Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Nvidia’s $589B DeepSeek rout

finance.yahoo.com

Post reply on HN
Print view View original

541–550 of 1001 posts

Re: Nvidia’s $589B DeepSeek rout

#541

Post by jiggawatts »

quesera wrote: post #534
jiggawatts wrote: post #500

Earlier quoted context omitted.

I recently went to the LLM chat arena and tried my "test input" against the latest frontier models that GPT 3 failed on. This test snippet simply repeats the same four-letter word in a paragraph many times using all of its various possible meanings simultaneously. The request to the AI is to put the meaning of each usage of the word next to it in brackets. None of the frontier models can do this perfectly. They all s…

> This test snippet simply repeats the same four-letter word in a paragraph many times using all of various possible meanings simultaneously This sounds like fun. How does it do with an arbitrary quantity of "buffalo"s?

That's a known thing that would be in its training set.

I just made up my own thing that no AI model would have seen anywhere before.

It's pretty easy to create your own, just pick a word that is highly overloaded. It helps if it is also used as proper names, business names, place names, etc...

Re: Nvidia’s $589B DeepSeek rout

#542

Post by TheAlchemist »

gitfan86 wrote: post #521
TheAlchemist wrote: post #474

Nvidia valuation = $3000B Nvidia profits during AI madness last year = $65B (last 4 quarters) Nvidia profits during normal year = $5B This stock could drop 90% from here, and still be expensive. The numbers are absolutely crazy and make no sense at all. Stargate project is aiming to invest $500B over 4 years. Those $500B are a pipe dream, but let's suppose for a second that all of that $500B will be Nvidia profits an…

Your numbers are wrong. The last quarter profit was $19B and the projected profit is 21B next quarter. That is 84B/year profit with zero growth. If META,STARGATE,xAI,etc.. all increased spending rapidly you could get to 200B profit rate in 2026. 3T / 200B = 15 That means they could return a 6.6 dividend which is higher than 10 year bonds and is in no way overvalued by historial standards. All that to say, I sold all…

You're right about the profit - I took last fiscal year. Still, it doesn't change anything in what I wrote.

What you just wrote is "IF the biggest companies on earth, and the US government decide to spend all of their money on a single chip maker, then you could get to 200B profit rate in 2026". I won't disagree with that.

Re: Nvidia’s $589B DeepSeek rout

#543

Post by lvl155 »

jhickok wrote: post #439
zozbot234 wrote: post #400

Earlier quoted context omitted.

Not surprising there - a maxed out Mac Studio is a great AI homelab, giving you way more bang for the buck than nVidia offerings.

Think that was the cause of their stock increase? I feel like investors use opportunities like this to pile money into safer bets rather than just bail on stocks altogether.

No of course not. Indexing in 2025 is a whole different world. It was most likely people covering Apple short they had paired with long QQQ.

Re: Nvidia’s $589B DeepSeek rout

#544

Post by CamperBob2 »

junon wrote: post #319
polygamous_bat wrote: post #225

Earlier quoted context omitted.

o1 does not show the reasoning trace at this point. You may be confusing the final answer for the reasoning trace in the middle, it's shown pretty clearly on r1.

I wasn't really referring much to the UI as I was the fact that it does it to begin with. The thinking in deepseek trails off into its own nonsense before it answers, whereas I feel openai's is way more structured.

All you get out of o1 is

    Reassessing directives

    Considering alternatives

    Exploring secondary and tertiary aspects

    Revising initial thoughts

    Confirming factual assertions

    Performing math

    Wasting electricity
... and other useless (and generally meaningless) placeholder updates. Nothing like what the output from DeepSeek's model demonstrates.

As Karpathy (among others) has noted, the output shows signs of genuine emergent behavior. Presumably the same thing is going on behind the scenes in the OpenAI omni reasoning models, but we have no way of knowing, because they consider revealing the CoT output to be "unsafe."

Re: Nvidia’s $589B DeepSeek rout

#545

Post by DennisP »

Analemma_ wrote: post #371
tgtweak wrote: post #354

Earlier quoted context omitted.

Hype buyers are also Hype sellers - anything Nvidia was last week is exactly what it is this week - DeepSeek doesn't really have any impact on Nvidia sales - Some argument could be made that this can shift compute off of cloud and onto end user devices, but that really seems like a stretch given what I've seen running this locally.

