Revolutionary Realtime Ray Tracing Acceleration

Incorporating new hardware-based ray tracing engines, Turing is the industry’s first graphics processor enabling real-time ray tracing. A single NVIDIA Quadro RTX 5000 board can render complex professional models with physically accurate shadows, reflections, and refractions to empower users with instant insight. Working in concert with applications leveraging APIs such as NVIDIA OptiX, Microsoft DXR and Vulkan ray tracing, systems based on NVIDIA Quadro RTX 5000 will power truly interactive design workflows to provide immediate feedback for unprecedented levels of productivity.

RT Cores

New dedicated hardware-based ray-tracing technology allows the GPU for the first time to real-time render film quality, photorealistic objects and environments with physically accurate shadows, reflections, and refractions. The real-time ray-tracing engine works with NVIDIA OptiX, Microsoft DXR, and Vulkan APIs to deliver a level of realism far beyond what is possible using traditional rendering techniques. RT cores accelerate the Bounding Volume Hierarchy (BVH) traversal and ray casting functions using low number of rays casted through a pixel.

Enhanced Tensor Cores

New mixed-precision cores purpose-built for deep learning matrix arithmetic, delivering 8x TFLOPS for training, compared to previous generation. NVIDIA Quadro RTX 5000 utilizes 384 Tensor Cores; each Tensor Core performs 64 floating point fused multiply-add (FMA) operations per clock, and each SM performs a total of 1024 individual floating point operations per clock. In addition to supporting FP16/FP32 matrix operations, new Tensor Cores added INT8 (2048 integer operations per clock) and experimental INT4 and INT1 (binary) precision modes for matrix operations.

Mesh Shading

Provides a compute-based geometry pipeline to speed processing and culling for geometrically complex models and scenes to improve performance by up to 2x.

Motion Adaptive Shading

Offers more granular control over how GPU horsepower is distributed (i.e. more cycles applied on the detailed areas of a scene and fewer on the less detailed areas) to increase performance and at the same image quality, or produce similar image quality with a 50% reduction in the time required to generate shaded pixels.

Variable Rate Shading (VRS)

More control over pixel shading rate; efficient for effects like motion, blur, foveated shading. This capability enables shading and geometry samples to process at different rates for more efficient execution.

Texture Space Shading

Decouples shading from screen space, improving shading efficiency and reuse.

Multi-View Rendering (MVR)

Extension of Single Pass Stereo rendering multiple views in a single pass with unique view origin positions or view directions.

High Speed GDDR6 Memory

Built with Turing’s vastly optimized 16GB GDDR6 memory subsystem for the industry’s fastest graphics memory (448 GB/s peak bandwidth), NVIDIA Quadro RTX 5000 is the ideal platform for latency-sensitive applications handling large datasets. NVIDIA Quadro RTX 5000 delivers greater than 50% more memory bandwidth compared to previous generation.

Error Correcting Code (ECC) on Graphics Memory

Meets strict data integrity requirements for mission critical applications with uncompromised computing accuracy and reliability for workstations.

H.264 and HEVC Encode/Decode Engines

Deliver faster than real-time performance for transcoding, video editing, and other encoding applications with two dedicated H.264 and HEVC encode engines and a dedicated decode engine that are independent of 3D/compute pipeline.

TURING VR

Ray tracing and variable rate shading enhance visual quality, while Multi-view rendering provides wider field of view, support for next-generation HMDs and displays. The RT Cores provide for accurate acoustic simulations and VirtualLink makes VR setup easier than ever with a single cable connection.

Full-Scene Antialiasing (FSAA)

Dramatically reduce visual aliasing artifacts or “jaggies” with up to 64x FSAA (128x with SLI Mode) for unparalleled image quality and highly realistic scenes.

32K Texture and Render Processing

Texture from and render to 32K x 32K surfaces to support applications that demand the highest resolution and quality image processing.

NVIDIA RTX Platform for Realtime Ray Tracing

Software framework that makes realtime ray tracing possible, portable, and presentable.

Provides interoperability between rasterization, ray tracing, compute and AI/Deep Learning.

New Turing ray tracing acceleration in OptiX, DXR and Vulkan.

NVIDIA MDL, now open source, and support for Pixar's Universal Scene Description (USD) promote portability and consistency.

NVIDIA NGX: AI for Visual Applications

Provides a neural graphics framework for image processing.

Brings AI-based features to end user applications to enhance graphics, photos, imaging, and video processing.

NGX features utilize Tensor Cores to maximize performance and efficiency.

Allows NVIDIA to bring advances in AI features to developers who place them into applications with NVIDIA pre-trained neutral networks.

NVIDIA NGX Technology

AI InPainting allows the removal of existing content from images and replaces it with realistic computer-generated alternatives.

AI Up-Res increases the resolution of an image or video by 2x, 4x, or 8x using AI to create new pixels by interpreting the image and intelligently placing data in the new image.

DLSS (Deep Learning Super Sample) removes jagged lines to smooth them, producing a higher quality image than by using other techniques

AI Slow-Motion inserts interpolated frames into a video stream to provide smooth, slow-motion video.

Software Optimized for AI

Deep learning frameworks such as Caffe2, MXNet, CNTK, TensorFlow, TensorFlow RT and others deliver dramatically faster training times and higher multi-node training performance. GPU accelerated libraries such as cuDNN, cuBLAS, and TensorRT deliver higher performance for both deep learning inference and High Performance Computing (HPC) applications.

NVIDIA CUDA Parallel Computing Platform

Natively execute standard programming languages like C/C++, Fortran, Python and APIs such as CUDA, OpenCL, OpenACC and Direct Compute to accelerate techniques such as ray tracing, video and image processing, or finite element analysis or computational fluid dynamics (among others).

Unified Memory

A single, seamless 49-bit virtual address space allows for the transparent migration of data between the full allocation of CPU and GPU memory.

NVIDIA GPUDirect for Video

GPUDirect for Video speeds communication between the GPU and video I/O devices by avoiding unnecessary system memory copies and CPU overhead.

NVIDIA NVLink High Speed Interface

Connect a pair of NVIDIA Quadro RTX 5000 boards with NVLink to essentially double the amount of GPU memory available, double CUDA, RT and Tensor Core counts, and scale application performance by enabling GPU-to-GPU data transfers at rates up to 100 GB/Sec.