저자: MeetKevin · 외국인
요약 보기
ARM IPO는 AI 붐에 편승한 SoftBank·임원 엑시트 성격이 강하므로 최소 6개월은 투자하지 말라고 주장. 매출 약 25%가 중국(ARM China)에 의존하며 중국 실적 투명성이 낮고, 로열티 모델이라 오픈소스 RISC-V 확산 위험이 있음. 엔비디아는 AI 수혜가 가장 크지만 AI가 일시적 유행일 수 있어 실적 사이클을 지켜봐야 한다고 봄. SoftBank는 주식 담보대출(85억달러)로 락업 위험이 있음.
다듬은 글
hey in this video I'm going to tell you everything that you need to know on the arm IPO maybe you don't even know what the heck arm is and don't worry I'll break that down for you so we'll start with what the heck is this company do they actually do AI which is the big buzzword these days which by the way is perfect timing especially since the CEO is going to get a nice 20 million dollar check just for getting this IPO done that's on top of stock options that are going to vest okay these folks are gonna make a killing on this IPO this IPO could be rivaling some of the money rays that we've seen from companies like Facebook rivian Alibaba and let's just say two of those three lost a lot of money so uh walk into this one with a little bit of caution but I'll also give you my opinion on how I think this play compares to something like Nvidia all right let's just get started with it this is the filing widely anticipated filing the arm company okay so what the heck is this company well here's their prospectus summary but let me just simplify this for you okay there are these things in computers we need called chips no not like Doritos but things that process computer information there are companies that are called designers of chips and designers are going to be companies like Nvidia really popular designer of these chips and then of course somebody has to make those chips and that's going to be like your Taiwan semiconductors which will actually make the chips uh and then of course you have companies that Supply materials to help make those chips like asml that is a company that supplies some of the advanced lithography machines ultraviolet lithography whatever basically fancy machines to help make smaller and smaller chips okay got it so that's asml that's a Dutch company here's Taiwan semi it's a Taiwanese company they do somewhere around 90 percent of advanced chips in the world Nvidia is sort of a chip designer and Nvidia in addition to designing chips also provides code for some of their chips like specifically their GPU chips which they use a language called cuda okay now how does that relate to arm so arm is a language provider language okay now this is where things could get a little confusing but we want to keep this simple okay there are basically three types of languages x86 and arm these guys go head to head all the time okay arm is based on a risk framework which this was written as risk that's because I think the stock might be risky I'll explain why but it's actually risc risk five anyway so you've got x86 this is your 1970s 1980s language that Intel founded and AMD uses today then you've got arm which got really popular because it optimizes what you actually need a chip for and that is really good for Energy Efficiency and so arm became very popular for mobile like smartphones right now arm helps provide code and gets about 50 percent of their revenues from smartphones and consumer electronics keep that stat in mind that'll be important for the whole AI thing okay so anyway you got x86 and arm kind of like head to head over here with these different languages for Designing and using chips now I said there were three but just focus on those two for a moment so you've got arm and then you've got x86 and these guys are sort of in a battle against each other to get designers to use this code so yeah that actually means Nvidia in for example the Nintendo 6 Tetra uh a CPU central processing unit has used arm language to make their chips who else uses arms language to manufacture their chips Samsung Amazon and their AWS their graviton AWS chip their Cloud compute chip is an Arm based language chip Nvidia has used them for many other purposes but also Apple the M1 and M2 chips we keep hearing about those are Arm based chips the old a chips that we saw for like the iPhone 12.
