Making Sense of the AI Cycle: Where are We Now?

Today, we're analyzing the soaring performance of semiconductor stocks fueled by the AI boom, spotlighting gains from SanDisk and Micron. With the substantial increase in tech giants' capital expenditure projections, now exceeding $800 billion, we can project out the trickle-down effects across the layers of the AI ecosystem. 

From potential bubble concerns, insights from thought leaders, and the future of memory supply in powering AI applications, we're going to get to the bottom of this AI cycle.

------
🌌 LIMITLESS HQ ⬇️

NEWSLETTER:    https://limitlessft.substack.com/
FOLLOW ON X:   https://x.com/LimitlessFT
SPOTIFY:             https://open.spotify.com/show/5oV29YUL8AzzwXkxEXlRMQ
APPLE:                 https://podcasts.apple.com/us/podcast/limitless-podcast/id1813210890
RSS FEED:           https://limitlessft.substack.com/

------
TIMESTAMPS

0:00 AI Investment Insights
3:40 Flow of AI Capital
5:46 The AI Layer Stack
9:47 Rise of Agentic AI
12:29 Demand for CPUs
14:33 Memory
20:54 AI Infrastructure
24:19 Navigating the AI Market

------
RESOURCES

Josh: https://x.com/JoshKale

Ejaaz: https://x.com/cryptopunk7213

------
Not financial or tax advice. See our investment disclosures here:
https://www.bankless.com/disclosures⁠

Creators and Guests

Host

Ejaaz Ahamadeen

Host

Josh Kale

What is Limitless: An AI Podcast?

Exploring the frontiers of Technology and AI

Josh:
If you bought SanDisk stock just a year ago, you're up 40 times your money.

Josh:
Micron, up eight times. Intel and AMD are both up almost four times what you

Josh:
would have put in just a year ago. And we thought this was an AI bubble.

Josh:
We were actually wrong about how it's going to play out. And we have the data

Josh:
to back this up. Now this year, five of the biggest companies in the world are

Josh:
going to spend close to a trillion dollars.

Josh:
As that trillion dollars is spent, it's going to trickle down the stack to a

Josh:
series of layers that are embedded within this AI ecosystem.

Josh:
So we're going to be detectives this episode and track down where all of that

Josh:
trillion dollars in CapEx is flowing.

Josh:
It's clearly leaving the large

Josh:
cap companies. Google is spending 90% of all the money that comes in.

Josh:
Where is it going? Well, we are going to walk you through everything.

Josh:
So hopefully by the end of this, you'll have a pretty good idea of the structure

Josh:
of what the AI investment universe looks like.

Josh:
And you can make up your own mind on where you think the best place is to allocate

Josh:
your dollars to collect the money that's flowing from these large cap companies.

Ejaaz:
Now, I've really had to ground myself over the last couple of weeks because

Ejaaz:
it felt like I'm in the land of make-believe.

Ejaaz:
Look at some of these stocks and how much they've increased over the past year.

Josh:
This is crazy.

Ejaaz:
This is absolutely insane. AMD is up three and a half X and 25% of that move

Ejaaz:
was literally over the last 12 hours, right?

Ejaaz:
Because they reported earnings, they crushed it. SanDisk is up 40%, as you mentioned.

Ejaaz:
ARM, Intel, all of these are in the infrastructure layer of AI,

Ejaaz:
which has become a very consensus trade.

Ejaaz:
And that's for many different reasons. And we're going to address that later on in the stack.

Ejaaz:
But there's one clear thing that changed over the last week.

Ejaaz:
And that came in the form of the major earnings reports for Q1 of 2026 from

Ejaaz:
four of the largest hyperscalers.

Ejaaz:
You've got Amazon, Meta, Microsoft, and Google. Now, there was a problem. There was an issue.

Ejaaz:
People thought that they were going to spend hundreds of billions of dollars.

Ejaaz:
And they originally committed to that being around $630 billion in 2026,

Ejaaz:
but no one knew how they were going to make money.

Ejaaz:
Q1 proved that them spending all that money not only resulted in more profit

Ejaaz:
and more revenue that they had generated, but also that they were going to revise this.

Ejaaz:
They were going to increase their spend in 2026 to a total of $800 billion,

Ejaaz:
with next year projecting $1.1 trillion.

Ejaaz:
This is just from four to five companies. I have to emphasize how nutty and crazy that is.

Ejaaz:
But again, this translates into actual revenue or money being earned.

Ejaaz:
I'm going to show you a few different stocks that kind of demonstrate or portray

Ejaaz:
that. So firstly, Sandisk.

