Live data from Hacker News

Nvidia’s $589B DeepSeek rout

finance.yahoo.com

111–120 of 1001 posts

Re: Nvidia’s $589B DeepSeek rout

#111

Earlier quoted context omitted.

While everything you say may be true, this shows a fundamental misunderstanding about how the modern stock market functions. How much value a company creates is at best tangential and often completely orthogonal to how much the stock is worth. The stock market has always been a Keynesian beauty contest, but in the past few decades, strongly shaped and morphed by attention economies. A good example of this is DJT, a c…

I understand that part, hence why this is very much buy the dip for me.

It's not just the usual herding behavior though. There's a convex response to news like this because people look at higher order effects like the growth of growth for stocks. Basically the DeepSeek story is about needing 40x fewer compute resources to run inference if their benchmarks are true. The dip doesn't mean that NVidia is now doomed, it simply means that if DeepSeek is legit, you need much less NV hardware to run the same amount of inference as before. Will the demand rise to still use up all the built hardware? Probably, but we went from a very stratospheric supply constraint to a slightly less stratospheric one, and this is reflected in the prices. Generally these moves are exaggerated initially, and it takes a bit of time for people to digest the information and the price to settle. It's an oscillating system with many feedback loops.

Re: Nvidia’s $589B DeepSeek rout

#112
post #32

Well this is very bad for the private AI companies but Big Tech will be fine, especially Meta and Apple. Everyone else who over-leveraged into the private AI companies (Anthropic, OpenAI) are going to have their valuations under scrutiny.

What does Apple have to do with AI

MSFT should be here instead of apple

Re: Nvidia’s $589B DeepSeek rout

#113

Good time to buy then, I don’t understand how stupid some traders can be. A more efficient model is better for NVIDIA not worse. More compute is still better given the same model. And as more efficient models proliferate it means more edge computing which means more customers with lower negotiating power than Meta and Google… This is like thinking that if people need to dig only 1 day instead of an entire month to ge…

> Good time to buy then

No. The stock is still x10 after this dip from 2 years ago and x40 from a few years ago.

Re: Nvidia’s $589B DeepSeek rout

#114
post #54

I find it interesting because the DeepSeek stuff, while very cool, doesn't seem invalidate that more compute wouldn't translate to even _higher_ capabilities? It's amazing what they did with a limited budget, but instead of the takeaway being "we don't need that much compute to achieve X", it could also be, "These new results show that we can achieve even 1000*X with our currently planned compute buildout" But perhap…

> doesn't seem invalidate that more compute wouldn't translate to even _higher_ capabilities?

That's how i understand it.

And since their current goal seems to be 'AGI' and their current plan for achieving it seems to be scaling LLMs (network depth wise and at inference time prompt wise), i don't see why it wouldn't hold.

Re: Nvidia’s $589B DeepSeek rout

#115
post #2

Nvidia -13% in Frankfurt stock market just now. Valuations of private unicorns like OpenAi and Anthropic must be in free fall. DeepSeek spends $6 million in old H800 hardware to develop open source model that overtakes ChatGPT. AI gets better, but profit margins sink with strong competition. Chinese AI startup DeepSeek overtakes ChatGPT on Apple App Store https://news.ycombinator.com/item?id=42839656 Edit: Nvidia now…

[flagged]

Please try to at least attempt to consider nuance. Do you seriously think that would happen? What is your point here? Do you think people in favor of restricting one thing are in favor of restricting everything?

Re: Nvidia’s $589B DeepSeek rout

#116
post #15

Earlier quoted context omitted.

Their model is open and they published paper describing it https://arxiv.org/pdf/2412.19437 The can't be far off, or it would be noticed. Even if they are heavily government subsidized for energy and hardware, I don see how the cost of training in the US would be more than double.

People on the internet can lie. Especially when such a lie could cause the nasdaq to dip multiple percent points. Not saying they are lying, but there incentives.

Much easier to identify the incentives of the people who just lost a lot of money who were betting on the idea that it was their money that was going to make artificial intelligence intelligent.

Everyone’s already begun trying this recipe in-house. Either it works with much less compute, or it doesn’t.