those are arm-based the s chips that were in the eye watches the Apple watches those were Arm based okay so hopefully this adds a little bit of clarity so what you have is you have this language and everybody in Mobility is really using arm these days because it's more energy efficient and Intel's kind of like crap man we're stuck with the x86 that battle is so important to understand because today Intel is seeing their revenues do this so why would somebody invest in Intel well listen to this Intel's gotten so smart they're now building new fabrication that is facilities where you make chips Fabs column they're building new fabrication facilities with stimulus money from Joe Biden and in those facilities guess what kind of chips they're going to manufacture remember they found an x86 are they going to manufacture x86 chips Only Hell No even they see the writing on the wall and instead Intel will manufacture chips that are on based on x86 that architecture and arm so what you could see is this competition for which language to use is a big deal but arm is the company that came up with and uses well it was a partnership back in the day in like the 80s Apple was actually involved in this but what's left over is basically the company arm now okay so you got the arm language and x86 that's just the lines of code that the chip designers are like okay cool let's uh let's put silicone together like this and what if we stack away for like this oh it does better oh let's do more of that uh or oh our uh our uh enemy over here AMD just came out with a and new chip well let's put it under an electron uh scanning microscope or whatever these folks use and go oh that's how they manipulated that graphine oh man I'm gonna bring this home to my wife tonight okay all right you get the idea okay this is all like a massively competing industry for chips okay everybody wants the best Dorito on the Block some people write the language some people design the chips some people make the chips some people Supply the machines that make the chips arm is simply and at this point only that it is just the language okay now that's really important because there is also now potentially a third language and that is also a it's well it's called risk five but arm the r in arm is because it's based on risk but risk five because it's an open source language risk five is sort of this new kid on the Block and people are like oh maybe we should use this but then companies are like no we're going to keep using arm because that's what we know how to use okay worth knowing though risk five is actually a risk to arm it's probably small it's probably going to happen over time where you see a transition between different languages but I think we can say with certainty a lot of the x86 business is going to arm or risk five most of it probably going to arm right now worth keeping that in mind so uh here are just some other examples I went through some of these Fire TV these are I just wrote some notes as I was doing my research on some of the things that are arm-based chips okay great in fact they go as far as saying 70 of the population probably uses arm-based products okay that does also make me wonder how are you still gonna grow if you're already doing that much that is also going to become very important here in just a moment here's some things that I wrote down though because artificial intelligence is going to be a big deal so what I wrote down is there's something to know there's a difference between a GPU and a CPU arm does a lot of CPU work okay this is where arm does a lot of work it's not to say they you can't use arm for gpus and I'm over simplifying this video otherwise this will be like a five hour long video most of arms work is going to be in the CPU those central processing units especially for mobile devices so you're going to hear a lot about energy saving keep that in mind because that's going to be a big deal too when we get to some of the pros and cons of this company Okay so CPUs can usually be used for AI inference models and gpus are usually used for training okay now I may have lost it there let me just put it this way okay if I said hey you're an AI and you need to tell me if the face that I'm drawing is a happy face or a frowny face and you started seeing me draw this what can you infer that the face is about to be well you're probably like well I mean it's starting to curve kind of down so it's probably going to be a frowny face right you already know that because you've already seen a million different frowny faces as you've grown up throughout your childhood hopefully you've also seen an equal amount of happy faces okay but the point is you know what a frowny face is a frowny face understanding what this is the first time is really what you're using gpus for today it's a training whereas once you kind of start drawing and providing oh what's it going to be what what oh What's it gonna be oh it's going to be frowny that's where CPUs come in more now don't get me wrong there are highly intensive visual applications that will require gpus to tell you that this is a frowning face think like Tesla self-driving will probably be more involved in the GPU world this means there is somewhere a space for CPUs in artificial intelligence but it's not entirely clear how different that is from just operating regular programs on computers today in fact what I can tell you is entirely clear is that if you wanted to build a new language model you would probably do it by renting server capacity and this is what the server would look like the top part of the server would generally be a CPU although even Nvidia is trying to get into that game right now everybody's trying to take everybody's bread okay this shouldn't be a surprise okay keep in mind I'm a licensed financial advisor and even though I can't give you personal financial advice it's my job to tell you you need to make sure how to get more bread in your life that's everyone's goal that's why we're here everybody wants more bread okay that's fair so the CPU might be at the top and let's say we're going to throw in an AMD CPU at the top of the server rack guess what the next 10 items in this server rack are well more often than not now they are Nvidia h100s though there are other weaker models that do good enough for most AI things that people probably need you probably don't need the latest and greatest but just to show you the latest and greatest Nvidia is probably about a thirty thousand dollar chip for each one of these lines right here in a server stack and that AMD CPU is probably about 15 grand so