These Nvidia Turing architecture based GPUs feature new RT Cores which help accelerate ray tracing along with new Tensor Cores for the AI. The RTX 8000, RTX 6000 and RTX 5000 all feature Samsung's new 16Gb GDDR6 memory, which brings along improved performance and efficiency.
Nvidia is calling Turing "the greatest leap in computer graphics in over a decade", noting that it is focusing on realistic ray-tracing (incredible lighting effects) for its "RTX" range of cards.
With its new architecture, Turing, NVIDIA is making the bold claim that it has reinvented computer graphics.
"Hybrid rendering will change the industry, opening up awesome possibilities that enhance our lives with more attractive designs, richer entertainment and more interactive experiences", he said.
The chief executive of Nvidia Jensen Huang, announced the rollout of the company's first chipsets based on its Turing architecture, which will be in a line of products meant for use in designing games and other content.
Nvidia Turing will combine both its RT Cores and Tensor Cores to usher in a new form of hybrid rendering that includes both simulation and enhanced rasterization.
The company has announced that Quadro professional workstation graphics cards with Turing technology will be made available in Q4.
I'm sure we'll find out much more about Nvidia's new gaming graphics cards during the event next week. These specialized cores essentially compute how light and sound travel in a 3D environment at a rate of up to 10 GigaRays. Real-time ray tracing could lead to more faster renders and more effective work with the technique.
A render farm of 240 Dual 12-core Skylake Intel CPU servers costs about $2 million, according to Huang, compared to a comparable setup of four Nvidia RTX 8-GPU servers that would cost $500,000, take up about a tenth of the space and require one-eleventh of the power to operate. The Turning architecture supports artificial intelligence capabilities, has dedicated ray tracing cores, and new creative tools powered by neural networks.
"We have been able to operate and optimise performance across the stack, removing bottlenecks, and creating new ideas that break Moore's law". With up to 4,608 CUDA cores, Turing supports up to 16 trillion floating point operations in parallel with 16 trillion integer operations per second.
The Santa Clara, California-based company's shares have risen more than 50 per cent over the past year, backed by data centre chip sales that more than doubled to $1.9 billion its last fiscal year.