Nvidia unveils new GPU architecture for computer graphics rendering

The Turing Architecture combines ray-tracing and AI inference for a new kind of hybrid rendering.
Written by Stephanie Condon, Senior Writer

NVIDIA Turing Architecture

Nvidia on Monday unveiled a new GPU architecture, designed for the $250 billiion visual effects industry, to power rich, cinematic-quality interactive experiences. The Turing GPU is the first ever GPU expressly for ray tracing -- a rendering technique that creates realistic lighting effects.

The Turing architecture combines dedicated hardware acceleration of four core elements: AI, ray tracing, programmable shading and simulation. It features new RT cores to accelerate ray tracing. These dedicated ray-tracing processors accelerate the computation of how light and sound travel in 3D environments at up to 10 GigaRays a second. Turing accelerates real-time ray tracing by up to 25x that of the previous Pascal generation.

The Turing architecture also features new Tensor cores for AI inferencing, providing up to 500 trillion tensor operations a second.

The combination of RT cores and Tensor cores makes real-time ray tracing possible.

"This fundamentally changes how computer graphics is going to be done," Nvidia founder and CEO Jensen Huang said at the SIGGRAPH conference in Vancouver, Canada.


Nvidia also announced its first Turing-based products -- the Quadro RTXTM 8000, Quadro RTX 6000 and Quadro RTX 5000 GPUs.

"Turing is Nvidia's most important innovation in computer graphics in more than a decade," Huang said in a statement. "Hybrid rendering will change the industry, opening up amazing possibilities that enhance our lives with more beautiful designs, richer entertainment and more interactive experiences. The arrival of real-time ray tracing is the Holy Grail of our industry."

With Turing's hybrid rendering capabilities, applications can simulate the physical world at 6x the speed of the previous Pascal GPU generation.

Nvidia also announced the Quadro RTX Server, a reference architecture for highly configurable, on-demand rendering and virtual workstation solutions from the datacenter.

Quadro GPUs based on Turing will be initially available in the fourth quarter.

Editorial standards