NVIDIA Officially Announces Turing GPU Architecture at SIGGRAPH 2018

Isaac Cain
August 16, 2018

The Quadro RTX GPUs work with Nvidia's Quadro Infinity software both of which are due to reach the market in the first quarter of 2019.

"Turing is Nvidia's most important innovation in computer graphics in more than a decade", said the company's CEO Jensen Huang at the SIGGRAPH conference in Vancouver, where he talked up the new technology. The new Turing technology can capitalize on this with promises of a new generation of hybrid rendering that enables cinematic-quality interactive experiences, wonderful new effects powered by neural networks and fluid interactivity on highly complex models.

Real-time ray-tracing of objects and environments with physically accurate shadows, reflections, refractions and global illumination.

NVIDIA's new Turing architecture uses hybrid rendering to enable real-time ray tracing for video producers. It can simulate the physical world six times faster than the Pascal architecture and can do final-frame rendering for film effects 30 times faster than CPU nodes, according to the release.

As for the tech powering these new cards, we've heard rumblings about the newest line of Nvidia GPUs for some time now. With the new product, Nvidia clocks its ray tracing capabilities at 10 gigarays a second, with more general performance at 16 teraflops. The platform also supports Nvidia's new open-source Material Definition Language software development kit, which is a suite of tools that integrate the precise look and feel of real-world materials into rendering applications. NVIDIA states that Turing has RT cores for accelerating ray traces and Tensor Cores for AI inferencing.

NVIDIA Officially Announces Turing GPU Architecture at SIGGRAPH 2018

Now, these RT cores will perform the ray calculations themselves at what NVIDIA is claiming is up to 10 GigaRays/second, or up to 25X the performance of the current Pascal architecture.

On top of this specialized ray tracing technology, Nvidia has also integrated the Tensor Cores seen in Volta to inject artificial intelligence into this GPU architecture.

The Tensor Cores accelerate deep-learning training and they provide up to 500 trillion tensor operations per second.

Thanks to Samsung's industry-leading 16Gb GDDR6 memory, end users can expect improved performance and energy efficiency in the widest array of graphics-intensive applications, including computer-aided design (CAD), digital content creation (DCC) and scientific visualization applications. Other applications include "denoising", resolution scaling and video retiming. Translated, "Aqui" means "Here" and plopping the coordinates into Google Maps reveals a location in Cologne, Germany, possibly where Nvidia's "Gaming celebration" will be held.

In any case, we'll know much more when NVIDIA's Gamescom event kicks off next week at 6 p.m. CET (1 p.m. ET) on August 20.

More news: Government wants to revoke operator’s contract after Genoa bridge collapse

Other reports by LeisureTravelAid

Discuss This Article