Limitless: An AI Podcast

Ejaaz:
Earlier this week, the most powerful people in finance stood around a table

Ejaaz:
next to Jensen Huang and announced they'd raised $500 billion to buy NVIDIA GPUs.

Ejaaz:
Now, if you're listening to this and you're thinking this is just an AI bubble

Ejaaz:
circular economy type thing, you might not actually be wrong.

Ejaaz:
Larry Fink, the head of BlackRock, actually likened this deal to mortgage-backed

Ejaaz:
debt securities of the early 2000s.

Ejaaz:
And if that sounds familiar, to which he created. Yes, to which he created.

Ejaaz:
And if you're likening that to a kind of like a PTSD flashback,

Ejaaz:
that's because that's exactly what happened in the 2008 financial crisis itself.

Ejaaz:
But if you look at the news in general, if you look at the way that this deal

Ejaaz:
is structured, it might actually hint at something completely different. In fact, the opposite.

Ejaaz:
GPU prices for renting has gone sky high. It's up 40% on the year.

Ejaaz:
And there's not enough GPUs to back a lot of the deals that Microsoft,

Ejaaz:
Google, Anthropic and OpenAI are signing with NVIDIA. So the question that we're

Ejaaz:
going to unpack on the show is, is this very much a bubble back deal?

Ejaaz:
Or is this something completely different that we're missing?

Josh:
A new paradigm of investing, a new paradigm of financial manufacturing and construction.

Josh:
This is a new investable asset class. Yeah, this is an entirely new thing.

Josh:
GPUs. Who would have thought?

Josh:
Michael Burry, the guy who's responsible for the big short, he was like,

Josh:
no, the price of these things are going down only. It turns out he could not have been more wrong.

Josh:
And now Jensen has assembled Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR.

Ejaaz:
It's the Avengers of finance. It's the Avengers of finance.

Josh:
It's like Apollo alone has a trillion dollars.

Josh:
Of assets managed. Blackstone has over 1.3 trillion.

Josh:
Brookfield is over a trillion. And combined, that's 3.4 trillion.

Josh:
BlackRock is bigger than all three of those combined.

Josh:
And they're all doing this together. And together, they've signed this thing

Josh:
called a memorandum of understanding. Now, I had to actually look up what this

Josh:
means because I had no idea.

Josh:
A memorandum of understanding, or an MOU, is a formal, usually non-binding document

Josh:
signed by two or more groups.

Josh:
It shows that the groups share a common goal and plan to work together.

Josh:
So this is not a contractual obligation. We have to start with that.

Josh:
This is not a guarantee that $500 billion is going to flow into this new economy.

Josh:
But it is an intention that all of these people are going to be aligned to work

Josh:
together towards funding this next build out of AI. And what I found most interesting

Josh:
is that this was actually orchestrated entirely by Jensen.

Josh:
Jensen reached out to all of these banks himself personally.

Josh:
And he said, hey, I'd like to work together on this thing. And not a single

Josh:
bank that he reached out to said no.

Josh:
So here we are now with a moment on CNBC in which they're all sitting around

Josh:
a table talking about how they are committing $500 billion to this new asset class.

Josh:
And this is unbelievable. This feels like a, for better or worse,

Josh:
a brand new paradigm for the AI era in which now the collective force of the

Josh:
United States banking system is like starting to get behind this.

Josh:
And I should say this is not for AI as a whole. This is purely for NVIDIA as a company.

Ejaaz:
Yeah, and I want to take a moment to actually explain what's happening here,

Ejaaz:
because I think there's a lot of confusion.

Ejaaz:
There's a lot of headlines that people are getting worried over.

Ejaaz:
NVIDIA stock tanks 4% on the news, but I think that's a little too early to

Ejaaz:
judge. So firstly, what does this structure sort of look like?

Ejaaz:
Well, it's what you're seeing on the screen right now, which is essentially.

Ejaaz:
There's a problem in AI right now, which is all these hyperscalers,

Ejaaz:
all these AI labs, Anthropic, OpenAI, Google, Meta, you name it,

Ejaaz:
have spent a lot of money to buy GPUs.

Ejaaz:
The reason why they're doing this is to train and inference brand new AI models,

Ejaaz:
which they have a lot of paying customers for.

