GPT-4 Makes Nvidia the BEST AI Company on Earth (NVDA Stock)
Вставка
- Опубліковано 8 лют 2025
- 🧠 Register for NVIDIA's GTC and you could win a FREE NVIDIA RTX 4080: nvda.ws/3iVdj7A and enter the giveaway at tickersymbolyo...
💰 Want my research and stock picks? Let me know: tickersymbolyo...
Want to get better at investing? Start learning about AI. Watching Nvidia's GTC 2023 is a great start because #nvidia ( #nvda stock ) is going to cover the biggest AI topics that are making the biggest impacts on every industry right now.
#openai just released #gpt4 and it's insanely powerful. We're going to see it in every Microsoft ( #msft stock ) product very soon. But did you know that @OpenAI's GPT-4 was trained on Nvidia's hardware? In fact, Nvidia has over a 90% market share for enterprise GPUs. So, everything from @MicrosoftAzure to @amazonwebservices will have to double down on Nvidia's GPUs over the next few years as AI businesses explode. That's why Nvidia stock ( NVDA stock ) could be one of the best stocks to buy now and hold forever.
💬 Join The Conversation & Stay Up to Date 💬
Follow me on Twitter: / tickersymbolyou
Join me on Discord: / discord
Support on Patreon: / tickersymbolyou
📝 Resources & References 📝
@NVIDIA GTC Sept 2022 Keynote with NVIDIA CEO Jensen Huang: • GTC Sept 2022 Keynote ...
@CNBC Why OpenAI’s ChatGPT Is Such A Big Deal (CNBC): • Why OpenAI’s ChatGPT I...
@ARKInvest2015 Big Ideas 2023: ark-invest.com...
@OpenAI DALL·E 2 Explained: • DALL·E 2 Explained
@OpenAI GPT-4 Developer Livestream: • GPT-4 Developer Livest...
Using OpenAI's ChatGPT to generate business ideas: / 1599809546519601159
📜 Disclosures 📜
All content on this channel is for informational purposes only and should not be construed as professional financial advice or recommendation to buy or sell any securities. Trading stocks, ETFs, other securities, and/or cryptocurrencies poses a considerable risk of loss. Neither host or guests can be held responsible for any direct or incidental loss incurred by applying any of the information offered. The content of this video is solely the opinions of the speaker who is not a licensed financial advisor or registered investment advisor. Should you need such advice, consult a licensed financial or tax advisor. When you make purchases through links in this video description, the author may earn a commission.
🙏 Thanks for watching!
🧠 Register for NVIDIA's GTC and you could win a FREE NVIDIA RTX 4080: nvda.ws/3iVdj7A and enter the giveaway at tickersymbolyou.com/gtc/
⚠ CORRECTION: OpenAI's DALL-E is based on a Diffusion Model, not a GAN. Sorry about that!
Yes sir make more videos on Nvidia about their earning profits etc
PE of 148. What a steal!
Great video Alex! One correction: the rumor of GPT-4 having 100T parameters was referred as “ridiculous” by Sam Altman during an interview with StrictlyVC on January 18th. While GPT-4 is likely to be bigger than GPT-3.5, 60X bigger would be prohibitive in terms of training cost and time. The technical report OpenAI released was purposely hiding all details as they say it’s becoming too dangerous to share - a dangerous trend in itself if you ask me as it will push all companies not to share new model architectures - so nobody knows the actual number of parameters. You are really good at not spreading fuds and sticking to real facts, so you might want to look into it and correct it at some point.
Ok but this video never says GPT-4 has 100T parameters, so… not sure you’re making this correction.
@@TickerSymbolYOU I don’t think they reveled the actual parameters of GPT-4 so i am not sure how u got 10 trillion parameters.
@@TickerSymbolYOU my apologies, you did say 10T. A disclaimer “this is a guess” as no official info from OpenAI has been released to date would be great. Thanks for replying both here and on Twitter, you are super responsive!
i agree with a lot of it, although on some level, these superlatives “most important company in the world” cannot be taken seriously when literally months ago, you made a video about how NVDIA is in trouble.