I agree hype is a big portion of it, but if DeepSeek really has found a way to train models just as good as frontier ones for a hundredth of the hardware investment, that is a substantial material difference for Nvidia's future earnings.

Or Nvidia keeps its earnings and our best frontier models get a hundred times better.

Re: Nvidia’s $589B DeepSeek rout

#546

Post by KiwiJohnno »

donsupreme wrote: post #478
KiwiJohnno wrote: post #399

Earlier quoted context omitted.

Not quite, I believe this sell off was caused by DeepSeek showing with their new model that the hardware demands of AI are not necessarily as high as everyone has assumed (as required by competing models). I've tried their 7b model, running locally on a 6gb laptop GPU. Its not fast, but the results I've had have rivaled GPT4. Its impressive.

> I've tried their 7b model Anything other than their 671b model are just distilled models on top of Qwen and Llama using their 671b reasoning data output, right?

Correct. Its the best model I've been able to run locally, by a long shot

Re: Nvidia’s $589B DeepSeek rout

#547

Post by llm_trw »

int_19h wrote: post #520
elihu wrote: post #450

Earlier quoted context omitted.

Jevon's paradox would imply that there's good reason to think that demand for shovels will increase. AI doesn't seem to be one of those things where society as a whole will say, "we have enough of that; we don't need any more". (Many individual people are already saying that, but they aren't the people buying the GPUs for this in the first place. Steam engines weren't universally popular either when they were introdu…

The other thing is that if this pushes the envelope further on what AI models can do given a certain hardware budget, this might actually change minds. The pushback against generative AI today is that much of it is deployed in ways that are ultimately useless and annoying at best, and that in turn is because the capabilities of those models are vastly oversold (including internally in companies that ship products wit…

An rag model can already sort your email.

Its just that it costs too much to do that for the hoi polloi who think everything digital should be free forever.

Re: Nvidia’s $589B DeepSeek rout

#549

Post by puppymaster »

elihu wrote: post #450
ozten wrote: post #431

NVIDIA sells shovels to the gold rush. One miner (Liang Wenfeng), who has previously purchased at least 10,000 A100 shovels... has a "side project" where they figured out how to dig really well with a shovel and shared their secrets. The gold rush, wether real or a bubble is still there! NVIDA will still sell every shovel they can manufacture, as soon as it is available in inventory. Fortune 100 companies will still…

Jevon's paradox would imply that there's good reason to think that demand for shovels will increase. AI doesn't seem to be one of those things where society as a whole will say, "we have enough of that; we don't need any more". (Many individual people are already saying that, but they aren't the people buying the GPUs for this in the first place. Steam engines weren't universally popular either when they were introdu…

I also dont get how this is bearish for NVDA. Before this, small to mid companies would give up on finetuning their own model because openai is just so much better and cheaper. Now deepseek SOTA model gives them much better quality baseline model to train on. Wouldn't more people want to RAG on top of deepseek? or some startups accountant would run the numbers and figures we can just inference the shit out of deepseek locally and in the long run we still come out ahead of using oenai api.

Either way that means a lot more NVDA hardware being sold. You still need CUDAs as rocm is still not there yet. In fact NVDA needs to churn out more CUDAs than ever.

Re: Nvidia’s $589B DeepSeek rout

#550

Post by chatmasta »

kd913 wrote: post #394

The biggest discussion I have been on having this is the implications on Deepseek for say the RoI H100. Will a sudden spike in available GPUs and reduction in demand (from efficient GPU usage) dramatically shock the cost per hour to rent a GPU. This I think is the critical value for measuring the investment value for Blackwell now. The price for a H100 per hour has gone from the peak of $8.42 to about $1.80. A H100 c…

Why is there this implicit assumption that more efficient training/inference will reduce GPU demand? It seems more likely - based on historical precedent in the computing industry - that demand will expand to fill the available hardware.

We can do more inference and more training on fewer GPUs. That doesn’t mean we need to stop buying GPUs. Unless people think we’re already doing the most training/inference we’ll ever need to do…

“640KB ought to be enough for anybody.”

Post reply on HN