Ejaaz:
But the issue they're facing is the money they've invested, which is now to

Ejaaz:
the tune of $2.6 trillion converted over the next couple of years,

Ejaaz:
I believe, is not enough for them.

Ejaaz:
So much so that they're going into negative cash flow. So what happens when

Ejaaz:
you've spent all the money that you have in your company, in your balance sheet?

Ejaaz:
You need to go to Wall Street. That's exactly what NVIDIA, specifically Jensen, has brokered.

Ejaaz:
He's gone to Wall Street and he said, listen, we need more money to build more

Ejaaz:
GPUs to sell to these different customers so that they can produce their products

Ejaaz:
and services, their new models.

Ejaaz:
And Wall Street has gone back and said, I have an issue with this,

Ejaaz:
Jensen, which is GPUs aren't a versatile asset. like they can only be used for

Ejaaz:
one thing specifically, which is either training a model or inferencing a model,

Ejaaz:
and it's only customer-specific.

Ejaaz:
And Jensen responded to them and said, that's not true at all.

Ejaaz:
GPUs, specifically NVIDIA GPUs, are the most versatile asset out there.

Ejaaz:
You can use it for anything. You can use it for training, you can use it for

Ejaaz:
inference, and you can use it for any model, whether it's Claude,

Ejaaz:
whether it's GPT, whether it's Gemini, whatever. You can use it for it,

Ejaaz:
which means that it's a versatile customer base.

Ejaaz:
Earns a lot of money. And then Wall Street shot back at him and said,

Ejaaz:
well, hang on a second, these GPUs die after a couple of years.

Ejaaz:
And Jensen goes, that's not actually true. In fact, we have 10-year-old GPUs

Ejaaz:
that are being re-signed for another 10 years right now today at a higher price

Ejaaz:
than they sold earlier on.

Ejaaz:
So basically what he's pitched them is this is a new asset class and it can earn a ton of money.

Ejaaz:
And so Wall Street looked at this, some of the biggest financial powerhouses

Ejaaz:
in the world and thought, you know what?

Ejaaz:
He might be right. This is an asset class that can be likened to property or

Ejaaz:
railroads back in the day.

Ejaaz:
And that's why Larry Fink is comparing it to the 2008 mortgage-backed securities.

Ejaaz:
Now, if you're wondering, okay, well, this is like a financial crisis type thing,

Ejaaz:
you might be right, except there was like a few different things going on there,

Ejaaz:
which we'll unpack later in the episode.

Josh:
Yeah, it feels very much like AI compute is equivalent to revenue.

Josh:
And these are very now like durable, appreciating assets that can yield value over time.

Josh:
So when you think of like a bond, per se, the value of a bond is implied to

Josh:
go down over time, the underlying asset like the US dollar due to inflation,

Josh:
but the yield it will come up with is going to outpace that and then some hopefully

Josh:
the construction of a GPU is that not only do you get a yield in terms of the

Josh:
value creation off the back of token generation.

Josh:
But you now also have an asset that is likely going to appreciate.

Josh:
And even in the face that it doesn't, Jensen is giving plunge protection.

Josh:
So 25%. Yes. Depreciation insurance of up to 25% to help the banks get these

Josh:
marginal deals over time. So banks initially were concerned.

Josh:
They don't want to fund this because they don't want

Josh:
nvidia to come out with the new gpu that's a thousand times better than this

Josh:
one and it's going to knock all the margin out from underneath them

Josh:
who knows the roadmap about nvidia gpus better than anyone else it's jensen

Josh:
the guy who's building it so baked into this contract is the idea that

Josh:
jensen will ensure you he will make sure that hey your gpus are not going to

Josh:
fall drastically over time everything is going to be smooth and predictable

Josh:
and we'll work together to fund these companies that don't have

Josh:
the ability to do so so this comes in the form of like this long-term

Josh:
Debt and asset backed structures.

Josh:
And you think of it like if you're not a hyperscaler, if you're not Google who

Josh:
has a couple hundred billion dollars to spend off your balance sheet this year,

Josh:
but you still want to compete in the world of AI, you still need GPUs, these are who you go to.

Josh:
And they will offer you GPUs in exchange for interest on these GPUs.

Josh:
And in the worst case that it doesn't work out, they can just claw back the GPUs.