Ejaaz:
40x over the last year. This is just like an astounding thing where like,

Ejaaz:
you know, a 40x return on any kind of major cap stock is just completely unheard of.

Ejaaz:
Then you have Intel that operates in the CPU kind of infrastructure packaging

Ejaaz:
layer of AI GPUs is up four and a half X. What else have we got? We've got AMD.

Ejaaz:
I mentioned that it's up 320% over the last year, but literally today it is

Ejaaz:
up 17%, absolutely crushing it.

Ejaaz:
And then we have Micron technology, which plays in the AI memory trade is up

Ejaaz:
7.25x over the last year.

Ejaaz:
And then there's this really cool investment vehicle where there's a bunch of

Ejaaz:
different AI memory trades.

Ejaaz:
You can get access to just a basket via this ETF, Roundhill Memory ETF,

Ejaaz:
something I participated in.

Ejaaz:
That only launched just over a month ago, Josh, and it is up 72%.

Ejaaz:
Just insane gains on these stocks.

Josh:
Do you have FOMO? Because I do. I'm feeling the FOMO. And I think the question

Josh:
and part of the concern that a lot of people are going to be feeling as they're

Josh:
seeing these numbers is like, oh, my God, did I miss this?

Josh:
It seems like very obviously we're in some sort of a bubble.

Josh:
But what stage of that bubble are we in?

Josh:
How should we deal with this? How should we navigate this? Well,

Josh:
Brad Gerstner, actually, Altimeter Capital investor, he's like a very prominent

Josh:
thought leader in the space.

Josh:
He went on CNBC yesterday to talk just about this, about where we are in the cycle.

Josh:
And his idea is that this time is different.

Josh:
And these are very famous words. We've heard this time is different a lot,

Josh:
but he explained why. And a lot of it comes from this capital expenditure.

Josh:
I mean, he just mentioned $800 billion is being spent this year.

Josh:
Over a trillion dollars is being spent next year. And a lot of that money is

Josh:
flowing from these large mega cap companies into infrastructure required to

Josh:
build out more AI use cases.

Josh:
As of now, these AI use cases are actually printing a good bit of money.

Josh:
I mean, it comes from that top where we see Anthropic and OpenAI printing billions

Josh:
of dollars of ARR per week. Google just became the most valuable company in the world.

Josh:
And it seems to be like we have this flywheel that is sustainable so long as AI is useful.

Josh:
And in the case that AI continues to be useful and we unlock new use cases that

Josh:
are more valuable than what they are today, then the amount of money people

Josh:
will pay for it will continue to go up.

Josh:
And those capital expenditures that are being priced into these earnings reports

Josh:
can then be priced into these further concentric circles out from the core.

Josh:
So if you'll notice, we didn't mention NVIDIA on one of these most valuable,

Josh:
like the largest gainer stocks, because they're kind of sitting at the top in

Josh:
which a lot of the capex is coming from.

Josh:
So what we're going to do now is kind of walk through the layers and the stack

Josh:
of where this money flows.

Josh:
So if you think about it, all of the money comes in through companies like Google,

Josh:
through companies like Amazon, who are collecting income from retail.

Josh:
Retail is spending on their goods and services.

Josh:
They're delivering a lot of value to them they take that money they redistribute

Josh:
it where is that going is the question we're going to answer so over the last

Josh:
few months we've seen this happen in semis a lot of those um

Josh:
A lot of those companies that you showed us, EJES, are semiconductor stocks. It's moving into CPUs.

Josh:
It's moving into GPUs. But let's start with the layer zero that we have here

Josh:
on the artifact, EJES. What's going on with the layer zero?

Josh:
What is at the base layer of this stack?

Ejaaz:
Okay, so you mentioned that four of the hyperscalers, Google,

Ejaaz:
Microsoft, Meta, and Amazon, are kind of taking inflows from retail.

Ejaaz:
There are currently two ways, or rather two startups and companies,

Ejaaz:
that funnel 99% of that retail. And they're called OpenAI and Anthropic.

Ejaaz:
They're both private companies. So you can't even access them publicly via equities

Ejaaz:
right now, but they contain the bulk of ChatGPT and Claude users, right?

Ejaaz:
So that in itself is kind of scary to kind of read, but the fact is they have

Ejaaz:
paying customers, they have paying enterprises, and that money is where the

Ejaaz:
fountain starts, the waterfall starts.

Ejaaz:
And from OpenAI and Anthropic, it flows down into what we're calling the layer

Ejaaz:
one, which is kind of like the platforms, hyperscalers, and also Model Labs,

Ejaaz:
Google counts itself as one as well.