I didn't realize things can't change over months.
We are all servants to the algorithm.
@@TickerSymbolYOU lol, that’s not a sufficient explanation
@@TickerSymbolYOU you can change your mind and I’m not a fan of arguing over semantics. What changed wasnt things as much as your mind, and i just think these statements are more palatable when you at least acknowledge that you had a different opinion and that you are changing your mind, as opposed to just creating quick videos to get views.
@@investormd4436 It's hardly a quick video. He clearly puts a huge amount of work and effort into each video.
Mate, this aged so well!!
I sold my stock at 300 and haven't bought back in. Too expensive for me. Most stocks are and I am a cheap dude. My strategy is simple. When my tech stocks go to lala land - I sell them and buy what it is on sale. I was bagholding META at 120 cost basis and that was from averaging down from 240. I sold half recently.
Right now it is bank stocks on sale and I hate owning bank stocks. I bought some BAC. ARCC and MAIN.
The final sequence with AI building for optimization was mind blowing stuff. Props !
TSM can also become an important company because it manufactures Nvidia chips. Also, ASML is an important company as it manufactures sophisticated machines used by TSM. I have positions on all (Nvidia, TSM, and ASML). Is anyone aware of more companies that can become extremely important?
I would like to know also
@@ifoundthistoday I now know one of the companies, it is Synopsys.
@@niteshroyal30 oh thx
Pretty much once a stock it's on this channel it's sure to go down.
NVDA is highly overvalued. 147 PE ratio
TSM has a way better PE and it makes NVIDIA’s chips
I would change one word . . . Take out “highly” and insert “extremely”.
You don't value growth stocks with the PE ratio. You use forward PE or the PEG ratio.
@@baneinvestment917 Agree. Seems like a bargain considering they are the only game in town for the most advanced chips. Maybe the geopolitical situation is holding TSM’s value down.
Nvidia's earnings are about to explode.
AWS and Google make their own chips. They did so specifically because they can poach chip designers from Nvidia and make chips at lower costs better suited for them and their compute architecture for their cloud. I love Nvidia, but Nvidia is overvalued. As soon as Nvidia becomes too profitable, they will be dropped because the giants will want to save on margins. I have a hard time imagining Microsoft having any loyalty to Nvidia in the long term.
Yes they do make in-house use only chips, but they are still buying record amount of chips from Nvidia. They make ASIC chips only that can't be updated, and don't have the high performance as Nvidia. A & G are also laying off thousands of people. Having the top performing chips is much more desirable than lower margins. It just doesn't make much sense to spend billions every year making your own chips while it's much cheaper to buy off-the-shelf solutions from Nvidia. You don't hear very much anymore about Google's TPU that they used to boast about being close to Nvidia performance. All the giants will continue to work with Nvidia so they won't be left way behind in cutting-edge performance. Nvidia does not lay any of their people off...so why would anybody leave? Nvidia and MSFT are currently have a monster partnership on multiple projects. The keynote will be another reminder on why the giants can't even begin to effectively compete with Nvidia. Nobody has a complete solution like Nvidia!
I don't think neither AWS or Google have their own fabs? Time frame through design to bug free silicon is not insignificant. Shorter time frame and reduced risk is a big deal. Chips are not their core competency, but what the chips deliver for their customers. It appears Nvidia is in a pretty good spot at the moment.
@@scrout TSM makes the chips for everyone at this level.
@@scrout Nevertheless, it is way overvalued.
Eventually, a company needs to pay its workers more, consumers can no longer afford its prices, new competitors come around, old competitors come out with better products, things happen. NVIDIA is actually in a really bad spot where everything has been going right for so long that it's almost guaranteed something will go wrong.
@@pinbot454 Google is a poorly run company that has been resting on its massive profits and us government funding for far too long
Imagine an AI creating an entire story mode game by itself and making the characters adapt to the players experience.