Josh:
NVIDIA can claw back those GPUs and turn it into their own NeoCloud,

Josh:
give it to another neocloud but the idea is that these assets are valuable they're increasing in money

Josh:
they are transferable in an easy way that you can just take the gpu and plug

Josh:
it in somewhere else or give the actual data center control over to someone else

Josh:
and it's a really lucrative kind of bizarre thing it's like okay if you're a

Josh:
bank 500 billion dollars

Josh:
you have insurance you are now able to allocate this

Josh:
incredibly valuable capital resource to anybody who you want and collect a pretty

Josh:
high rate of return on top of that

Josh:
and nvidia wants in on this too so initially it wasn't for nvidia nvidia now

Josh:
is given the option to backstop up to 25 of each opportunity so that is 125

Josh:
billion dollars at the ceiling

Josh:
and basically now nvidia and co the avengers get to roam around choose who they

Josh:
would like to give these gpus to and wrap it up in a really interesting financial

Josh:
product that they can go off and, I guess, monetize.

Ejaaz:
And I want to stress that this is only for NVIDIA GPUs specifically.

Ejaaz:
Jensen brokered this deal for his company only.

Ejaaz:
And he has a reason to do that because GPUs for the longest time has been very

Ejaaz:
broad based. If you look at some of the GPUs that Google makes or that Meta

Ejaaz:
is making or that even OpenAIR and Anthropic are reportedly working on their own specialized chips.

Ejaaz:
These are exactly what I just said. They're specialized. They can't be used

Ejaaz:
for many other models. It's only used specifically for their things.

Ejaaz:
So it's a much more niche case to create a type of loan or credit-backed security for.

Ejaaz:
Jensen has the opposite issue, which is like, it's too broad.

Ejaaz:
But that makes it an amazing financial asset. So in effect, NVIDIA is sort of becoming a bank.

Ejaaz:
And I wouldn't be surprised if Jensen starts to make a lot of money from this.

Ejaaz:
Over the last couple of weeks, something that he's also started doing is backstopping

Ejaaz:
specific Frontier AI labs and saying, hey, don't worry, I got you.

Ejaaz:
I'll front up the money that you need to purchase my GPUs.

Ejaaz:
And in return, whatever money you make on the products that you're building,

Ejaaz:
you can give me a revenue split from that.

Ejaaz:
I think he signed like a reportedly 10% revenue split from Safe Super Intelligence,

Ejaaz:
which is Ilya Sutskiver's new lab for their breakthrough that they're launching pretty soon.

Ejaaz:
And I think he's gonna do the same for a lot of neoclouts like CoreWeave,

Ejaaz:
Nebius and such like that, that are reporting crazy earnings.

Ejaaz:
I think this morning, CoreWeave reported 464% increase in revenue year upon

Ejaaz:
year, which is just a precursor to like the insane demand that they're seeing right there.

Ejaaz:
But then a question that comes into mind is, where on earth is this money coming from?

Ejaaz:
And on the screen here, it's like the main claimants are pension funds, right?

Ejaaz:
So pension funds who have amassed a large amount of wealth and typically don't

Ejaaz:
invest in high volatile type assets.

Ejaaz:
They kind of stick to real estate, very low interest types of things,

Ejaaz:
are the ones that are going to be backing a lot of this new GPU asset class.

Ejaaz:
And you have the biggest, most

Ejaaz:
powerful financial people in the world that are kind of pushing this on.

Ejaaz:
And so I'm thinking, is this reckless behavior?

Ejaaz:
Well, if you take the word of Goldman Sachs CEO David Solomon,

Ejaaz:
he goes, $500 billion sounds like a lot, but there are $9 trillion in US money

Ejaaz:
market funds and more than $100 trillion in US equities.

Ejaaz:
He has a deep belief in this opportunity and Goldman brings its extraordinary

Ejaaz:
distribution network. Larry Fink, CEO of BlackRock, also says,

Ejaaz:
he said it's a very attractive opportunity with long-dated, long-term returns.

Ejaaz:
They will be talking to pension funds. So it seems like the two most powerful

Ejaaz:
financial connoisseurs in the world are convinced that this new asset class

Ejaaz:
is a very real thing, which means that they've probably looked at the balance sheets.