if you add this up you're looking at about 15 grand to 300 Grand in income so in case you're wondering why Nvidia has gone to the moon it's probably because of this now they've got earnings coming up they're always going to have earnings coming up this video is not to speculate on Nvidia earnings I've got a position in Nvidia I want to be clear about that obviously I don't have a position in arm because it's not public yet my position isn't massive in Nvidia I think it's gotten a little ahead of itself though I'll tell you if they can prove that they're not a flash in the pan Nvidia is going to be a great investment and and I'll explain more about what I mean with the Flash and the pan problem in just a moment but again the whole point of this is to say if you're doing language work for this part on AI servers you're probably missing where the real money is you see what I mean and where's who's Pro what's the programming language here it's Cuda okay and who owns Cuda it's Nvidia again I I prom I'm not trying to chill in video I would love to have other stocks I'm just looking at this like oh man people are gonna hype this up like it's a Nvidia uh you know play uh on uh like an alternative to Nvidia and personally I think that could end up being a mistake now I'm open to your commentary in the comments but we got a lot more to cover okay so let's look at some more some more notes here right ready for this so arm right now says they want to get into Data Centers so the CPU strategy of getting into Data Centers you have to kind of think okay who's the data center world for AI right now well maybe you could do big data like palantir or snowflake or of course your OG names the Google the Microsoft the Amazon right that's great but how different is this new amount of money that uh arm wants to raise now from what they've been doing for Amazon's Cloud compute so you remember what an IPO is an IPO is not an opportunity for you to get rich as an investor an IPO is an opportunity for the Insiders to finally dump their shares on you this is why usually IPOs are not a great thing to invest in I generally stay away from IPOs and even though I've been saying that for years I still have taken a bite on some IPOs only to get burned and learn my lesson so I'm I'm a little jaded when it comes to IPS just gonna throw that out there now it's also worth noting that hey ai's been great but how much has it really boosted our world's productivity these days my take is potentially somewhere around 10 percent maybe 15 if you use some of the strategies that we talk about in my courses on building your wealth where we integrate here's how to actually use AI to be productive but okay is it really like revolutionary is everybody really waking up going oh my gosh is AI no of course not because the reality is AI has been big data and neural Nets they've existed for 10 plus years in fact we've been working on our own artificial intelligence for a real estate startup for four or five years we just sort of keep evolving it and this whole AI craze in March changed nothing so it's kind of like yeah okay you found a sexy way to label AI it's one of the concerns I have for NVIDIA okay it's actually why I'm not kind of going like oh let's go all in on a video right like there's a little bit of gun shyness that I have and I actually think there'll be better opportunities to buy Nvidia in the future and I'll explain why okay so now we understand a little bit about uh this this arm versus CPU GPU thing uh now what we should do is understand a little bit about where this money is going to go because they want to raise you know about 10 billion dollars would be nice and so why do they want to raise money well they tell you here that developing Leading Edge chips is getting exponentially more expensive Ah that's actually going to be a red flag keep in mind this company tells us that 80 of their staff is sitting in research and development to design a seven nanometer chip the company spent about a quarter of a billion dollars to design a two nanometer chip they spent about three times that so this is where they talk about this exponential growth in costs ah so why do you want an IPO well because you're probably at the peak AI hype cycle and it's a perfect time to dump on the normies okay I'm sorry I don't want to sound like bearish maybe this thing will go to the Moon I'm just saying if if I were uh you know one of the executives over at arm and I was calling up my buddies over at SoftBank which is not a bank it's a Japanese investment firm be like hey everybody's you know losing sleep over how great AI is and how revolutionary it is you know it's it's amazing amazing uh how about we just IPO now it's the perfect time to do it remember this is a company worth noting this is a company that was acquired for between 31 to 32 billion dollars in 2016 by SoftBank SoftBank not a bank remember that agreed to sell this company in 2020 to Nvidia for guess how much they agreed to sell it for 40 billion dollars that was in 2020 2021-ish but the deal fell apart because the FTC sued and Regulators got involved it would have just been too big uh like Nvidia would have just gotten too big now they're kind of like okay well forget about that 40 billion dollar valuation when everybody was printing money now that everybody's so hyped up on AI why don't we try to sell it at a 60 to 70 billion dollar valuation oh yeah baby give me those dandies this way SoftBank can exit with a sweet sexy double now you can't blame them I mean nobody's gonna blame anybody for for making money but what I did think was interesting going through the filing and everything I say in this video you should assume is an opinion because I can't remember every I spent like two or three hours on this I could be making a mistake on some of this so just use some of this as an inspiration for you to go through the information yourself in other words arm and soft don't sue me bro anyway SoftBank has basically carved out a little exception for themselves here's what's going on usually when a company IPOs there's something known as a lockup for example everyone on the inside like all these r d people who have been getting stock based comp for years or decades are finally like yeah baby my ship is coming in we're gonna get rich off the IPO and we're gonna sell those shares as soon as we can maybe and those folks would be able to do so usually within about 180 days so in other words you IPO and you kind of have to wait six months for the market to settle down and then people on the inside can start selling except for SoftBank