New details reveal that former NVIDIA CEO Jensen Huang used circular financing to fund AI startups and absorb excess TSMC wafers and memory rather than letting GPU prices drop
Semi-reluctant AMD ‘fanboy’ here, have hated Nvidia for almost a decade now.
All of this, where we are now with LLMs… this all started because Huang decided consumer grade GPUs can actually do full real time raytraced lighting, the movie grade shit that you normally do frame by frame with a farm of gpus.
That didn’t work well, so, Neural Upscaling. That was also fairly shitty, so, gen 2 gen 3 gen 4 gen 5, still too shitty, framgen, multiframegen! Which is also often still shitty.
Oh also, because Jensen says this is the new paradigm, uh, i guess fuck everything thats ever been done in terms of optimizing lighting in game engines, throw that all out, make it work with new paradigm.
Oh thats basically fundamentally impossible to do sanely and performantly, on affordable hardware? Clearly you just don’t understand how to use Unreal Engine, what do you mean you don’t own a GPU that costs as much as your last entire PC builf?
4k60fps ultra? Fuck you! That’ll be $4000.
(not counting fake frames…)
This is all Huang’s fault.
And thats without even going into creating the largest financial bubble in the history of mankind which will pop into the Greatest Depression in recorded history, very likely end the USD’s status as world reserve currency.
… so yeah, I semi-reluctantly ‘fanboy’ for AMD for a few reasons, some of those reasons being they didn’t initiate all of the above.
Anyone who is an Nvidia fanboy is definitionally an idiot that understands nothing about this industry.
You’re very much right. But I’ve become disillusioned even with AMD (or rather the Radeon part). They are in lockstep with whatever dumb direction nvidia decides, slap a slight discount on it and do the same shit.
They had shot down promising internal projects time and again. Why? To create the exact copy of what nvidia already did. (ray tracing, FSR, anti-lag…) They opted for one-gen-behind feature parity instead of offering something that would distinguish them and give an edge. (You know actual R&D, not copy and develop)
I was really hopeful for Intel shaking things up but their momentum ran out fast, they didn’t even release the second gen full-die card (which already targets mid-range) B770.
After the entire… userbenchmark + related shennanigannery debacle… I really can’t believe that I’m now actually rooting for Intel as an underdog, in any capacity.
Yet, here we are, rofl.
But hey, at least AMD continues to be waaaaaaay better at being open source than Nvidia.
Semi-reluctant AMD ‘fanboy’ here, have hated Nvidia for almost a decade now.
All of this, where we are now with LLMs… this all started because Huang decided consumer grade GPUs can actually do full real time raytraced lighting, the movie grade shit that you normally do frame by frame with a farm of gpus.
That didn’t work well, so, Neural Upscaling. That was also fairly shitty, so, gen 2 gen 3 gen 4 gen 5, still too shitty, framgen, multiframegen! Which is also often still shitty.
Oh also, because Jensen says this is the new paradigm, uh, i guess fuck everything thats ever been done in terms of optimizing lighting in game engines, throw that all out, make it work with new paradigm.
Oh thats basically fundamentally impossible to do sanely and performantly, on affordable hardware? Clearly you just don’t understand how to use Unreal Engine, what do you mean you don’t own a GPU that costs as much as your last entire PC builf?
4k60fps ultra? Fuck you! That’ll be $4000.
(not counting fake frames…)
This is all Huang’s fault.
And thats without even going into creating the largest financial bubble in the history of mankind which will pop into the Greatest Depression in recorded history, very likely end the USD’s status as world reserve currency.
… so yeah, I semi-reluctantly ‘fanboy’ for AMD for a few reasons, some of those reasons being they didn’t initiate all of the above.
Anyone who is an Nvidia fanboy is definitionally an idiot that understands nothing about this industry.
You’re very much right. But I’ve become disillusioned even with AMD (or rather the Radeon part). They are in lockstep with whatever dumb direction nvidia decides, slap a slight discount on it and do the same shit.
They had shot down promising internal projects time and again. Why? To create the exact copy of what nvidia already did. (ray tracing, FSR, anti-lag…) They opted for one-gen-behind feature parity instead of offering something that would distinguish them and give an edge. (You know actual R&D, not copy and develop)
I was really hopeful for Intel shaking things up but their momentum ran out fast, they didn’t even release the second gen full-die card (which already targets mid-range) B770.
Yeeeeep.
After the entire… userbenchmark + related shennanigannery debacle… I really can’t believe that I’m now actually rooting for Intel as an underdog, in any capacity.
Yet, here we are, rofl.
But hey, at least AMD continues to be waaaaaaay better at being open source than Nvidia.
Intentionally or unintentionally hahahahah!
Yo! I’m also a “fan” of his cousin’s company for the same reasons as you. The Nvidia stans have gaslit me into thinking I was some kind of freak
Your take needs to be in the history books that cover this fiasco
I still can’t beleive the AMD head lady Lisa Su is his cousin.
Reality needs better scriptwriters, this shit is too ridiculous and wacky to be real, but as far as I can tell… it is real!
Fuck man, I dunno, I’m going with “2012 was the apocalypse, we phase shifted into the clown dimension” at this point.
Imagine a world in which AMD bought Nvidia instead of ATI.
Imagine a world where AMD built their own GPU division and ATI remained independent
… why do you hurt me like this … ?