Ejaaz:
They've probably looked at the revenue demand that a lot of these frontier labs

Ejaaz:
that are meant to be purchasing these things are going to do.

Ejaaz:
And they're looking at it and they're saying, this is an obvious no-brainer.

Ejaaz:
Now, if you're listening to me and you're thinking, dude, this happened with

Ejaaz:
railroads and it didn't work out.

Ejaaz:
This happened with the housing environment, mortgage-backed debt securities

Ejaaz:
in 2008, that didn't work out.

Ejaaz:
I have to say, it's a very different story on our end.

Josh:
Well, that's what I was going to ask you. I was going to say like,

Josh:
hey, obviously they're going to come out and say these things.

Josh:
I mean, we've seen them manipulate markets for a long time. We just saw what

Josh:
Citadel did to Leopold. It's like everyone is very clearly out in their own

Josh:
best interest. So if we look at this deal, okay, they're not going to say it's

Josh:
anything less than exceptional.

Josh:
So how do we kind of vet this? How do we fit this into a specific piece of context

Josh:
that I guess we could reference?

Josh:
And there's an interesting example of like aircraft finance versus mortgage

Josh:
finance, because this is something that has happened in the past where

Josh:
when you have an expensive standardized asset that's transferable between operators

Josh:
and has a lot of demand for these secondary markets, it creates this interesting marketplace that

Josh:
I think is much more comparable to aircrafts than mortgages.

Josh:
And I'll explain. So like a Boeing 747 or 737 or whatever, that's been built

Josh:
like 20 years ago. And trust me, you've flown on these. The airlines kind of

Josh:
suck. You're flying in some old planes.

Josh:
That is just as valuable today

Josh:
As it was 20 years ago, because they're able to derive so much value from it.

Josh:
It does the same exact job.

Josh:
The same plane that was built today is doing the same job that was built 20

Josh:
years ago. And sure, perhaps you would prefer to fly on the newer plane.

Josh:
But the reality is, is that tickets are sold out on the 2005 plane and the 2025 plane.

Josh:
And when you think of GPUs, they exhibit a lot of the same traits and characteristics

Josh:
as an airplane, where it's expensive, standardized, it's transferable.

Josh:
It has a lot of liquidity in secondary markets.

Josh:
And I think this is an interesting way of looking at it relative to mortgage

Josh:
finances, which is where we got in trouble. And this isn't the first time this has happened before.

Josh:
There is something that has been done similar to this with Broadcom,

Josh:
where like that Google Anthropic structure actually runs through this thing

Josh:
called an SPV, a special purpose vehicle that buys TPUs and leases them with

Josh:
Broadcom providing the residual value guarantees,

Josh:
and then Apollo and Blackstone supplying the private credit to fund all of this.

Josh:
So people have experimented with these structures, before. It has worked.

Josh:
We haven't seen it at this scale. I mean, the alarm bells are partially ringing.

Josh:
I'm like, just out of instinct, like out of an e-joke reaction,

Josh:
like, oh, wow, this is a lot of money. This is a lot of powerful people who

Josh:
can very much control and sway the way the market moves.

Josh:
So far, it seems like a pretty reasonable thing. It's like, hey, we need GPUs.

Josh:
GPUs are transferable. They're kind of like they're fungible,

Josh:
I guess. I'm like thinking of the word. I'm like, well, this feels kind of crypto adjacent.

Josh:
There's like these fungible assets that can be transferred that are valuable.

Josh:
So I don't know. There's a chance this goes over. OK.

Ejaaz:
And it's important to not extrapolate too far into the future,

Ejaaz:
like what we can feasibly attain from the data, which, by the way,

Ejaaz:
is publicly available. If you're listening to this and you don't believe anything that we're saying.

Ejaaz:
Maybe we should actually link to a bunch of sources. Maybe we'll link this artifact

Ejaaz:
that you're seeing on the screen right now.

Ejaaz:
The data is all available through quarterly earnings of every single company

Ejaaz:
that is leading at every layer of the AI stack. So you can see the data,

Ejaaz:
digest it yourself and figure it out for yourself.

Ejaaz:
But what I will say is when you look at the 2008 financial crisis,

Ejaaz:
when you look at the railroad crisis, when you look at the telecom crisis back

Ejaaz:
then there was a huge amount of oversupply.