SoftBank at least based on what I saw only has about a 60 day lock up that is because they wrote in their lockup expires 60 days after a public offering interesting okay so a lot lower in other words SoftBank can get out before the Insiders okay interesting what else I noticed is that SoftBank then said and this is what I thought was really interesting maybe it's normal softbank's got a bunch of debt remember they're not a bank they're an investment firm and basically what they're going to do is they're going to pledge eight and a half billion dollars of their shares as collateral to basically borrow against all right that's a whole lot of words let me put it this way the IPO is expected to raise 10 billion dollars SoftBank owns a ton of shares tied up by this uh you know lock-up agreement of 60 days or whatever but they're basically going to take a loan out on day one and if the stock tanks the bank that gave the loan can liquidate the shares right away so in other words SoftBank gets their money right away on a loan if the shares tank then the bank can Margin Call SoftBank right away to cover the bills basically so if it tanked on IPO day the lockup agreement means nothing those shares start getting dumped to the tune of eight and a half billion dollars now I don't actually think the shares will tank on day one mostly because quite frankly it is now the job of the executives at this company to market the snot out of this company you're gonna see banners from this company about how great this IPO is how you should invest in this IPO how you should invest in the next generation of AI chips and arm 9 and all the latest and greatest crap from them but the reality is when you go through their documents there are some more problems uh okay let's get into some more of the problems here and then we're actually going to get into some of the revenue okay let's get to some of the juice you ready for some of the Juice by the way if you like this kind of video subscribe because there's more where this came from this is what I try to do all day long every day try to provide value to even people who invest with me uh or to my subscribers or the people who are so gracious to sign up for the courses linked down below so take a look at this all right these are some of my initial thoughts that I wrote down this has to be marketed uh in my opinion if Nvidia misses on earnings either Q2 Q3 Q4 it doesn't matter when any Nvidia Miss is probably bad for the entire industry and Nvidia beat is going to be good for arm not because arm is necessarily going to capitalize like Nvidia but it's just going to reignite the whole Ai fomo and AI fomo is going to be perfect I guarantee you all of the executives that arm are like come on baby come on and video it okay all right now it's worth noting that about 24.5 of revenues come from China from this company but what's a little scary about this here's the segment section by the way 40 from the US what's a little scary about the 24.5 that comes from China is that the part that comes from China is basically only determined based on what China tells them listen to this there's a whole section you can look this up it's on page 89 too bad it's not on 69.
anyway performance of armed China okay we depend on arm China for basically a quarter of our Revenue okay and this company gets licensing Revenue I didn't even mention that yet they don't make chips they just get royalties from the code they wrote we go we wrote the best code and everybody uses it well I don't think it's going to get replaced anytime soon maybe there could be a gradual transition to open source but honestly a lot of people used arm it's good code it's good technology I'm not gonna bag on that do I really think it's like the next AI Wonder no of course not uh but then again nothing really is me except for maybe Nvidia if they could pull off the amount of server sales that they're talking about okay so what is this right here so here they say they're dependent on China armed China right but listen to this I'm just going to read you this okay ready for this all right here it is under our partnership agreement armed China's payments do to us our determined based on the financial information that armed China provides to us accordingly we are dependent on armed China providing us with reliable And Timely financial information in other words if there are any issues in China and China's lying to us we would have no idea okay given that a quarter of your revenue is based on we have no idea it's a little bit of a red flag to me okay furthermore in light of these issues we expect to continue to see declining royalty revenues derived from the People's Republic of China and we could see declining and Licensing revenues in other words a of their business they think is shrinking so part of me is like trying to like jump out of the ship or something like that I don't know anywho let's keep going here so this arm thing is a risk all right they talked about 30 billion of a arm-based chips shipped last year that's not all their revenue right because they just get royalty on that right so even if you had a million AI chips that shipped that would be one one thousandth uh of of the 30 billion a million dollars of AI chips right that'd be one one thousandth of the companies 30 billion now if you had a billion dollars great that'd be one thirtieth my point is NVIDIA expects to ship 7 billion dollars of AI related data center chips per quarter that's about 28 billion per year that actually means there could be a real upside here if they could actually really ship uh AI chips but I'm not really seeing that yet so I want to be clear here because that might be a little confusing there is real potential upside if they can convert their pipeline to really amazing top producing AI chips but wait a minute wait a minute they're not making the chips that was the whole point of that little setup I just played I'm sorry I set you up I built you up just to tear you down they don't design the chips they do the code and somebody else does the chips they literally say in their investor relations document we rely on our customers too design manufacture and sell quote unquote chips so really even if you sell 28 billion dollars of AI chips in a year arm just gets royalty revenues on them so a percentage as a thank you for using the code but if that code is actually using Cuda and Nvidia or something else and is this code being used for gpus nominally it's mostly used for CPUs and that's really important to remember now do keep in mind they do talk about potential GPU deployments in fact