Not far away in my opinion
Create ai image. Make a 3d model with ai. Add it in the game you programmmed or modded with... AI. Hm. I see a pattern.
Already got chat bots creating ai npc dialogue
Careful: Advances in AI could slash the training or inference compute.
The software platforms are still interesting though.
Google doesn't SEO. Google does page ranking. SEO is the craft of maximizing the page ranking (and improving the search result) of some site by adapting it to what the page ranking algorithm might do.
I totally agree . I’ve been wanting to invest for months yet with a P/E ratio of 133 it’s overvalued by the market . Wait for a crash they buy.
still waiting?
"The results were consistent across tests. All four tests, the Pew Research Political Typology Quiz, the Political Compass Test, the World's Smallest Political Quiz and the Political Spectrum Quiz classified ChatGPT's answers to their questions as left-leaning." ~ David Rozado, 12/13/2022
Just to clarify:
Rule-based systems are definitely considered Artificial intelligence (like the one you mentioned in the video game or Google SEO examples).
AI is a generalized concept, there are several layers depending on the degree of complexity.
If-then statements aren't considered AI.
Jim Cramer said a couple of years ago that nvda will be the first ten trillion dllr company
He also called NVDA a dog and said he’s short late last year.
Thanks for that stock statement I bought three shares last week😊
Question? I thought you had an interview with PLTR? Is the still on? I know they have AI in their system.
If you use AI for content creation, I would argue that copyright/trademark laws will be challenged and ownership might revert to AI companies. Thoughts?
copyright protects the original creator. AI is now the creator - the owner of the AI model can claim copyright.
@@phvaessen One could argue that it is also a collaborative work between the AI and the prompter.
Quite a while ago you spoke about making a video about your experience with Palantir. Did you ever do that video? Thanks for your content.
It keeps getting pushed back but I promise it's in the works.
Semi-conductors is such a difficult industry: competition is FIERCE; you are only as good as your most cutting edge offering; cyclical earnings in the extreme; and all the money goes into R&D and capex not the share holders.
Google doesn't use AI in Search and uses SEO instead? That's completely wrong. You literally showed an visual example of transformers on Google Search at 4:55
how do Nvdia's supercomputer and others compare to DOJO?
What is the sustainable competative moat? Aren’t chip makers prone to disruption easily every few yrs?
Beware pump and dump artists NVDA CEO is the master of it as is this guy
Wow ! 🤯🤯🤯🤯🤯. Generative Pre-trained Transformer. MIT - Nice! Nvidia stock went down, then back up. 😂 I might buy a few shares then wait for it to go down then buy some more, if it doesn’t just go up with the way AI is rapidly evolving. Most of my investment is in Tesla (AI + FSD + Optimus robots + everything else they’re working on). A great deal of good information from your video - robots also. I have go back and watch and take notes. Thank you !!! MIT !
the stock price of nvidia is going crazy already
NVDA is the most important company in the world. I still play games. FF7 Remake is master piece and god of war.
Do not forget Nvidia relies on TSMC to make AI chip, and TSM P/E ratio is around 14 which is undervalued
Thanks Alex great stuff. Just wondering if these AI companies are using AI to build their transformer models ?
I'm sure AI is being used in more and more parts of the process. Coding co-pilots are popping up everywhere.
At least there's alternatives to NVIDIA. AMD's high end GPU's are also used in supercomputers, Cerebras could even replace NVIDIA for many AI applications (the Cerebras-GPT was trained super fast), plus Google & Tesla are developing their own AI chips. Of course they're all made by TSMC with ASML equipment.
That said, Open AI uses Azure, and Azure uses NVIDIA, so NVIDIA will be huge for the next couple years.
More important than TSLA?
This guys has some calls in NVDA
Great video as always. I'm curious if you'd recommend getting a grad degree in AI at this point?
I honestly think that’s one of the few grad degrees still worth it. AI is such a rich field. Not sure there are broad “AI degrees” but I’d recommend most related computer science or electrical engineering degrees for sure
@@TickerSymbolYOU Ok cool. Yeah Cornell, Carnegie, Columbia and U of Texas offer broad AI masters.