Ejaaz:
Which didn't have the back demand that it stated it had so 2008 people assumed

Ejaaz:
that property prices would just keep going up and at some point

Ejaaz:
that got two head over heels if you look at the railroad they built too much

Ejaaz:
if you look at the telecom they built too many cables right in this case

Ejaaz:
we're constrained by a few things number one physically it takes so many different

Ejaaz:
substrate layers to build a gpu.

Ejaaz:
Every single layer right now in the world of physical atoms is incredibly constrained.

Ejaaz:
There's not enough. The GPU demand is overweight, the actual supply that is available.

Ejaaz:
Number two, the AI demand, which is driving GPU demand, is accelerating much

Ejaaz:
faster than the supply itself can.

Ejaaz:
So if you look at memory as a basis for this, they can increase capacity.

Ejaaz:
This is the top docs, top memory manufacturers can increase capacity around 20% per year.

Ejaaz:
But demand is compounding at 45 to 60% per year.

Ejaaz:
So if you do the math, if that continues, you're going to be in a constrained

Ejaaz:
supply state for like at least until 2028 or 2029 until some of these other

Ejaaz:
chip fabs can get increased. So we're still kind of in a holding period, right?

Ejaaz:
Now, if you look at the backlog for some of these companies.

Ejaaz:
You might be like, well, customers don't want AI as much as these guys are making

Ejaaz:
it out to be. They're just kind of like pushing their bags.

Ejaaz:
Well, look at Google's backlog. It doubled this year in a matter of months.

Ejaaz:
It is now at $460 billion, and we're at the halfway mark of this year.

Ejaaz:
It's probably going to increase even more. This is the case across Microsoft

Ejaaz:
and a bunch of other hyperscalers as well.

Ejaaz:
Then if you look at the GPU rental prices, my favorite thing,

Ejaaz:
which kind of came out this morning, Josh, or maybe yesterday,

Ejaaz:
CallWeave had their earnings, and they said, we recently signed an A100 contract that extends into 2029.

Ejaaz:
For those of you who don't know, and A100 is an NVIDIA GPU that was created in 2020.

Ejaaz:
And its lifecycle prediction back then was three and a half years.

Ejaaz:
Now they're predicting that it's actually going to be good to go until 2029.

Ejaaz:
That's because it's not just being used for bleeding edge training.

Ejaaz:
It's being used for inference and a bunch of other stuff. So the point is,

Ejaaz:
these GPUs are very versatile, NVIDIA specifically, and that's why they raised a crap ton of money.

Josh:
Yeah, and more valuable over time. It's this really bizarre thing in which the

Josh:
useful lifecycle of a hardware object is increasing instead of depreciating for the first time.

Josh:
And we've never really seen this phenomenon at scale before.

Josh:
But like you mentioned, there's not many reasons in which it's going to slow

Josh:
down in the near term future. When I look at this, I'm kind of looking at it

Josh:
like around the corner. And then you can't really see around the next corner,

Josh:
but we have an idea of the first corner. And that first corner is sold out supply

Josh:
for at least 2027, likely 2028.

Josh:
And then by the end of the decade, 2029, 2020, 2030, we start to run into larger

Josh:
constraints, mostly around energy and power.

Josh:
And we start to like run into resource constraints that we don't quite have now.

Ejaaz:
So are you basically saying there's like multiple corners, Josh?

Ejaaz:
Like I'm curious, like, for the GPU specifically, do you think it's like six

Ejaaz:
months? Do you think it's also like 12 months? Like, what's your guess if you had to...

Josh:
Well, there's a somewhat clear trajectory for the next 24 months,

Josh:
like 18 to 24 months in terms of...

Josh:
It seems fairly predictable where we know how many, like the lithography machines,

Josh:
we know how many chips they can create.

Josh:
Then we know how many of those chips can be packaged into usable chips.

Josh:
And then we know roughly how many data centers can be built that can actually

Josh:
turn those chips on and power them.

Josh:
And you can somewhat project that out up to 24 months loosely,

Josh:
very loosely, because there is only so much throughput for these machines.

Josh:
So if you assume everyone's operating at full capacity, you can kind of work

Josh:
those numbers backwards and understand like, OK, they're sold out and they're

Josh:
still not going to be enough to satisfy the demand, assuming the demand continues,

Josh:
which there is no signs of slowing down.

