Nvidia isn't hurting for customers and they can barely keep up with demand. So yes, you're overthinking it.
>>476016>Nvidia isn't hurting for customers and they can barely keep up with demand.
That's what every capitalist wants, it means higher profit margins, why would they be ok with letting that go ?
>>476017>I have heard certain policies the US has taken called a "controlled demolition" of their economy.
That's a fitting expression but what is causing them to want to demolish their economy ?
>I think whoever decided USA policy either ate their own propaganda and believed it.
They all drank the cool-aid ? There is nobody else left ? Wow that's scary shit.
>But as you said if this isn't in the usual playbook for imperialists, then the imperialists have now recognized that they cannot bully China.
That would be a fair analysis if they weren't also surrounding China with a string of military bases.
>>476043>why would they be ok with letting that go ?
He's saying they're shut out of a market they can't really exploit right now anyway. That's why they don't care.>>476045>what is causing them to want to demolish their economy ?
Globalists don't want strong independent nation states they want weak failed states who cannot resist being absorbed into a bigger political structure.
>That would be a fair analysis if they weren't also surrounding China with a string of military bases.
We can't stop the military industrial complex. They get their money regardless of how much their actions help or hurt the country.
>dedicated AI chips
I'mma stop you right there OP. What the hell is an "AI chip"? Is this just another buzzword from Silicon Valley hucksters?
It's a processor that accelerates training models. Nvidia's H100 for example. But there are also chips inside phones that help make things like speech synthesis and video processing faster, which is using AI models.
I fail to see what makes one of these different than any other processor.
It has circuits that lets them do certain operations involved in AI much faster. In general hardware is often built to speed-up specific software, like video cards are created in response to games and rendering becoming a thing.
I guess the difference between video cards and AI processors is that you can buy a video card and AI accelerators are either chips developed internally by companies like Google and aren't for sale, or they are super expensive like the nvidia chips. So it's kind of hard to play with this technology at home.
>>476147>I fail to see what makes one of these different than any other processor.
special purpose coprocessors are older than your boomer ass
operations implemented in hardware are always more efficient than operations implemented in software
I guess you don't believe in GPUs either?
Well lets give that anon the benefit of doubt.
AMD's APUs i.e. processors with really powerful integrated GPUS might eventually get good enough to kill dedicated GPUs
Computers used to have dedicated sound cards and network-adapter cards for decades, you know.
Maybe that anon means that the dedicated machine-learning accelerator hardware in AI-Chips gets added to processors and that just becomes another part of general purpose processors.