This week at SIGGRAPH 2018 NVIDIA has proclaimed that computer graphics have made the greatest leap since the invention of the CUDA GPU in 2006.
NVIDIA has officially launched the NVIDIA Turing GPU Architecture, a new revolutionary class of graphics processing units.
Turing Class GPUs
Turing-class GPUs fuse together dedicated RT Cores with new Tensor Cores for AI inferencing, which together for the first time make real-time ray tracing possible.
NVIDIA has found a way to accelerate ray tracing rendering by up to 25x – 30x faster than its Pascal generation GPUs by tapping the power of specialized ray trace rendering engines engineered into the silicon of the chip. New AI-powered features further speed up, and importantly, enhance image quality, resulting in the ground-breaking performance.
RT Cores are dedicated silicon for ray tracing, engineered directly into the new GPU’s architecture. The two engines—along with a suite of more powerful technology for computing for simulation and enhanced rasterization—usher in a new generation of hybrid rendering to address the $250 billion visual effects industry.
“Turing is NVIDIA’s most important innovation in computer graphics in more than a decade,” said Jensen Huang, founder, and CEO of NVIDIA, speaking at the start of the annual SIGGRAPH conference. “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.”
Newly Unveiled Turing Class GPU Products from NVIDIA
The graphics giant also unveiled its initial Turing-based products—the NVIDIA Quadro RTX 8000, Quadro RTX 6000, and the Quadro RTX 5000—which will revolutionize the work of some 50 million designers and artists across multiple industries.
NVIDIA’s eighth-generation GPU architecture, Turing enables the world’s first ray-tracing GPU and is the result of more than 10,000 engineering-years of effort. By using Turing’s hybrid rendering capabilities, applications can simulate the physical world at 6x the speed of the previous Pascal™ generation.
You can learn more directly about each product above here.
What are RT Cores / What are Tensor Cores
RT Cores are dedicated ray-tracing processors in silicon on the GPU chip, which accelerate the computation of how light and sound travel in three-dimensional environments. The acceleration is up to 10 GigaRays a second.
The Turing generation GPUs crunch through ray tracing operations at up to 25x faster than the previous Pascal generation GPUs from NVIDIA. And final rendering for film effects is more than 30x faster.
“Cinesite is proud to partner with Autodesk and NVIDIA to bring Arnold to the GPU, but we never expected to see results this dramatic,” said Michele Sciolette, CTO of Cinesite. “This means we can iterate faster, more frequently and with higher quality settings. This will completely change how our artists work.”
Tensor Cores are specific areas of the chip dedicated to accelerating deep learning training and inferencing, providing up to 500 trillion tensor operations per second. The deep learning is targeted at learning anti-aliasing, which is a breakthrough in motion image generation, denoising, resolution scaling and video re-timing. All of these features are part of the NVIDIA NGX software development kit, a new deep learning-powered technology stack that enables developers to easily integrate accelerated, enhanced graphics, photo imaging and video processing into applications with pre-trained networks.
You can read the full press release here.