Josh:
All these use cases for AI, particularly around agentic AI, require a tremendous

Josh:
amount of inference and even including efficiency upgrades to the software,

Josh:
something similar to what we imagine SSI is working on.

Josh:
There's still a huge amount of demand that will fulfill that this is jevin's

Josh:
paradox which we really should name jensen's paradox because that seems to be

Josh:
a little more accurate in terms of how this is working

Josh:
but we can kind of project out till then and we know all right gpus fully sold

Josh:
out fully constrained after that things get a little

Josh:
more hairy right it's because you have to assume by that time we'll have something

Josh:
similar to agi asi self-recursive improvements we should be getting a lot of

Josh:
innovation breakthroughs around efficiency and software and we don't really know what

Josh:
the power market's going to look like. We're not sure if we're able to make enough

Josh:
Energy to satisfy the demand of the gpu centers that are being projected out

Josh:
into 20 to 30 so it seems like this is a very long duration thing that's going

Josh:
to need to play out but in the short term in the intermediary term it seems like i mean

Josh:
i'd like to find the steel man against this because this seems important and

Josh:
we should talk about like what are the possible ways in which this breaks

Josh:
but just looking at demand of inference and our capability of serving inference

Josh:
and there is a huge mismatch in the case of inference demand that's just not

Josh:
going to be met for a really long time.

Josh:
So those H100s from 2020, or the A100s, I should say, from 2020,

Josh:
are still going to be useful in 2027, 2028.

Josh:
And that is particularly valuable when you're investing in GPUs at this scale.

Ejaaz:
Well, actually, now that you say it, a lot of that inference demand,

Ejaaz:
at least in the next six months or so, is going to come from AI agents.

Ejaaz:
I'm in no doubt about that.

Josh:
You said the word agents. It's funny you should mention agents because we have

Josh:
something to say about agents from our sponsor of this episode, Ledger.

Josh:
If you're building with AI agents, you are probably worried about security and

Josh:
rightfully so because as we've seen recently, these agents have been kind of

Josh:
doing some funky things.

Josh:
So Ledger has this three-step approach to solving this. The first is that the

Josh:
agent proposes a change, then the human approves the change,

Josh:
and then the Ledger signer actually enforces this change.

Josh:
There's a three-step process to make sure your agent doesn't do anything you

Josh:
do not want it to do they have this thing called the ledger agent stack which

Josh:
is an open source software stack that gives you a series of tools to help you navigate

Josh:
your journey with your agents it works with cloud code it works with codex cursor anywhere that you use

Josh:
your ai models it's available now it is open source you can find the link in

Josh:
the description down below and thank you so much to ledger for sponsoring this episode the agent

Josh:
barrage? I guess we have to talk about the downside effects,

Josh:
right? What could happen if things don't go as well as planned?

Josh:
And where does that risk actually live? So, Ijaz, it seems like this has been

Josh:
fully prepared, lovely by our clawed artifact right here. So,

Josh:
what is actually the downside risk? What do we need to look out for when we're

Josh:
evaluating how to invest around this?

Ejaaz:
Okay. So, we are the Limitless Show. And as anyone who's listened to us for

Ejaaz:
a while knows, we are grounded or we are trying to ground ourselves a lot more from the bullish case.

Josh:
I mean, if you got to say it. Yeah.

Ejaaz:
So there are a few ways where this can obviously go wrong.

Ejaaz:
And I want to kind of like walk through some of these and get your take on this,

Ejaaz:
Josh. So number one, the thing that's like blaring to me is this is all based

Ejaaz:
on the fact that AI demand not only is sustained.

Ejaaz:
So you have paying customers to buy cloud subscriptions, GPT subscriptions,

Ejaaz:
companies paying tens to hundreds of millions of dollars a year for API access. But...

Ejaaz:
That it increases. Right now, it's increasing at a crazy rate.

Ejaaz:
We see all these quarterly earnings, revenues compounded between 100% to 500% year upon year.

Ejaaz:
It is insane, but that's not sustainable. It's not going to keep doing that.

Ejaaz:
It'll presumably eventually plateau.

Ejaaz:
So if that does plateau, or in worst case, if that plummets,

Ejaaz:
then these companies are going to need fewer GPUs, which means that Jensen's

Ejaaz:
$500 billion, these debt-backed securities are going to be in less demand.

