384-bit memory bus,

  • Excellent energy efficiency.

This model is designed for graphics-intensive tasks including 3D rendering, big data processing, and ray-traced gaming.

GeForce RTX 5080 for laptops

RTX 5080 laptops cost $2 and feature:

  • 16GB GDDR7 memory,

  • 1350 AI TOPS of performance,

  • 256-bit memory bus.

These devices are ideal for content creation and high-performance gaming, offering DLSS 4.0 technologies and stunning image quality.

GeForce RTX 5070 Ti for Laptops

RTX 5070 Ti laptop models will retail for $1. Specifications:

  • 12GB GDDR7 memory,

  • 1000 AI TOPS,

  • 192-bit memory bus.

GeForce RTX 5070 for laptops

The most affordable RTX 5070 laptops will cost $1. They will offer:

  • 12GB GDDR7 memory,

  • 800 AI TOPS of performance.

All RTX 5000 laptops support ray tracing, DLSS 4.0, and new AI technologies. NVIDIA's RTX 5000 series promises to be the market leader, offering the perfect balance of price and performance.

NVIDIA DLSS 4 Boosts Performance Up to 8x

DLSS 4 introduces the first multi-frame generation technology, Multi Frame Generation, which uses AI to create up to three additional frames for every one rendered, dramatically increasing frame rates. This technology works in unison with other DLSS 4 technologies to deliver a combined performance gain of up to 8x over traditional rendering while maintaining the responsiveness of NVIDIA Reflex.

dlss 4

DLSS 4 also introduces the industry’s first real-time Transformer AI models. New Transformer-based DLSS Ray Reconstruction and Super Resolution models use 2X more parameters and 4X more compute power to deliver improved image stability, reduced ghosting, greater motion detail, and improved edge anti-aliasing. DLSS 4 will be available on GeForce RTX 50-series GPUs in over 75 games and applications at launch.

dlss 4 2

NVIDIA Reflex 2 debuts Frame Warp, an innovative latency-reducing gaming technique that updates the last rendered frame based on a new control signal just before the frame is sent to the screen. Reflex 2 can reduce latency by up to 75%. The increased responsiveness of NVIDIA Reflex gives you an edge in multiplayer games and makes single-player games more enjoyable to play.

Blackwell Brings Artificial Intelligence to Shaders

Twenty-five years ago, NVIDIA introduced GeForce 3 and programmable shaders, which laid the foundation for two decades of graphics innovation, from pixel shading to compute shaders to real-time ray-traced rendering. Along with graphics cards
With GeForce RTX 50, NVIDIA introduces RTX Neural Shaders, which integrate small neural networks into programmable shaders to bring cinematic-quality materials, lighting, and other visual effects to games.

Rendering game characters is one of the most challenging tasks in graphics, as we are able to notice the smallest errors and inaccuracies in images of digital people. RTX Neural Faces takes a rasterized face and 3D pose data as input, then uses generative AI to create a temporarily stable, high-quality digital face in real time.

RTX Neural Faces is complemented by new RTX ray tracing technologies for hair and skin. Together with the new RTX Mega Geometry technology, which increases the number of ray-traced triangles in a scene by up to 100 times, these advanced capabilities will provide a significant leap in the realism of game characters and environments.

The capabilities of neural rendering, DLSS 4, and the DLSS Transformer model on GeForce RTX 50-series graphics cards are showcased in Zorah, a stunning new graphics demo from NVIDIA.

Autonomous game characters

GeForce RTX 50 Series GPUs deliver industry-leading TOPS (Total Opportunity Performance) to create autonomous game characters that are simulated in parallel with game rendering.

RTX AI PC Image

NVIDIA is introducing a new suite of NVIDIA ACE technologies that enable in-game characters to perceive, plan, and act just like real players. Autonomous characters powered by ACE are being integrated into PUBG: BATTLEGROUNDS and InZOI, an upcoming life simulation game from publisher KRAFTON, as well as MIR5 from Wemade Next.

In PUBG, NVIDIA ACE-powered companions plan and execute strategic moves, dynamically interacting with real players to ensure the survival of the team. InZOI introduces “Smart Zoi” characters that autonomously change their behavior based on life goals and in-game events. In MIR5, raid bosses powered by Large Language Models (LLM) adapt tactics in response to player actions, creating more dynamic and challenging battles.

Fundamental AI Models for PC with RTX

The application areas of the fundamental models include large language models (LLM), multimodal models, image generation, speech processing, models for creating embeddings in generation tasks using advanced search, data extraction from PDF, and computer vision. NIM microservices include all the necessary models and components for running AI on RTX PCs and are optimized for deployment on NVIDIA GPUs.

To demonstrate how enthusiasts and developers can use NIM microservices to build AI agents and assistants on the RTX platform, NVIDIA will release a set of NVIDIA NIM microservices and AI Blueprints from leading companies: Black Forest Labs, Meta, Mistral, and Stability.AI. To demonstrate the power of NIM for building AI agents and assistants, NVIDIA today announced a preview of Project R2X, a computer vision-enabled PC avatar that can provide information to the user, assist with applications and video conferencing, read and summarize documents, and perform many other tasks.

AI-powered tools for creativity

GeForce RTX 50 Series graphics cards dramatically accelerate all creative workflows. RTX 50 Series graphics cards are the first consumer GPU with support for FP4 precision, which increases AI image generation performance by 2x in models like FLUX and allows generative AI models to be run locally with less video memory than previous-generation GPUs.

The NVIDIA Broadcast app has received two new AI-powered beta features for streamers: Studio Voice, which improves microphone audio quality, and
Virtual Key Light, which allows you to illuminate faces for a more professional look. While Streamlabs introduced an intelligent streaming assistant powered by NVIDIA ACE and Inworld AI that acts as a co-host, producer, and technical assistant, improving the quality of live broadcasts.

Availability and prices

For desktop users, the GeForce RTX 5090 with 3 AI TOPS and GeForce RTX 352 with 5080 AI TOPS will go on sale on January 1 for $801 and $30, respectively.

5000

The GeForce RTX 5070 Ti with 1 AI TOPS and GeForce RTX 406 with 5070 AI TOPS will be available in February for $988 and $749, respectively.

Founders Edition models of NVIDIA's GeForce RTX 5090, RTX 5080, and RTX 5070 graphics cards will be available for purchase directly from NVIDIA.com and select retailers worldwide.

Factory and stock clocked models will be offered by leading graphics card manufacturers such as ASUS, Colorful, Gainward, GALAX, GIGABYTE, INNO3D, KFA2, MSI, Palit, PNY and ZOTAC, as well as in desktop PCs 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 graphics cards will be available in March for $2,899, $2,199, and $1,599, respectively, while RTX 5070 models will go on sale in April for $1,299 from leading global manufacturers including Acer, ASUS, Dell, GIGABYTE, HP, Lenovo, MECHREVO, MSI, and Razer.