For instance, HKUST just did an experiment where small weak base models trained with DeepSeek’s method beat stronger small base models being trained with much more costly RL methods. Already this seems like it is enough to upend the low end models niche market, things like haiku and 4o-mini.

Be really skeptical why the people who should be making tons of money by realizing actually it was all a mirage and that they can now get the real stuff for even cheaper, would spend so much effort shouting about this, in order to undercut their own profitability..

Re: Nvidia’s $589B DeepSeek rout

#117
post #78

Not sure what the fuss is. I tried Deepseek earlier today for the first time and it was even worse than o1 when it came to reasoning skills and following my requests for how I wanted to engage with it. o1 at least gives it to me straight. When I ask it to engage in more back and forth before assuming what I'm after, it tends to follow through. Deepseek seemed immediately eager to (very slowly) feed me a bunch of made…

The excitement isn't the capabilities of the model, it's how efficiently it was created. One of the major lessons in AI in the last couple of years was that scale mattered - you would want to throw more and more compute at a problem and that has turned into incredible share prices for Nvidia and incredible investments in data centre and energy generation. If it turns out that actually we didn't need quite such incredible scale to get these results and actually we were just missing some really quite basic efficiency optimizations then the entire investment cycle into Nvidia, data centres and energy generation is going to whipsaw in an incredible way.

Re: Nvidia’s $589B DeepSeek rout

#118
post #84
post #55

Earlier quoted context omitted.

Put more simply: The stock market is not a reflection of the economy. Wall Street is not Main Street.

It's a reflection of expectations about the future economy. Obviously, such expectations are not always accurate because humans are quite fallible when trying to predict the future. This is even more true when there is a lot of hype about a certain product. Yesterdays price of (say) NVidia was based on the expectation that companies would need to buy N billion of USD of GPUs per year. Now Deepseek comes out and makes…

> OpenAI and Anthropic, who now suddenly have to compete against a competitor that can produce the same product at (apparently) a fraction of the price.

OpenAI and Anthropic can react by adopting DeepSeek's compute enhancements and using them to build even better models. AI training is still very clearly compute-limited from their POV (they have more data than they know what to do with already, and training "reasoning"/chains-of-thought requires a lot of reinforcement learning which is especially hard) so any improvement in compute efficiency is great news no matter where it comes from.

Re: Nvidia’s $589B DeepSeek rout

#119
post #78

Not sure what the fuss is. I tried Deepseek earlier today for the first time and it was even worse than o1 when it came to reasoning skills and following my requests for how I wanted to engage with it. o1 at least gives it to me straight. When I ask it to engage in more back and forth before assuming what I'm after, it tends to follow through. Deepseek seemed immediately eager to (very slowly) feed me a bunch of made…

It can be a win for nvidia but bad for nvidia stock if the win is smaller than the initially expected win.

The business is doing good, the wannabe traders not so much

Re: Nvidia’s $589B DeepSeek rout

#120

Good time to buy then, I don’t understand how stupid some traders can be. A more efficient model is better for NVIDIA not worse. More compute is still better given the same model. And as more efficient models proliferate it means more edge computing which means more customers with lower negotiating power than Meta and Google… This is like thinking that if people need to dig only 1 day instead of an entire month to ge…

maybe Nvidia is fine but I don't understand this logic. suppose it turns out GPUs are unnecessary at all, but they still provide a performance boost, but you can do everything you can do right now with CPUs w.r.t performance. would that be good or bad for Nvidia? unless it can be said we need more performance than is currently possible, e.g. new demand, it would be catastrophic. it is unclear that throwing more compu…

I could see arguments be made in both ways here. If GPUs end up being more efficient/powerful (like today) it could induce even more demand, but also if CPU gets within ~20% of how fast you can do something with a GPU, people might start opting for something like Macs with unified memory instead of GPUs.

Today a CPU setup is still nowhere near as fast as a GPU setup (for ML/AI), but who knows how it looks like in the future.

> it is unclear that throwing more compute actually expands what is possible

Wasn't that demonstrated to be true already in the GPT1/2 days? AFAIK, LLMs became a thing very much because OpenAI "discovered" that "throwing more compute (and training data) at the problem/solution expands what is possible"

Post reply on HN