Nvidia has announced the availability of its Nvidia Quadro GV100 GPU with Nvidia RTX technology to artists and designers from all sectors, in what the company labelled the biggest advancement in computer graphics in nearly two decades.
"We're starting this year off with a big bang with what we believe is the biggest advancement of computer graphics in the last 15 years since Nvidia invented the programmable shader," Nvidia VP of Professional Visualization Bob Pette told press during a briefing at Nvidia GTC in San Jose.
According to Pette, it was clear to Nvidia that it had to continue to push the envelope as its customers continued to push the envelope. That meant extending the availability beyond game developers.
"They wanted photorealism, the democratisation of VR allowed them to do more in VR than they were ever capable of doing," he said of artists and designers. "They don't want to dumb-down their models, they don't want to decimate their models ... and they don't just want to look at it in standard colours, they want to look at it with photorealistic rendering, with photorealistic light -- they want to look at a car on a road in the rain at night just as the human eye would perceive it."
Pette said it was the work of Nvidia customers that has led the company to do a lot of work both within the GPU and significant amount of work with the software stack, noting that customers want the same level of photorealistic rendering that is typically seen in the movies.
"That takes hours and hours of CPU rendering just to get a single frame," he added. "It's about bringing all the goodness, all the features of GPU architecture, to desktop and our virtual products.
"Over 6 billion hours of rendering time producing billions of images all need to be done fast."
But it isn't just companies like Pixar that now require the technology; Pette said industries such as fashion and car manufacturing require image rendering before it can reach production stage.
"Nvidia RTX technology will, for the first time ever, enable true, real-time cinematic ray tracing. It's not only for gamers; it's for movie makers, designers, architects, for people building cars and for people building self-driving cars -- the industry impact of this is going to be phenomenal given the amount of computing that each of those industries spend on CPU rendering cycles today," he explained.
"In a workflow, it is now becoming way too long for them to be able to make the rapid changes that their customers are requiring."
He said Nvidia has been in development on RTX for a decade.
"Recreating VR is one of the most daunting tasks we know," CEO and founder Jensen Huang added. "It's beautiful, it's just super hard to produce.
"It's a gap we've been trying to close literally for four decades."
The company has also worked with Microsoft on its DirectX Raytracing technology and will be available on Vulkan, which is the PC maker's open, cross-platform graphics standard.
The Quadro GV100 with 32GB of memory is scalable to 64GB with multiple Quadro GPUs using Nvidia NVLink interconnect technology.
It's based on Nvidia's Volta GPU architecture and packs 7.4 teraflops of double-precision, 14.8 teraflops of single-precision, and 118.5 teraflops of deep learning performance.
Additionally, the Nvidia OptiX AI-denoiser built into Nvidia RTX delivers nearly 100x the performance of CPUs for real-time, noise-free rendering.
The Quadro GV100 is already available through Nvidia, and from next month through workstation manufacturers such as Dell EMC, HP, Lenovo, and Fujitsu.
Disclaimer: Asha Barbaschow travelled to GTC as a guest of Nvidia
- Nvidia redefines autonomous vehicle testing with VR simulation system
- Nvidia and Arm want to bring deep learning to your IoT projects (TechRepublic)
- Nvidia doubles down on AI
- Nvidia's Titan V giant: $3,000 buys you 'most powerful PC GPU ever'
- Nvidia morphs from graphics and gaming to AI and deep learning
- Nvidia expands new GPU cloud to HPC applications
- Windows 10: Movie-grade, real-time 3D graphics near with DirectX Raytracing API