NVIDIA RTX Technology
Fusing AI-driven neural rendering and ray tracing.
NVIDIA RTX™ technology is the most advanced platform for full ray tracing and AI-powered neural rendering technologies that revolutionize how we create and work. Over 100 professional and creative applications leverage RTX to accelerate performance, deliver lifelike graphics with physically accurate lighting, and power AI-driven workflows—from generating 3D assets to rendering complex simulations to running AI agents that automate steps across creative pipelines. With the latest enhancements in AI, ray tracing, and simulation, RTX technology enables incredible 3D designs, photorealistic simulations, and stunning visual effects faster than ever.
Technologies
RTX Core Technologies
Create rich, photorealistic visuals and true-to-life simulations with the latest NVIDIA RTX™ technologies.
Fifth-Generation Tensor Cores
Fifth-generation Tensor Cores deliver up to 3x higher throughput for deep learning, with new FP4 support for massive performance gains alongside TF32, BF16, FP16, FP8, and FP6 precision. A second-generation FP8 Transformer Engine—shared with data center-class NVIDIA Blackwell GPUs—accelerates LLM prototyping, inference, and AI development across frameworks like TensorFlow and PyTorch.
Fourth-Generation Ray Tracing Cores
Fourth-generation RT Cores deliver up to 2x faster ray tracing, rendering photorealistic scenes with physically accurate shadows, reflections, and refractions. Built for neural graphics technologies like RTX Mega Geometry and APIs including NVIDIA OptiX™, Microsoft DXR, and Vulkan, they power interactive design workflows with immediate feedback and dramatically shorter iteration cycles.
Blackwell Streaming Multiprocessor (SM)
The NVIDIA Blackwell SM unites next-gen CUDA™ cores and neural shaders for AI-augmented graphics and compute, scaling to 24,064 CUDA cores and 125 TFLOPS of FP32 performance—up to 1.4x higher than the previous generation. Embedded neural shaders bring AI directly into the graphics pipeline to power technologies like RTX Mega Geometry and DLSS. And because the SM runs ray tracing, AI inferencing, and compute concurrently, it accelerates AI-augmented workflows across AEC, manufacturing, and media workflows.
Deep Learning Super Sampling
Deep Learning Super Sampling (DLSS) is a revolutionary suite of neural rendering technologies that uses AI to boost FPS, reduce latency, and improve image quality. The latest release, DLSS 4.5, introduces a second-generation transformer model for Ray Reconstruction and Super Resolution—delivering superior lighting accuracy, temporal stability, and motion clarity in ray-traced and path-traced content—along with enhanced Multi Frame Generation, powered by NVIDIA RTX PRO™ Blackwell Series GPUs and fifth-generation Tensor Cores.
RTX Neural Shaders
RTX Neural Shaders bring small neural networks into programmable shaders, unlocking new techniques for radiance caching, texture and material compression, radiance fields, and next-generation asset generation. The RTX Neural Shaders SDK lets developers train shader data on an RTX PRO workstation and accelerate neural representations with NVIDIA Tensor Cores at runtime.
Mega Geometry Technology
RTX Mega Geometry dramatically increases the geometric detail possible in ray-traced scenes, accelerating BVH building to enable up to 100x more ray-traced triangles with better performance. It’s available today through NVIDIA RTX Kit, a suite of technologies for AI-accelerated ray tracing, immense geometry, and photorealistic visuals.
GPU Architecture
Fueling Accelerated Computing and Agentic AI With Unparalleled Performance, Efficiency, and Scale
NVIDIA Blackwell
The NVIDIA Blackwell architecture combines breakthrough AI, ray tracing, and neural graphics technology with massive performance and memory increases to drive cutting-edge professional creative, design, and engineering workflows from anywhere. NVIDIA RTX PRO Blackwell GPUs are the most advanced RTX GPUs offering high-performance, real-time ray tracing, and AI-accelerated compute to run agentic AI and neural rendering directly from your desktop.
NVIDIA Omniverse
Give AI agents tools to build, inspect, and prepare 3D worlds for physical AI simulation. NVIDIA Omniverse™ libraries help developers add RTX sensor simulation, GPU-accelerated physics, OpenUSD interoperability, and simulation-ready asset validation to existing applications and workflows, moving industrial digital twin and robotics content from visually complete to ready for virtual testing. Paired with NVIDIA Cosmos™ world foundation models, RTX-powered rendering helps generate the synthetic worlds that train physical AI.
