?
They're not doing this for people in the industry. This is their own agenda. You think they care what the industry wants when they're the only people they can turn to? You and the rest of the general populace romanticize this too much and it's why they get away with this. They'll make what they want as the future so long as it generates them enormous amounts of cash and enforces their monopoly. The onus is also on software developers which are forced to use the RTX API over the regular Microsoft DXR one just to get access to RTX AI denoising for barely usable performance on the current cards.
Don't get me wrong; ray tracing is the future, and NVIDIA wants to control how late that future is and how much money they can milk out of it.
If you didn't realize, we were always going to get accelerators for ray tracing. This is just how computing works. NVIDIA is just going to make this a very expensive and long process while also making their cards have some very extreme built in obsolescence. You don't go up to Intel or AMD and say "wow, you dedicated die space for an AES 256 accelerator just for my own security? Wow, aren't you amazing!"
You're falling really damn hard for these benchmarks. I just explained how they are comparing this to an old architecture which had no ray tracing accelerators to begin with. Any tiny amount of logic for ray tracing will give you an enormous improvement, but just because there's improvement doesn't mean the actual performance is usable. Besides, if you're working in this field, whichever company you're working for is supposed to have render farms anyways so it's not even any of your concern. You won't be getting your dream of a nice desktop raytracing rendering monster with Turing. You're going to have to wait another decade or two for that.
But hey, feel free to waste $700-$1200+ on a "ray tracing" card that'll just get outdated with the next card which will probably cost $800-$1500+ instead.
