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. NVIDIA Quadro RTX 8000 passive 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 Vulcan ray tracing, servers utilizing the NVIDIA Quadro RTX 8000 passive and NVIDIA Quadro vDWS (NVIDIA Quadro Virtual Data Center Workstation) GPU virtualization software powers 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.
Multi-Precision Tensor Cores
Purpose-built for deep learning matrix arithmetic at the heart of neural network training and inferencing functions, Turing GPUs include enhanced Tensor Cores that accelerate FP16 / FP32 matrix operations in addition to INT8 and two new INT4 and INT1 (binary) precision modes. Independent floating-point and integer data paths allow more efficient execution of workloads using a mix of computation and addressing calculations.
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 48GB of state-of-the-art GDDR6 memory delivering 55% greater throughput than the previous-generation GDDR5x technology, NVIDIA Quadro RTX 8000 Passive is well-equipped to handle latency-sensitive applications handling large datasets.
Error Correcting Code (ECC) on Graphics Memory
Meets strict data integrity requirements for mission critical applications with uncompromised computing accuracy and reliability for workstations.
Encode/Decode Engine Enhancements
H.264 encode bitrate/quality improvements to modulate performance to reduce bitrate (e.g. ~6% at the same performance level, ~11% with reduced performance).
Vidmem output for encoded bitstream and motion vectors to accelerate post-processing for end-to-end GPU workflows.
Optical flow at 4 x 4 granularity for improved video frame interpolation/extrapolation, object tracking, motion segmentation, action recognition, etc.
I-frame only decode for H.264 and HEVC to allow inferencing with I-frame only to speed use cases such as video surveillance.
VP9 10-/12-bit decode to enable VP9 HBD content decoding.
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.
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.