Ejaaz:
Okay, so you've got Google, which doesn't actually just act as a Model Lab.

Ejaaz:
They have TPUs, they have the infrastructure, they have the distribution layer,

Ejaaz:
they have the cloud infrastructure, which is the main funneler of all this revenue

Ejaaz:
that they're making from their recent earnings.

Ejaaz:
You've got Amazon that's mentioned here as well, which is doing the same with AWS.

Ejaaz:
And you have Microsoft, which is doing the same through Azure.

Ejaaz:
Meta, who's doing the same from their social media platform.

Ejaaz:
And we've included opening an Anthropic here because some of them sometimes

Ejaaz:
play in the infrastructure layer.

Ejaaz:
So the first stack is, okay, we have OpenAI and Anthropic dealing with the retail.

Ejaaz:
That money and revenue flows into the hyperscalers that can provide compute

Ejaaz:
and distribute their models.

Ejaaz:
So very important. Google Cloud, AWS, and Microsoft Azure, all their cloud computing

Ejaaz:
services, distribute the model to all their enterprise customers and governments

Ejaaz:
and customers that actually want to, retail customers that actually want to

Ejaaz:
end up using these things.

Ejaaz:
And then, as you mentioned, Josh, we move into layer two, which is what we're

Ejaaz:
calling kind of the GPUs and semiconductor area, right? Now,

Ejaaz:
this is where the narrative breaks.

Ejaaz:
And classically, we have been told AI is going to do so well.

Ejaaz:
Demand is going to increase. Where's the best way to buy the picks and shovels?

Ejaaz:
Well, it's NVIDIA, of course, right? They're the most valuable company in the world, right, Josh?

Josh:
And that was correct. Yes, they were a $4 trillion company.

Josh:
As of today, that has changed. Google has taken the crown again.

Josh:
And I think it's because we're kind of reaching further out of this risk curve.

Josh:
The money is starting to propagate further out. So as it started,

Josh:
we needed to go from zero to one.

Josh:
We needed to go from no AI to AI. And what was required of that?

Josh:
Lots and lots of GPUs, lots and lots of intelligence. And that's where NVIDIA

Josh:
was most valuable because they were able to provide the GPUs to spin up the

Josh:
AI, to take the industry from zero to one.

Josh:
Now that we have AI, now that we have established business and revenue streams,

Josh:
that money is flowing to optimize the stack.

Josh:
So Jensen and NVIDIA are going to continue to make a tremendous amount of money,

Josh:
but that's kind of baked into the valuation and we've seen that run up.

Josh:
I mean, they're the second most valuable company in the world right now.

Josh:
It's a tremendous amount of market cap that they've absorbed over the last, call it 36 months.

Josh:
Now the time is come to cycle further

Josh:
out this risk curve to move beyond the picks and shovels into this

Josh:
next layer of the stack that even sits below gpu consumption and my understanding

Josh:
he does is it's not just gpus now there's also cpus and that has to do a lot

Josh:
with the the new agentic trend that we're seeing where agents are kind of being

Josh:
proactive and they need orchestrators and while gpus are really good at solving

Josh:
complex uh mathematical problems to produce inference

Josh:
You need smarts and you need brains in order to kind of orchestrate these models.

Josh:
And that's what's key about this next infrastructure layer that's going to really

Josh:
be a pretty big deal in the space.

Ejaaz:
That's exactly it. So I've got this article pulled up where Jensen went on a

Ejaaz:
stage right now and he talked about GPUs and the demand increasing ever so highly.

Ejaaz:
And he goes, consumption is going through the roof for GPUs with the rise of

Ejaaz:
agentic AI in the last several months. Now, I want to unpack this phrase very

Ejaaz:
specifically, because it explains why Intel and AMD are absolutely skyrocketing right now.

Ejaaz:
Agentic AI refers to AI agents. AI agents is kind of like, think of like an

Ejaaz:
instance of a ChatGPT or a Claude code that can just go off and do a bunch of

Ejaaz:
things autonomously for you.

Ejaaz:
So you don't have to type to it, you don't have to speak to it,

Ejaaz:
you don't have to prompt it.

Ejaaz:
It just goes off and works autonomously. Now, you can imagine the market for an AI autonomous worker

Ejaaz:
It's pretty large. The TAM is pretty huge. It's pretty much any sector that

Ejaaz:
involves a computer for now until robots actually become a thing.

Ejaaz:
But there's one kicker. This is the narrative violation, Josh.