A lot of colleges are going out of business, they’re getting less and less students every year. Every time a college goes out of business, that degree holds less value because the college system as a whole gets a weaker reputation. Best to learn this on your own, there’s plenty of resources, that’s where society is heading.
So no updates on foundry?
Great take on this subject!!!!!
Awesome story. And I have an upcoming interview with a IBM Head, will definately want to know where they plan to get in with AI.
That's awesome! Best of luck!
Neeeeh. AMD will rock it as well, Lisa is there ❤
AMD will show its charm
OpenAi's DaALL-E is not GAN. It's a diffusion model
There are so many incorrect things in this video
Besides Dall-E being a diffusion model, what's one other incorrect thing?
Thank you for all this information.
Brainchip uses neurotrophic processing in the chips which is superior in my view.
I find Google to be inept in providing relevant information. Whatever algorithm they use they don’t provide the most substantive information. Seems like popular or paid info is prioritized!
ItS because Google censors results that dont suit the companys agenda. Conservative s are 100% censored i dont care what the lefty Lol "fact" checkers say.....
True, they have adopted a tictoc model now instead of a search model.
I'm sorry but nobody except OpenAI knows parameter count, also trillions of parameters seems a little off given the chinchilla scaling laws
It's not off. GPT-4 can understand images; GPT-3.5 can't. That requires a much bigger model.
Not sure if the latest RTX 4090 will be able to power gpt4, definitely not as powerful as the H100, but shouldn't be slower by much?
NVIDIA has different cards for their data centers. The RTX cards are consumer-focused. Their data center cards are a different animal altogether.
Does Nvidia use any of this AI within the PC graphic cards, and is any games using the built in AI during game play?
Yes and it is very powerful, DLSS.
The idea is you render the graphics properly at a lower resolution, lets say 1080p, then upscale the image to 2160p (4K) via AI.
This means you only have to render 1/4 of the pixels, and can have a MUCH higher frame rate, the AI does a brilliant job at upscaling and sharpening the image to a greater resolution in real time as you game.
Lets not forget AMD. they have their Instinct lineup too and ,also, nvidia chose amd epyc cpus for the dgx cards. I d also like to mention dojo from tesla, but I think they re yet to show their progress on this product/project
yes, AMD recent bull run is wonderfull, long both. But in the recent years AMD gave more dips to buy than NVDA
AMD is an excellent company, but there offer is not comparable to NVIDIA's. NVIDIA is a platform, not "just" GPU's. Because of that, NVIDIA is a long term winner. I own both, by the way 🙂
@@phvaessen Yes, but AMD´s market isnt only gpus... AMD is developong ROCm software
@@marcos14223what os ROCm?
@@phvaessen AMD is not on the same level on the software side, for sure, but they have products in most if not all hardware branches. Recently they kind of beat Intel in the gaming and I d argue in the server sector as well (see the epyc lineup and ryzen 7000 series, namely 7800x3d which most likely now is the fastest gaming cpu, beating intel after many years of dominance)
There is also Nvidia's supplier, Taiwan's tsmc..
Designed by Nvidia, manufactured by tsmc..
Yep! I cover TSMC a lot!
I thought Frontier by the US Energy Dept is the fastest super computer currently. I’ll have to check out this one you mentioned in Japan.
1:45 - name of the game please
Mass Effect Andromeda
Very interesting. Subscribed.
And then we forget how to do all this. Dark ages all over again. The evolution is mind blowing, WALL-E?
Thank you for your honesty thank you for the videos
I thought the trend was for big tech, at least, was to develop their own chips to customize the chips to optimize to specific applications. Is that not correct?
NVIDIA's unique position is NOT its GPU hardware, but the software that uses it. NVIDIA has developed an entire ecosystem around its platform. Several hundred thousand AI researchers are developing industry-specific AI models using NVIA's open software platform. Some Big tech will make specific cards for their needs, but nowhere close to what NVIDIA is offering. NVIDIA's platform fits many applications and will adapt to new needs faster and at lower cost, this is essential in a fast evolving environment.
