[rl3]

The Turing architecture is also used in Quadro RTX cards, and those have a ridiculous amount of VRAM.

Is there any professional/computational use for these RT cores beyond raytracing?

One case that comes to mind is perhaps raytracing acoustics, and although interesting it's technically still raytracing.

As far as gaming is concerned, personally I'd love if the RT cores could contribute—however inefficiently—to rendering workload in non-RTX games. It's annoying that 50% of the die is allocated to hardware that requires feature-specific implementations.

[jsheard]

The whitepaper brings up physics simulation and occlusion/visibility testing as possible non-raytracing applications of the RT hardware, plus acoustic simulation as you said.

Page 30: https://www.nvidia.com/content/dam/en-zz/Solutions/design-vi...

[kbwt]

> It's annoying that 50% of the die is allocated to hardware that requires feature-specific implementations.

That's the future. While we may be able to cram more transistors onto "7nm" chips, only a tiny fraction of the chip area can be powered on because leakage currents are no longer decreasing with transistor size [1]. Hence Apple's Neural Engine and Nvidia's RTX. You have to waste the extra transistor count on something specialized.

[1] https://semiengineering.com/is-dark-silicon-wasted-silicon/

[rl3]

The RT and Tensor cores are primarily intended to power raytracing and DLSS, respectively. Both of those features will be used in conjunction with traditional shading/compute units, so the entire die is utilized at once, at least at a high level.

It would be really interesting to know the power consumption of a Turing card maxing out just its shading units, versus full utilization with RTX/DLSS.

[twtw]

> It's annoying that 50% of the die is allocated to hardware

Source on 50%?

[rl3]

Roughly* 50%:

https://i.kinja-img.com/gawker-media/image/upload/t_original...

[twtw]

A) that's a marketing image. The green box is placed over a random portion of the die, not just the raytracing units. (They are distributed throughout the die, not shoved on one side).

B) that said, looks more like 30% to me.

[rl3]

Good point. Here's another marketing image that illustrates what you're saying:

https://i.redd.it/gxmyxo2c5yf11.png

Just by looking at that you can intuitively see that the colored boxes in the previous image appear to be arbitrary. On the other hand, the previous image does seem to be trying to convey that a substantial portion of the die isn't shading units, even if the distribution is misrepresented.

>B) that said, looks more like 30% to me.

Was talking about RT and Tensor combined, not just RT.