Ejaaz:
Guess what you need to allow the AI agents to use the tools to go off and do that work?

Josh:
They need the brain. They need the CPU.

Ejaaz:
The CPU.

Josh:
Computer processing unit versus the graphical processing unit.

Ejaaz:
Right. So let me take you through a little historical context here.

Ejaaz:
Now, if we rewind to, let's take GPT 4.0, right? It was a breakthrough model. Everyone loved it.

Ejaaz:
Can you guess how many CPUs were used to train or inference that entire model?

Ejaaz:
Like, what's the ratio roughly? Do you think how many- CPUs or GPUs?

Ejaaz:
CPUs. So how many CPUs did you need for like the average GPU?

Ejaaz:
Like, what do you think that ratio was?

Josh:
I would guess close to zero.

Ejaaz:
Yeah, you would be correct. We barely even used it.

Ejaaz:
Fast forward to today, that ratio is almost one-to-one. Let me rephrase that. Let me reframe that.

Ejaaz:
You need one CPU core per GPU, and that trend is going to flip over the next

Ejaaz:
six months where the number of CPUs will outweigh GPUs.

Ejaaz:
So basically, Intel and AMD, who kind of built their bread and butter of profit

Ejaaz:
margins off of CPUs that were used for gaming and a bunch of other stuff,

Ejaaz:
now has found themselves in an absolute goldmine of an industry which requires

Ejaaz:
the things that they've been building for over decades.

Ejaaz:
So Intel and AMD are like, well, okay, I'll spin up as many CPUs as you want.

Ejaaz:
And Jensen's like, I need all these CPUs so I can spin up all these data racks.

Ejaaz:
Now, to help you understand why it is needed specifically, Josh,

Ejaaz:
you mentioned AI agent orchestration, right?

Ejaaz:
So let's say you spin up a bunch of AI agents. They need to,

Ejaaz:
guess what, interact with other AI agents. They also need to use tools.

Ejaaz:
And most importantly, they need to use these tools faster than humans can themselves.

Ejaaz:
Josh, on yesterday's episode, you mentioned what you loved about Codex specifically

Ejaaz:
was the browser use and the fact that it's so quick.

Ejaaz:
You love that. The reason why it's able to do that is because it has access to more CPUs.

Ejaaz:
It allows you to kind of run C compiler and a bunch of other things.

Ejaaz:
So the long story short is CPUs are in huge demand. And that's why Hymdi stock is up 15% today.

Ejaaz:
And it's probably going to be up what's up three and a half X over the last couple of years.

Ejaaz:
So they just reported their earnings. Intel did the same about two weeks ago.

Ejaaz:
And there's a huge demand for CPUs in general.

Josh:
And this is kind of the natural extension of the way that AI is moving.

Josh:
This hasn't always been the case, but we went from basically LLMs,

Josh:
right? Where we just used GPUs. It was a chatbot. You type into an interface.

Josh:
Then we went into reasoning and chain of thought where there was a lot more

Josh:
tokens needed in order to answer the same question.

Josh:
Then we moved into the agentic era where things like OpenClaw and the claw agentic

Josh:
swarm operating system came into play. And that's when this trend changed again.

Josh:
So assuming these trends are going to continue changing, the paradigms are going

Josh:
to continue to shift. But the one singular core truth throughout this is that

Josh:
we need more tokens. And no matter how we get them is up for debate,

Josh:
like how those are going to be generated.

Josh:
But the fact is that no matter what we do, every paradigm shift has resulted

Josh:
in a gigantic increase in more tokens, which seems like is a trend that you continue to bet on.

Josh:
And one person who bet on this pretty bigly was Donald Trump and the US government,

Josh:
like us, for the first time, like we're making money.

Josh:
Trump bought Intel and he's up what, 500% so far?

Ejaaz:
500 percent um so uh the government

Ejaaz:
uh about a year and a half ago

Ejaaz:
took a 10 stake in intel and

Ejaaz:
their primary reasoning for that was we have

Ejaaz:
too much reliance currently in the u.s on this one company called tsmc which

Ejaaz:
is based in taiwan which china presumably wants to take over at some point it's

Ejaaz:
too much of a geographical and national risk and so we wanted to kind of bring

Ejaaz:
a bunch of tsmc's capabilities on shore and that was expressed in the form of

Ejaaz:
intel who doesn't just build CPUs, by the way,

Ejaaz:
they're working on building a bunch of frontier GPUs. And that's a story for another day.

Ejaaz:
But the point is, Trump bought 10% via the government, and they are up 5x.

