![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
I owned one
. It was a good card, and nVidia's 5 series FX cards were bad. However, as good as it was, it wasn't "complete different performance class" is was just superior. And, of course, nVidia had a better product with the 6th gen cards. I mean we saw it kind of happen again with the 480 and 5870, but again, it didn't kill nVidia.
Dear gosh were the launch drivers disastrous on those. I returned mine and went 470 sli.
For sure they were. I didn't return mine, but I did end up getting a 580 (or was it 570, can't remember) when it came out because damn.
My memory's faulty... googled myself and I sold it, not returned
. I made a thread here about the driver glitches I was having, it was crazy.and for investors. Whether it be the big holders, whom they actually answer to. Or individuals, whom may FOMO their way into buying some shares, now.
I wouldn't call it serious competition. Competition, but not too serious. NV holds a huge lead and doesn't appear to be complacent as far as I can tell. Could they falter, sure. At this moment they are the market leader for the foreseeable future.
This has been talked about in other thread topics. But it also relates to "what do AI researches need to continue to do AI development."
This enters the idea of general-purpose AIs vs research AIs like the Facebook LLM and their dedicated accelerators are inferior in every measurable way to the Nvidia offerings, but they are 1/5'th the price, and while completely proprietary they are also not a source of any measurable bottleneck.
I'm not so sure that is true. Thing with the AI startups. There is so much money flying around cause everyone knows its an arms race.
The issue with your point here is that you first have to express what is the goal of all of these various companies buying H100 chips is.
Again, beat who to do what? The whole point of my post was discussing that the direct issue is that a lot of people bought into AI and they don't know why they did it. Just like the .com bust, a bunch of companies just made websites with no real purpose or point and no real way of generating money through their site.AI... should someone get to the holy grail general intelligence they win. That is the problem for us gamers right now. The AI rush at the high end isn't just about Microsoft getting a better chat bot... or some other company building a tool to make 10s high end company logos from a doodle. The big money in the field feel there is no second place finish. No one uses Yahoo search... even Bing struggles cause Google is Google. Second place is not an option even if the less you spend the more you save.
Like a "Dual Use" export control on GPU's like similar export controls on Thermal Cameras and (Maybe?) Drones?
Forgot the /s
I believe the same logic applies however. Tesla is not the only company attempting to create a self driving car. I don't believe coming in second in that race is seen as an option by anyone in the race either.
As much of a prick he is, you have to respect him. This's a madddd mann filled with passionate lust and undying thirst for building empires. He won't stop at nothing until the goal is completed. That's the kind of person I wanted to work with achieving domination and building a legacy. https://videocardz.com/newz/jensen-huang-wants-to-lead-nvidia-for-another-30-to-40-years
AMD doesn’t make top tier graphics cards? 7900 XTX beats the 4090 at some games, impossible for them to trade blows if they aren’t top tier. Not to mention AMD’s gpus are cheaper as well. Price to performance does matter.
I had several hundred shares of nvda. Purchased them at $9 per share and sold at a little more than $27 per share in part to put down a downpayment on my current house. I wish I would have kept those.
![Zarathustra[H]](https://cdn.hardforum.com/data/avatars/m/9/9298.jpg?1707758471){kind=avatar}
Go back a few years and I'm sure he said they're moving to a Crypto Metaverse. Nvidia is a company that moves towards fads, and right now it's AI. They're a graphics card company, when the AI fad fades. Either because people have their fill of AI, or because competitors caught up and flooded the market with cheap AI hardware. Nvidia never wants to go back to be a graphics company because literally everyone is making their own GPU's. Also because the past several years their growth was heavily tied to crypto, and going back to their roots means a sharp decline in revenue. The AI situation is so stupid that even OpenAI can't decide if they want to hire their CEO. Give it time, the AI market will crash and Nvidia will be once again a graphics company. At least until they find another niche to try and capitalize on.
I am pretty sure it is an old email (probably around when they decided stuff about the Turing architecture circa 2016), back when crypto was the big fad, it would be dishonest to say Nvidia was not ahead of the curve with the competition about AI potential or thinking they just jumped on it.
He has those old-schooled Taiwanese semi-capitalist/communist family oriented mindsets. To me, he's both workaholic and excellent visionary (sharp and cunning as fox) who's willing to go to extreme extent to achieve his goals.
This really is not true right now, or for the forseeable future. The majority of games still ship with visual effects meant for non-RT graphics cards and will continue to do so for a long time. In those games (the majority of games) RT is additional.
It will start with Unreal 5 lumen first, there at least one case that we saw some Unreal 5 title being Lumen only I think, I feel we saw it with some of those recent Unreal 5, time saving for dev but less optimized for player game trade-off made possible for small studio to do quite impressive stuff.
What you're referring to is first mover advantage. And yes, it gives an advantage. But every major market has competitors that are in it.
That's great, except by this standard not everyone can win. So it's a zero sum game. And neither I nor you think that that is really true. There's a lot of people including myself that try to limit the Google products we use. And I don't use Microsoft's products really anymore either. And thank God for alternatives.
Apple will still be a computer company and Nvidia will still be a graphics company. Apple's phone sales dwarf their computer sales, but who's to say that'll always be the situation? Remember when tablets were going to displace PC's, and become the future? Apple was a computer company, who became a portable music player company, then a phone company, then a tablet company, back to a phone company. The tablet thing isn't working out so well. Nvidia went from being a graphics company to a crypto company, to an AI company. All of which is fueled by their graphics development. You will likely see them being a graphics company once again, unless there's another use for GPU's that Nvidia discovers? At which point Intel will also make graphics, and will also take a slice of that pie.
The big orders are done, they sold just short of 600,000 H100’s this quarter, the big players have what they need, Blackwell will be slower coming out and heavily constrained because of all the new logistics and such. Q1 and Q2 are highly unlikely to see these sorts of figures.
The overlap silicon wise however is the issue. 8 Billion is great and all... but if 2 billion of that is chips that could have made them 8 billion had they been put in data center cards the choice is clear.