Would you say this software ecosystem is similar in loyalty adoption like Apple IOS?
@@SumitGupta-nh1ki I am not a fan of APPLE so I have no valid opnion.
I like Bill Gates definition of "platform": A platform is when the economic value of everybody that uses it, exceeds the value of the company that creates it. Then it's a platform.”
The term "platform" is much, much more than just a piece of software. It is the common interest of many players around the globe that support the common success that matters. For example: NVIDIA depends on TSM for their chips, but NVIDIA also delivers the super computer (hardware+software+network+optimisation services) that helps TSM works better, cheaper and faster. Same with Microsoft: Nvidia built (most of) it's supercomputer, and now rents a part of Microsoft's free ressources to his own customers that want to play with AI. It's a win-win, and also there is no more a customer-supplier relationship, it's intermixed.
*bro.. i thought you were AI person talking, i know now you should be real human.*
How do you know gpt4 has 10 trillion parameters?
I did the math (and Nvidia has a plot showing the growth of these models over time).
sell sell sell: goes up
buy buy buy: goes down
Story of my life.
I feel like this guy makes videos on stocks that just ran
I’ve been making videos on Nvidia for over a year.
but amd is faster and lower electric?
There is a fundamental problem on UA-cam. That’s trust. And that’s because it’s a no win situation.
UA-camrs have to keep putting out videos otherwise their channel dies. It is as simple as that. Engagement goes down which means views go down which means income go down. It’s a hard gig as it is.
So when you see a video like this, it’s hard to know if it’s genuinely important info or just another piece in the unending chain. It’s especially true when dealing with opinion piece like this.
Its not a prob with just this channel, every channel except very very few.
So yeah, don’t ban me if I don’t fully accept everything said in the video.
"Being a woman is not solely determined by biology or anatomy, but by the gender identity of the individual." ~ChatGPT
Lol let’s create something more powerful than us and fill it with lies.. should work out
Beam NG drive 'UTHA' map clips used for AI route guiding
No NVIDIA without TSMC
How come apples phone continues to get worse at auto spell, and auto prediction. It’s complete junk.
The valuation of NVIDIA rn 😅🤯
2:20 are you saying it was a hoax ???
Why did you insert such a despicable statement?
@@Marmots4reFun In some countries he could be locked up for this
NVIDIA is way too expensive to buy from a stock perspective. Good company but crazy high valuation.
First bought Nvidia in 2018 and they were saying the same exact same thing. A 300% gain later I'm laughing all the way to the bank
@@mcc5746 Suuuuure you did.
And even if assuming what you're saying is true...the same can be said for Microsoft, AMD, and Google. It doesn't mean they aren't overvalued. If anything, their extreme growth in such a short period of time should serve as a warning that they are grossly overvalued.
AI, AI, AI... Most things you mentioned are not AI but simple conditional logic.
I literally say that in this video, yes.
The valuation is going to come down. AI is going to be a bit of a waiting game and the valuation is way too high compared to revenue which is expected to be hard to come by in the sector. Planning to pick up some shares on a pullback.
Yeah, but down 50-75% probably not. Maybe 20-30%? And then we can just buy more.
@@mjregan88 Yea, technical pullback to 190 easy. I just feel like so much has to go right for NVDA at these levels to make new highs and we're almost there. I'd personally rather go AMD than NVDA just based on valuation although I love NVDA over AMD much more in terms of possibilities of TAM expansion in an AI dominated tech market in the future. In all honesty it depends on when AI is really going to have a massive impact on our daily lives. Sure Chat-GPT is rad, some self-driving is cool, but I'm talking autonomy. It's a matter of when these opportunities actualize and I see them at least 3-5 years out which is why I'm more focused on valuation in the semi space. I sold a sweet trade on AMD, waiting for pullback, and then I will buy back in.
Won't Amazon AWS chips be a big contender in the artificial intelligence space for their cost and energy efficient qualities?