they talk about arm gpus so they exist it's just not a primary product see it says right here our primary product offerings are leading CPU products complementary products are gpus all right good fine so uh and here's by the way where it says that 80 of their employees Global employees are focused on R and D uh worth noting and their revenue was relatively flat but that's okay because the whole chip industry has somewhat slowed down so I'm not like horribly worried about their revenue down one percent gross profit year over year eight point eight percent external year or um external Revenue year over year that's not a big deal the whole chip industry has slowed down uh not so worried about that uh at about 50 cents in earnings per share if this company IPOs at the valuation we think it will around maybe you know 60 to 70 billion bucks uh maybe 60 ish dollars per share that puts you around 120 times earnings because it's a software company this is an insane valuation for a company that actually expects to kagger compounded annual rate of growth for their compute business grow at 15 percent okay that's really really expensive I mean you're talking about over a three peg here for those of you keeping track Nvidia trading at a lot lower peg in fact let's go ahead and pull up the Nvidia Peg just so you can compare for a moment the Nvidia evaluation so remember that Peg is priced to earnings growth let me pull this up for you okay so this right here is nvidia's latest quarterly document uh you can see their revenue year over year is actually down 13 this whole industry has gone through a correction so it is not surprising to see some negatives here we talked about the h100 revenue if that h100 ordering is a flash in the pan Nvidia gets screwed however if Nvidia starts growing at 50 percent oh it's gonna be juicy why because right now the stock market Wall Street consensus is that Nvidia is going to make over eight dollars per share this year over eight dollars per share puts them at about a 57 price to earnings ratio which is about half of the arm price to earnings ratio with a 25 growth factor which is about 80 more than arms they're in about a 2.3 Peg so you have like a three peg for arm 2.3 Peg for NVIDIA but if Nvidia is not a flash in the pan with the h100s and they start growing at 30 40 50 this Peg could go all the way down to one which means Nvidia stock could double there's some real potential Nvidia or it's a flash in the pan we'll find out over the next probably year as the earnings come in so that's why I wrote at the bottom the danger is The Flash and the pan for NVIDIA Nvidia does have about three to four times the revenue that arm does three to four times the free cash flow that arm does uh but let's be fair to arm here arm does have good free cash flow I can't take that away from them and they're well capitalized they've got about 3.2 billion dollars in current assets versus 2.8 billion in total deaths this is great it's actually very good 675 million of free cash flow no borrowing no dividends they don't plan to do any dividends just keep in mind a lot of employees are probably going to end up dumping shares once they are able to do so okay what else do we have here oh boy I'm telling you I spent way too long on this okay I'm gonna need like an AI beer or something after this all right so we got a lot of information here oh okay I'm sorry their cloud kagger is expected to be 16.6 I think I said 15 whatever but the growth is really unclear for AI and they talk a lot about energy saving for servers their pitch goes something like this hi everyone we did really great with energy saving and chips for things like watches and smartphones so uh let us help you save on energy costs for your serversal it's not really like hey we have the best AI technology that's going to make your servers better it's just going to make them more energy efficient is the idea in fact they actually say that right here in the graviton data center chip they don't talk about how it's better they talk about Better Price performance over comparable x86 systems that's basically incorporating the price of the code the chip and the energy consumption now I know that's not super clear but let me just put it this way after having gone through this this company talks about the energy saving aspects of the arm Orchestra architecture more than they talk about artificial intelligence so maybe they're trying to go for a high ESG score which maybe that's realistic that's the thing like maybe maybe we shouldn't be expecting them to be like oh we're really I play we should be expecting them be no like forget about AI for a moment maybe AI is a flash in the pan in which case who has the best potential technology for mobile chips well hands down its arm who has the best potential technology and and code technology for CPUs in servers and computers and otherwise it's arm sorry Intel I know Intel chips x86 are good too but they're just not as good as the risk-based chips at this point maybe that'll change Intel's not going to try to give up on their x86 and those are some risks right x86 from Intel could start being a lot stronger or better or whatever or companies like AMD could start switching to an open source version of risk language writing which then removes the royalty revenues that arm gets remember they're just writing the code for these chips all right what else I do not believe that Tesla uses arm for their self-driving uh Tesla I believe uses their own code but they probably do have arm and some other generic uh chips so that's worth watching uh personally before investing in arm I wrote down that I would let this get through the first six months of lock-ups so that way the executives could dump SoftBank could dump you get rid of the Margin Call risk which would be a huge downside uh you get rid of the employees dumping and in the meantime as everyone else buys arm because they don't actually do their DD which is fine maybe they'll make a lot more money than I'm going to on this and you know what I hope you do I am happy for other people to make money but uh you know I think as people throw money at arm they'll actually be a buying opportunity and the other competitors like AMD Nvidia Intel I don't care okay I really don't care what you buy uh okay so uh they do mention they have some limited customer risk that might be apple they don't specifically say but like 30 of their accounts receivable as of the second quarter uh so ending like June 30th we're from One customer they didn't say which one and then the next biggest customer was slightly under 10.