Ejaaz:
And that's where you might see a default.

Ejaaz:
The second major thing here is that the GPUs themselves depreciate a lot faster.

Ejaaz:
And that has been the Michael Burry, the guy that did the famous big short back in 2008.

Ejaaz:
That's been his view this entire time. He says that the upgrade cycle for a

Ejaaz:
lot of these NVIDIA GPUs are actually a lot shorter than what Jensen NVIDIA claims.

Ejaaz:
However, in practicality, this seems to not be the case.

Ejaaz:
However, NVIDIA is now releasing a lot of GPUs at a much more higher frequency

Ejaaz:
rate, which means that they're going to replace more of the GPUs in the prior

Ejaaz:
market, in the prior cycle, and they'll start flooding the market.

Ejaaz:
My countess of that is simply you can't make GPUs that quickly.

Ejaaz:
It takes a lot of technical expertise, and it is limited by the likes of TSMC

Ejaaz:
and wafer capacity and a bunch of other technical stuff, which I don't want to get into on this show.

Ejaaz:
So I'm struggling to actually see how these two factors might actually be triggered.

Ejaaz:
But I don't know if you have a different opinion, Josh.

Josh:
Yeah, the thing that I'm looking out for most is the return on invested capital

Josh:
from the large hyperscalers.

Josh:
It feels like they just run the world. They're spending all the capex.

Josh:
They are basically floating the entire economy right now.

Josh:
And if the returns on that investment start to go down, for example,

Josh:
that seems like a very scary thing. So looking at Google's earnings reports,

Josh:
we see like, okay, They have $514 billion dollars.

Josh:
They're spending. Can they keep returning revenue on that on schedule?

Josh:
If the answer is yes, if there's still revenue to be made on AI spend, that is amazing.

Josh:
In the case that that turns and we start seeing earnings reports from companies

Josh:
who are spending huge amounts of capex saying, our margins are actually shrinking.

Josh:
Our revenue is not coming at the multiple that we expected.

Josh:
That seems to be a red flag because that will slow down spending significantly across the board.

Josh:
Basically, we want to make sure that all this stays profitable.

Josh:
So we want to make sure that Inference demand is actually continuing.

Josh:
Companies are actually able to meaningfully monetize this.

Josh:
The enterprises that are spending billions, hundreds of billions of dollars

Josh:
a year on AI spend, we need to make sure that they're actually getting value.

Josh:
Otherwise, they're going to cut those contracts. That is probably the most important thing.

Josh:
The second is just monitoring the rental rates. Like currently,

Josh:
a lot of companies are terrified to re-sign their long-term GP rental deals

Josh:
because the price that they're going to get them at this time around is going

Josh:
to be double the price that they got originally.

Josh:
That is a phenomenon that like no one was really expecting, but here we are.

Josh:
And if that trend continues as well that's something i'm kind of looking out

Josh:
for so i'm looking at what is the hour double

Ejaaz:
Like three years from now right i mean it.

Josh:
Might it might but i'm saying this this is just something that like you should

Josh:
keep a close eye on you know how long these

Ejaaz:
Contracts are for josh that they're signing i know.

Josh:
They vary quite a bit like some are short they're like a year some are longer

Josh:
out to like three years maybe um but i they're they're variable and i know that

Josh:
when the time has is coming to kind of re-sign this contract

Josh:
the price is higher not lower for the same supply so ensuring that this

Josh:
continues this trend continues and even if it doesn't continue making sure it

Josh:
doesn't flip negative because that might change things and granted nvidia has

Josh:
your 25 plunge protection service you're available but you don't really want to put that 125

Ejaaz:
Billion dollars by the way for those of you trying to do the math.

Josh:
That's a lot of money man a lot of money um and then third is just like

Josh:
what the yields actually is from these gpus like how much the real yield yeah

Josh:
the actual real yield and

Josh:
these are also like those mous this is an assigned deal and we've had something

Josh:
similar to this before remember that crazy project back in the day called project stargate where elon and

Josh:
masa or not sorry not elon sam altman and masayoshi son and even donald trump

Josh:
all stood in an office together and they said

Josh:
we're going to spend x billion dollars on this data center build out

Josh:
it hasn't really happened as planned so this is not a contractual obligation

Josh:
to spend 500 billion dollars this is a hey dude

Josh:
We're all rich. We'll commit to like $500 billion and we'll see how it goes.