Josh:
You know what price you bought at $20.47? It's trading at $111 today is what

Josh:
the ticker is trading at.

Josh:
So shout out to the US government pumping our bags and winning on behalf of

Josh:
the people. That's very exciting.

Ejaaz:
Yeah. Okay, so let's continue down this waterfall. So what we've started off

Ejaaz:
with is Anthropic and OpenAI as the retail that flows down into the hyperscalers,

Ejaaz:
the Google, the Amazons, the cloud distributors, their revenue margins are expanding.

Ejaaz:
Okay, this is amazing, but they rely on semiconductors.

Ejaaz:
They need Jensen Huang's GPUs, but Jensen Huang needs all these CPUs because

Ejaaz:
all these AI agents are using all these different tools.

Ejaaz:
Okay, so we need CPUs. Now, can you guess? There's another component,

Ejaaz:
Josh, that makes up 50% of the bill of materials cost for a GPU. 50%.

Ejaaz:
Guess what it is?

Josh:
What does every single GPU on the planet need? Jensen, what does he need?

Josh:
What is his bottleneck choke point that doesn't work?

Josh:
Like the company doesn't work in the absence of it's memory.

Josh:
Memory is the biggest thing in the world. It's impossible to make enough memory. No one has it.

Josh:
And therefore any of the memory stocks that you have invested in,

Josh:
in the last year have gone absolutely nuclear.

Josh:
And these are probably among the biggest winners. When you think about SanDisk

Josh:
delivering a 40 times return, Micron is looking like a seven times return.

Josh:
Memory has been the choke point of all of these because memory is the most important thing.

Josh:
It's where we get our context windows from. It's where a lot of the training data is stored.

Josh:
Memory is the next layer of the stack.

Ejaaz:
Yes, exactly. And if you thought NVIDIA had a monopoly on GPUs,

Ejaaz:
let me introduce you to the secret monopoly of memory.

Ejaaz:
So there are basically four companies which dominate in the AI memory landscape.

Ejaaz:
Their names are Micron, which is a U.S.

Ejaaz:
Company, SK Hynix, which is, by the way, the biggest memory provider,

Ejaaz:
Korean company, Samsung, second biggest, and then we have SanDisk.

Ejaaz:
Now, you might assume that they all make the same types of memory. Three of them do.

Ejaaz:
Micron, SK Hynix, and Samsung make something called high-bandwidth memory.

Ejaaz:
This is the premium memory that goes into making the GPUs, the Rubin Ultras,

Ejaaz:
the fashionably new GPUs that NVIDIA releases every single year.

Ejaaz:
That incorporates 50% of it.

Ejaaz:
Our bill of materials is this fancy high bandwidth memory.

Ejaaz:
It is an extremely complex thing to make. The supply is super constrained and

Ejaaz:
only these three companies dominate in making it.

Ejaaz:
And that is for a different reason. I'll explain the memory cycle later on.

Ejaaz:
But then you have SanDisk, which is up 40%. And the reason why they're up 40%

Ejaaz:
is because there's a second type of memory which is required by these AI models.

Ejaaz:
And it's called NAND or N-A-N-D, NAND flash. Now, here's the difference between the two.

Ejaaz:
High bandwidth memory, which is the original GPU premium, basically allows you

Ejaaz:
to move data really quickly in the AI models.

Ejaaz:
Think about it, right? These AI models are like 10 trillion parameter large,

Ejaaz:
at least Mythos and GPT 5.5.

Ejaaz:
They need access to data very quickly, and it's clunky. It's very big.

Ejaaz:
High bandwidth memory basically solves that. It allows you to kind of store

Ejaaz:
the memory and like move that memory really quickly.

Ejaaz:
But there's the second type of memory, which SanDisk specializes in, called NAND.

Ejaaz:
And what that does is when you're having a conversation, Josh,

Ejaaz:
have you ever noticed that the models now are really good at maintaining the

Ejaaz:
context and remembering things

Ejaaz:
that you mentioned like a few sentences before? Have you noticed that?

Josh:
For a longer period of time, yes. The context windows have expanded quite a bit now.

Ejaaz:
Yes, exactly. Now, the main enabler for that is this NAND storage,

Ejaaz:
which is kind of like a temporary memory storage.

Ejaaz:
It's more of a commodity. It's not as sexy as high bandwidth memory,

Ejaaz:
but it's super important.

Ejaaz:
SanDisk dominates that entire sector, which is why it's up 40X.

Ejaaz:
So memory has famously gone under a huge supply constraint.

Ejaaz:
These providers are sold out, and I'm not exaggerating this, until the end of 2028.