uh they emphasize sustainable growth probably because they're realistic about AI not exploding growth that's actually probably a good thing I don't need a flash in the pan investment okay I want something that's going to do well for a long term so like I don't want to sound too bearish because ai's probably flash in the pan but they are going to Brand this as an AI Nvidia is the only one that might not end up being a flash in the pan for AI right maybe you could argue uh the Microsoft Google's Amazon's I'm not convinced okay anyway SoftBank doesn't necessarily align with your interest right of course softbank's gonna dump out of this so they can go on to uh other Investments this 60 day is actually asterisk because of their margin loan I found out about that later because it's obviously in a different part of the report which is annoying 50 of the revenue comes from smartphones and consumer electronics we already know that section is going through a downturn uh say you know I mean just look up iPhone 15 sales expectations there are any warnings for weakness on this side right multiple references to decarbonization it's it's ESG promoting is what it is uh big bonuses tied to the IPO price specifically whatever price they get on this IPO probably increases what the executives are being paid in addition to the CEO getting 20 mil just cause uh right this is technically a British company okay we talked here some more percentages fine whatever okay so uh what else did I write down I wrote over here oh a little bit of History how Nvidia tried to uh acquire this company and I think as far as asterisks that was sort of the bulk of it oh they did ask asterisk Taiwan as well as a potential risk factor war with China okay that was an insane amount of talking and I'm just gonna bottom line this for you is this an AI play I don't think so is AI a flash in a pan probably except for NVIDIA though we will determine if the Nvidia flash on the pan is a flash in the pan over the next few earning Cycles I have absolutely no idea I just know if somebody's gonna make a crapload of money from this AI Revolution it's gonna be Nvidia how long that lasts I have no idea because here's the reality maybe everybody bought AI chips in q1 because everybody wanted Ai and then all of a sudden Google Amazon and Microsoft are like oh no our AI usage from our customers is actually falling off a cliff all right let's cancel some of the Nvidia orders we had well if people start canceling some of the Nvidia orders they have yeah nvidia's gonna dump it's gonna dump so fast and quite frankly ask yourself how often are you using AI on a daily basis probably very little maybe because you have learned productive ways to use it on a daily basis like I use it on a daily basis but that's okay uh we teach that in the the how to make more money and get shnit done faster course link down below but anyway look bottom line arm IPO I would not touch it until at least six months after second that Chinese revenue is gonna get destroyed especially with China's economy going to crap and I just don't trust it I don't have the transparency into that and I know SoftBank is gonna dump on you because they want to get their money out of this they tried getting their money out with uh Nvidia now they think they can strike it big on the AI hype and I think it's kind of like brilliant for them but it pisses me off because I feel like it's just sort of taking advantage of the hype cycle but then again if you were in their shoes you would do the same damn thing you can't blame them so um wouldn't touch it for six months I want to see growth expectations want to see what happens with China want to see what happens with Nvidia I'm actually not convinced that they're going to be around forever because open source technology means companies can kind of get rid of them and then not have to pay royalties to them if Nvidia can make it in like a Nintendo switch Tetra chip without arm and then they use risk five now they don't pay royalties oh come on man even if that's like 10 that bumps margins bro I'm sorry this was so long if you found it useful consider subscribing oka
Comments 0