I think it's nice #nvidia tries to further research innovative tech, I'd exclusively buy their products, no matter the price, if they had better open standards practices

I mean they played a huge role, especially in GPGPU and up until today, hw accel, physics, video enc/dec, compute units, ray tracing, upscaling
like you guys wanna do cool tech ? awesome, but at least open this shit up a bit

atm that's more like *voodoo* magic in black box cards
I'll stick with #teamred in the meantime

There's close to no reason nvidia tech can't run on non-nvidia hardware.
Please don't contribute to nvidia locking away technology, whether it is algorithms, models, research, given how many abstraction layer there are, in the end it's just the software and data that matters
recent example: RTX5000 series 32-bits PhysX EOL
sure it's 2 decades old tech, I didn't assume they'd keep supporting the thing 'til the end of times
(despite how fun dropping a cop from the top of a building through plastic wrap in mirror's edge was)
I just wished they had released this tech as an open source standard in a lib and allowed that features on ATI cards (which btw would have eased the strain on devs) rather than a corporate hardware-exclusive clunky additional layered framework
also nvidia fucking stock is the worst hardware bubble of this decade, profits vastly from the whole fashionable AI situation, ain't gonna end well either way

also, letting all those (not) fancy proprietary firmware blobs and ML tech in the render pipelines stack is calling for drm's (bad ones) within the next 10 years in the GPU market
as in "pay that overpriced monthly nvidia subscription or we cut access to over the network real-time streamed encrypted DLSS model weights and leave you with glitchy unlit low polys for free tier"
they've already hooked the pipelines with remix

let's re-read #PirateCinema from Cory Doctorow
https://craphound.com/pc/download/

Download Pirate Cinema For Free | Cory Doctorow's craphound.com

AMD is barely better on that matter as they're ramping up on ML trying to catch up because they envy the nvidia stock big money