Ejaaz:
They have signed customer contracts until the end of 2028.

Ejaaz:
They don't have enough memory, they don't have enough supply,

Ejaaz:
and so ramping that up is going to be the key focus over the next couple of years.

Josh:
And this has been really felt in the retail market as well. If you use memory

Josh:
for traditional use cases, like building PCs or just general consumer products,

Josh:
all of those have either been delayed or the prices increased because it is

Josh:
so difficult to get your hands on this memory.

Josh:
And when you think about this memory, there's a pretty simple analogy that I

Josh:
was using earlier today when it came to describing and understanding how this works.

Josh:
So high bandwidth memory, like you mentioned, it's known as HBM for short.

Josh:
It's basically a series of DRAM stacks stacked on top of each other.

Josh:
And you can think of those as this like this huge desk right next to the AI

Josh:
model that they can very quickly reference.

Josh:
And then the NAND is the file cabinet that's kind of right next to it that is

Josh:
more persistent that holds these ideas for a longer period of time it's larger

Josh:
but it is a bit slower and the dynamic between the two is really interesting

Josh:
now one of the things that was

Josh:
kind of a narrative violation as i was reading about it is the idea that

Josh:
Deep Seek can actually create more efficient models in terms of how much they

Josh:
use memory, and yet the demand for memory goes up.

Josh:
So Deep Seek, they published this famous paper that allowed them to get basically

Josh:
frontier level intelligence using a small fraction of the amount of memory that

Josh:
traditional AI labs have used. This seems like a bad thing.

Josh:
We need less memory because we've become more efficient, but the reality is,

Josh:
is that the inverse actually happened, where now we have greater memory efficiency,

Josh:
but far greater memory demand.

Josh:
So maybe you could explain this dynamic that's happened with this,

Josh:
because there's a strange deep seek paradox going on that I think is a narrative

Josh:
violation in which people kind of look for when seeing things that can pop this bubble.

Josh:
And the reality is, is that this is this is not even a little small tear in the bubble.

Josh:
It's actually improving the quality of the money spent. How is this working

Josh:
with this memory dynamic here?

Ejaaz:
Yeah, it's basically Jevin's paradox. So the problem that you're explaining

Ejaaz:
is if 50 percent of this GPU thing is reliant on this one component supplied

Ejaaz:
by four different companies, they have to work on a workaround, right?

Ejaaz:
They're probably going to create GPUs or models that don't rely high on memory.

Ejaaz:
And Deep Seek version 4, which was released a few weeks ago,

Ejaaz:
was the instantiation of that.

Ejaaz:
It uses, I think, like 5% to 15% of a Claude Opus 4.7 model,

Ejaaz:
which is a drastic reduction, which may lead you to think that memory stocks

Ejaaz:
are going to crumble, except the actual opposite happened.

Ejaaz:
So I've got this block of text here, but I'm going to explain it in very simple terms.

Ejaaz:
What they found was the architecture unlock that DeepSeq v4 created actually

Ejaaz:
ended up using more of that NAND flash memory because their architecture change

Ejaaz:
was, we'll just use more agents.

Ejaaz:
We'll let more agents do the thinking before we give someone an answer to their prompt.

Ejaaz:
And it resulted in a smarter effort. But that didn't decrease the reliance on

Ejaaz:
HBM or memory in general.

Ejaaz:
Increased it overall. You'll see here that each prompt required 157 rounds and

Ejaaz:
a bunch of token contacts, all of that interfacing with these memory components.

Ejaaz:
So the actual opposite happened. And this is Jeven's paradox playing out in

Ejaaz:
reality, where if you assume that the demand for goods, or rather the cost of

Ejaaz:
goods goes down, you'll assume demand goes down because it's like cheaper,

Ejaaz:
you don't make as much money.

Ejaaz:
But the in fact, opposite happens where demand goes way higher,

Ejaaz:
because now you can do more things for cheaper. So Jevon's paradox playing out here.

Josh:
Okay, next layer of the stack, power generation and infrastructure.

Josh:
How are we going to plug in these GPUs? Well, we need infra for it.

Josh:
Otherwise, they're just going to be sitting there dark with no power and no

Josh:
ability to turn them on. This is layer five of the stack.

Josh:
This is where you will recognize names, perhaps like Bloom Energy,

Josh:
which we have famously mentioned.

Josh:
Our boy Leopold is up like what, four or five times return on Bloom Energy.

Ejaaz:
Yeah, like two bills.

