The cat’s finally out of the bag as Nvidia; after all the hype, rumors, and unhealthy speculations that ran through weeks and months – have made the huge news breakout in a rather well planned marketing movement that ran through this week.
Monday – August 13, the world stood still as the green team finally released the cat out of the bag in the SIGGRAPH professional graphics conference in Vancouver. Huang, Nvidia CEO; has unveiled Turing – Nvidia’s 8th generation graphics architecture that’s bringing ray tracing to real-time graphics to the table. Sure a new architecture is pretty cool, but how about hardware? Well, they have revealed the first GPUs to carry the new generation of architecture. the ($2,300) Quadro RTX 5000, ($6,300) Quadro RTX 6000 and ($10,000) Quadro RTX 8000 – These cards might be enterprise grade graphics hardware but this has kept the hype train going because as the new gen of Nvidia hardware and architecture is unveiled and confirmed this now means that the consumer grade cards are soon to follow.
Dell EMC, HP, Inc., Hewlett Packard Enterprise, Lenovo, Fujitsu, Boxx and SuperMicro are the confirmed vendors of the new gen Quadro cards that will be available starting the 4th quarter of this year.
“It has to be amazing at today’s applications, but utterly awesome at tomorrow’s,” Huang said.
Turing features new RT cores and Tensor cores that accelerate ray tracing and AI referencing – packaged together to make real-time tracing possible. Nvidia has announced that they are also open sourcing the Material Definition Language software development kit.
“It turns out it was running on this — one single GPU,” Huang said to wild applause as he playfully blinded the camera by angling the gleaming Quadro RTX 8000’s reflective outer shroud. “This is the world’s first ray-tracing GPU.” – Imagine running a ray-tracing demo that was previously powered by 4 Volta GPUS to only a single Turing GPU? That’s what happened during the show of force by playing the stunning Star Wars-themed Reflections ray-tracing demo.
The Turing architecture was made for the future – this was given extreme emphasis during the reveal and how was this displayed you asked? The Tensor cores of Turing is responsible to accelerate deep learning training and inferencing – about 500 trillion tensor operations per second to be exact; and you can harness that power through AI enhanced features through the NVIDIA NGX software development kit.
“At some point you can use AI or some heuristics to figure out what are the missing dots and how should we fill it all in, and it allows us to complete the frame a lot faster than we otherwise could,” Huang said
Turing cranks up the notch for rasterization – as the mainstay of graphics and being responsible for operations like variable-rate shading, texture-space shading and multi-view rendering; being 6x faster than Pascal – this provides more headspace for fluid operations with larger scaled models and scenes
Though everything looks nailed to enterprise Turing cards currently this has shed more light and bring the hype to a better position in the hopes for the consumer cards to play out in the market. But expect Nvidia to roll out the 80s and 70s series of card first just as they unveiled the higher enterprise grade cards first. Is this enough for you to whip out the green profile picture/banner? Or at least release the green army in you with a simple #BeForTheGame social media barrage?
One thing is certain, All the hype and rumor was somewhat true and indeed have the tables Turing’ed