some of that might be that their new 3k canadian graphics board is an okay, just okay upgrade on last gen in terms of performance
Nah, the 50 series is still a fair bit better than 40 series.
meh. the 30 to the 40, that's a significant jump. I don't call ~25% significant for that kind of dough
Market figured out that cheaper chips = lower CapEx hits for companies like Apple and Zuckturd. Those stocks are now nicely green.
/positions
ya I get it, the meh was meant only in the scenario of somebody who has the 4090 already and would be in that financial bracket. it hasn't impressed reviewers off the hopGamers weren't the reason Nvidia was worth 2.9 Trillion dollars though. AI firms are burning VC money, they're not using off the shelf cards. A new architecture capable of a 25% performance uplift is (well...was) worth it for them to upgrade into that architecture, and that's what fueled Nvidia's sky, sky high valuation.
Somewhat ironically, here's google's gen AI blurb on it:
View attachment 24003
That's why Deepseek is so dangerous to Nvidia. Apparently it didn't require Nvidia's newest and best tech for it's development and it has trimmed down models that you can run on a laptop. I saw a guy on youtube running a model using the laptop version of the 4070, which probably means the desktop 4060 or 4060ti can run it, as well as the 3070, 3080, and 3090. I've got a project I'm wrapping up this week and when I'm done, I'm going to take a run at trying that same model on a 1080ti and 2070 super to see how old you can really go.
American Energy Tech is underrated. At least if you listen to a guy like Chris Sacca.
Admittedly, I've only heard him speak a handful of times so my understanding of how deep he is into the tech end of it is super limited. But the few times I've seen him speak on energy it's about how much better a space energy is for VC investing in it is now, less capex intensive, faster to profit etc.
He sounds just like every other VC I've heard speak, who vastly overstates the importance of their role in the process, and vastly understates the value of the actual science.
I don't think that's going to win here, just like it didn't win solar, and now apparently, AI.