Josh:
Yeah, like doing pretty well. And this is the part of the stack that is,

Josh:
I guess, core and sits a little bit downstream of the GPUs. But again,

Josh:
contingent on it working.

Josh:
Tell me a little bit about this company, particularly, I mean,

Josh:
Corning. We had a deal with Corning Opio today and NVIDIA, which is pretty big.

Josh:
I would imagine most people listening to this have never even heard of Corning before.

Ejaaz:
Yes, exactly. So I have to be upfront. This is where I'm weakest in this stack or this.

Ejaaz:
So I have to be upfront and maybe we'll do an episode later down when we've

Ejaaz:
done a bit of research. But basically, if you look at the stack so far,

Ejaaz:
it flows down from retail, it goes through semiconductors, it goes through CPUs,

Ejaaz:
and it goes through memory.

Ejaaz:
Where does the puck flow next? Well, there's an issue.

Ejaaz:
I buy all these GPUs, I buy all these CPUs, I set them all up neatly in a warehouse,

Ejaaz:
in a stack, in a server rack.

Ejaaz:
But I don't have enough power to power the thing. Or if I do have the power,

Ejaaz:
I don't know how to regulate the power to the right GPUs and the right CPUs

Ejaaz:
at the right time to prevent a blackout, but not overheat it.

Ejaaz:
And then I have a cooling system.

Ejaaz:
There's a whole architecture around the GPUs and the CPUs operating at optimal form.

Ejaaz:
There was an article that was released by The Information this week,

Ejaaz:
which showed that XAI, which has the largest cluster of GPUs,

Ejaaz:
over a million high-grade NVIDIA GPUs, only utilize 11% of its power.

Ejaaz:
That's because they don't have enough power generation or the chips or the architecture

Ejaaz:
to allow this power to flow directly through GPUs.

Ejaaz:
I think this is where the puck is flowing next.

Ejaaz:
And today's announcement, where NVIDIA is partnering with Glassmaker Corning,

Ejaaz:
which is focused on optics specifically, is part of that puzzle.

Ejaaz:
But there's also another piece, which

Ejaaz:
is basically the power suppliers or infrastructure providers on its own,

Ejaaz:
mainly referenced by companies like GE, Vinova, CEG, Constellation Energy,

Ejaaz:
which we had on our stack just now,

Ejaaz:
which not only supply the power and make sure the power ends up at the data

Ejaaz:
center, but also regulates and makes sure that the power enters at the right

Ejaaz:
time to the CPUs, to the GPUs, to make sure they're optimally performing.

Josh:
And this is layer five of the six-layer stack, the sixth and final layer being the raw materials.

Josh:
Now, this is kind of the bare-bones foundational layer. When you think about

Josh:
The AI stack from first principles, what is required at that foundation,

Josh:
it is the raw materials in to get the intelligence out.

Josh:
We've turned sand and silicon into thought.

Josh:
And throughout that entire process, there is a tremendous amount of technology empowering it.

Josh:
Now, we are not proficient currently in materials, but if you would like us

Josh:
to be, and if this is an episode of interest, we can go deeper on layer five

Josh:
and layer six of the stack.

Josh:
Because I was looking at a chart this morning from of lithium carbonate,

Josh:
which is just a very critical component to a lot of the AI stack.

Josh:
I'm looking at a chart from November where it was priced at $75,000,

Josh:
and now it is at $187,000.

Josh:
So the gain and return on some of these materials has been just unbelievable.

Josh:
And this is the sixth and final layer of the stack. So perhaps in a future episode,

Josh:
but hopefully this has given you kind of a loose orientation of how you could

Josh:
think about the trickle down economics of AI, how it kind of starts from this

Josh:
layer zero and works its way through the infrastructure.

Josh:
Now, EJs, we have to kind of orient ourselves in reality.

Josh:
Good things don't last forever and there's

Josh:
a high probability that this one doesn't perhaps this time

Josh:
is different but there is a case to be made that it is not because

Josh:
traditionally memory has done well in the past there

Josh:
have been a series of memory booms that have done very well every single one

Josh:
of which has followed up with a bust so where would you say we are currently

Josh:
um in the cycle or at least what are the downsides that people should be looking

Josh:
out for and aware of that would let them know that hey maybe things are looking a little frothy here

Josh:
there's some cause for concern.

Ejaaz:
Yeah, so the first factor is the boom and bust cycles, particularly with memory,

Ejaaz:
which I think is, what was it, layer three, layer four of the sec that we just described.

Ejaaz:
It has traditionally gone through many boom and bust cycles,

Ejaaz:
which I'm showing on our screen here through this chart.