Definitely.
AI AI AI AI AI AI AI AI - NVDA ceo during their report
to be fair... that's all Nvidia does.
Thank you.
Of course Everyone can see that now, it’s follower mindset.
Ok but I was saying this years ago. I have the videos to prove it 😏
@@TickerSymbolYOU you mentioned a lot of stock, one of them could be right for sure.
So, TL;DW: all the hype about GPT boosts AI companies, thus NVIDIA? Did I miss a more direct link or did I get that right?
Direct link: these algorithms are all trained on Nvidia hardware
@@TickerSymbolYOU Ah! Might change though, as both Intel and Tesla each have developed dedicated hardware for ANN training (which have better performance and MUCH better power efficiency).
I think Nvidia's position as a leader in AI is a lot more fragile than people believe. There are many startups that could beat Nvidia for training such as Cerebras and Tenstorrent. People currently pays Nvidia for chips that can do a lot of calculations. "AIU" chips that are only optimized for matrix multiplication and do not need to support other operation types could greatly 🎉 them.
TSM makes the chips for NVIDIA, won’t they be a bigger winner
In my opinion, you're 1000% correct - This is what I don't get about these videos. I get his point, and he is correct, but to call it the most important/best AI company is a stretch. TSMC creates all of their chips, has pricing power and is overcoming many of the intrinsic risks e.g(China) that it currently faces. I'll stick to "buying the shovels" and keep adding to my TSMC stack for long term
Note: I am a biased TSMC holder
INTEL will make them all
TSM makes chips. NVIDIA offers a platform. Big difference. Both are excellent long term investments
@@phvaessen don’t get me wrong I love both but NVDA is very overvalued at this price, but has a bunch of long term appeal also
No because Nvidia also builds the software platforms people use
Absolutely!
If AI can design chips and TSMC builds them, why do we need Nvidia?
Does Nvidia know more about chip designs than other companies? Enough to make its AI chip designer smart than what other companies can build? Does it have a better relationship with TSMC?
Is Tesla further along in designing chips specifically for machine learning?
"Nvidia" will simply become the name of a very powerful super intelligent AI and will cease to be just a technology company.
Do you buy NVDA, or own the stock you cover?
Feel the same about this
SKYNET IS NOW ONLINE !
Is AMD going to come from behind again and trump Nvidia’s AI chips like they usually do or is this lead from NVDA more unassailable this time?
We’re talking server chips. So, different story this time
More people, less need from them. Where will this stop? It's cool stuff but... like your content
AI need super computer made with GPU
@TickerSymbolYou I'm gonna be embarrassingly honest; I spent a lot of time trying to figure out it you were an AI character. I was not settled on an answer, but decided to just trust the information anyway. This is truly ironic...
TRY DOJO AND D1 CHIP
Jensen is the Da Vinci of our time
Eh no...that would be Shigeru Miyamoto
buy NVIDIA stock now
Oh, all our jobs and industries that we’ve been working on our entire lives are going to wiped out. How exciting.
Good Video
title is a clicbait
NVidia: CEO thinks he is Tom Cruise so therefore so many photos of him in his leather jacket, but he is just fat version
Tensor core
Thank you for your video . The market is going down and a lot of people lost huge sums already . I Will advice traders especially newbies to have orientation of trading before they get involved in it because the Crypto market has been unstable , forget predictions and start making good profit now because future valuations are all guesses.
Google’s algorithms are poor.
'Akida'.
i have this feeling that amazon will sink and become just another fly by night memory, i cant even talk to alexa anymore, because i want chatgpt in there instead. old bezos dropped the ball and got left in the dirt. sell now before its zero
If trillions were true, then I suppose a cost-down so that decent AI lives in Notebooks and IOT.
As a comp architecture layman, I worry about "Graphics Processing Unit", an architecture I had on my 286 back then, as their architecture seems suboptimal for AI modelling to me, meaning something that has been repurposed. I get the weight matrix multiplication thing.
Excellent!