Ejaaz:
Now, what isn't represented here is.

Ejaaz:
At the start, on the left-hand side of this chart, we had 14 key memory providers.

Ejaaz:
Fast forward to today, and we have three for high bandwidth memory and one for NAD.

Ejaaz:
So four. And the reason why is every single boom and mainly the bust has crushed

Ejaaz:
out specific companies to leave like the three that we have right now.

Ejaaz:
And so history would tell us that the same thing is going to happen this time.

Ejaaz:
Now, I'm conflicted here, Josh, because I'm sitting here as a podcaster and

Ejaaz:
saying this, but I listened to three podcast episodes from the memory execs

Ejaaz:
from all of these different companies yesterday.

Ejaaz:
Guess what they said? They said, this time is different. They said,

Ejaaz:
this time is a very unique opportunity because not only is AI surging demand

Ejaaz:
for all of these memory chips and they've got payments for all of these things,

Ejaaz:
on the front end, the AI products themselves are being paid for.

Ejaaz:
They're actually making money. Now, the definition of a bubble is it's levered.

Ejaaz:
It's a bunch of hoopla, right? There isn't actually any money coming in.

Ejaaz:
This disproves that narrative, the bubble narrative, because there's money being

Ejaaz:
paid off. None of these companies are levered up.

Ejaaz:
That 800 billion CapEx number that we started this episode with is all coming

Ejaaz:
from cash flows that they have right now or cash reserves that they have right

Ejaaz:
now. No one is levered up. No one's borrowing money for this.

Josh:
Yeah, in 2026 and 2027, I mean, that's $2 trillion of guidance that comes from cash flow.

Josh:
It comes from Google spending 90% of the money they actually make.

Josh:
So there isn't leverage. This is somewhat sustainable.

Josh:
And now that we have the rough CapEx guidance, people can kind of price in what

Josh:
this will look like when it impacts the market.

Josh:
And that is the current layout of the land.

Josh:
This is everything you need to know about the full AI stack,

Josh:
what has been doing well, what hasn't.

Josh:
And I guess the question for you is now that you're oriented or whoever is listening

Josh:
is what part of the stack are you most interested specifically what companies

Josh:
are most interesting because there is this gigantic hot ball of money that is

Josh:
moving its way through the market for the first time in a very long time.

Josh:
This is a paradigm shift for the last decade or two.

Josh:
All of the money has kind of accumulated at the top of these funnels has accumulated

Josh:
to Google to Amazon to these large mega cap companies, but now it's working its way out.

Josh:
And where is that money going to end up? It is TBD. We've seen semiconductors.

Josh:
We've seen memory do really well will they continue tbd where

Josh:
else is the money flowing to we're not sure but the goal

Josh:
of this episode is just orientation right you kind of now have a lay of the

Josh:
land you have an idea of where everything is and can maybe go out and pick some

Josh:
winners so i guess the prompt for this episode is who are the winners who's

Josh:
going to win this um battle of capex accumulation is is what i would love to

Josh:
know because i will plan and allocate my portfolio accordingly

Ejaaz:
Yep, same here. My bags is the answer, Josh.

Ejaaz:
But right now, I think that we are kind of in a bubble, but it's a different

Ejaaz:
type of bubble than what we're used to because no one's levered up,

Ejaaz:
as I mentioned earlier on.

Ejaaz:
And it remains to be seen. Like every quarterly earnings, I'm going to evaluate

Ejaaz:
everyone's spec sheet and see if they're actually making money from this thing.

Ejaaz:
Q1 earnings, which released over the last two weeks, tells us that it is.

Ejaaz:
And listen, there are always levered bets everywhere.

Ejaaz:
Like if you don't buy any of the companies that we You mentioned there are smaller

Ejaaz:
cap stocks that you might go after.

Ejaaz:
They are high risk with all of these different types of investments and the

Ejaaz:
narrative can change, as we know, in AI in a second.

Ejaaz:
So the long story short is, neither of us know, but we will be keeping track

Ejaaz:
of everything and we'll be updating you.

Ejaaz:
Josh's prompt earlier on is genuine for me as well.

Ejaaz:
If you want to hear more about lower layers of the stack, we will do the research

Ejaaz:
and we'll deliver you that episode.

Ejaaz:
Let us know in the comments and let us know what stocks you're investing.

Ejaaz:
But aside from that, Josh, I think we're done.

Josh:
That's it. Thank you guys so much for watching. As always, share it with your

Josh:
friends who might be interested in an episode like this, and we will see you guys in the next one.

Ejaaz:
See you guys.