Josh:
And that's kind of what they have. It's a handshake deal to build this new financial

Josh:
economic instrument around the GPU, particularly as it relates to NVIDIA.

Josh:
So huge win for NVIDIA, probably a large win for a lot of the companies that

Josh:
are not able to afford this, and probably a huge win for the banks.

Josh:
At the end of the day, they seem to always win. And that's kind of what the deal is here.

Ejaaz:
I am really struggling to think about a world, an alternative scenario,

Ejaaz:
where AI doesn't require GPUs, specifically the monopolistic GPUs from NVIDIA.

Ejaaz:
They just have such a stronghold on the entire market.

Ejaaz:
And even if you have some kind of novel LLM architecture that gets created in

Ejaaz:
the future that completely disrupts the current paradigm, you're still going

Ejaaz:
to need hardware to run these things.

Ejaaz:
And that hardware is very much GPUs that are being designed and created by Jensen Huang.

Ejaaz:
So however way I skin this cat, I still think that you're going to need these GPUs.

Ejaaz:
You still need token generation. Gavin Baker has made this point across so many

Ejaaz:
other podcast episodes in the last two weeks that it's ingrained in my head at this point, right?

Ejaaz:
And then the other thing I think about is, okay, well, if NVIDIA becomes a bank

Ejaaz:
themselves and they start taking revenue splits from all these frontier AI labs,

Ejaaz:
that's a completely new revenue line for NVIDIA.

Ejaaz:
So when I think about this with my investing hat on, I'm thinking.

Ejaaz:
Okay, not only is NVIDIA supplying the foundational element that is required

Ejaaz:
to run and inference these GPUs, train these GPUs, but they're also

Ejaaz:
being the ones that are driving cost down per token, right? So like they're

Ejaaz:
doing this with their CUDA software mode.

Ejaaz:
And then I think about the financing side of things. So they're being the financiers

Ejaaz:
of this entire thing as well.

Ejaaz:
Now, that does sound like a house of cards if the demand wavers, if the demand plummets.

Ejaaz:
And I can easily see the market being very volatile and reacting to any kind

Ejaaz:
of headline like they did to this initial headline but i don't know it just

Ejaaz:
seems very bullish to me on nvidia at least and yeah i don't really know how

Ejaaz:
to think about it yeah yeah.

Josh:
And i mean like in this case like i do kind of lean on the opinions of people

Josh:
who are more in the know than me yeah someone like elon who is now exclusively

Josh:
committed to purchasing only nvidia gpus for the new data center build out

Josh:
and they are effectively the best data center builders in the world so i you

Josh:
have to like have a little bit of trust in the opinions of the

Josh:
true experts who are in the arena doing things when i look at that and i see

Josh:
like they exclusively want nvidia and they are building the best fastest most

Josh:
efficient data centers i'm like okay that's pretty good signal like micro hard

Josh:
the new data center that the spacex ai team is working on

Josh:
is i think like a third the footprint of macro

Ejaaz:
Hard right no micro hard.

Josh:
No no there is there's micro hard is

Ejaaz:
A micro hot.

Josh:
Yes and micro hard is about a third of the footprint i might be getting this

Josh:
wrong half or a third of the footprint of macro hard easy but it contains the

Josh:
same cluster of 200 000 gpus wow they just figured out how to do it much more

Josh:
efficiently and much more dense so

Josh:
there's a huge amount of innovation clearly they know things that the rest of

Josh:
the industry does not and when they come out the biggest purchaser right

Josh:
yeah and they're committed exclusively to the nvidia gpu wow and when vera rubin

Josh:
comes out at scale man oh my god i keep saying this for like holy smokes

Ejaaz:
Those models are gonna be insane those models are gonna be absolutely insane.

Josh:
So buckle up good time to be good time to be nvidia good time to be a GPU.

Josh:
Yeah, that's the update. So is this a house of cards?

Josh:
Is it all going to come tumbling down? Is this financial innovation in a new

Josh:
era of the United States of GPUs?

Josh:
Let us know in the comments down below if you enjoyed this episode.

Josh:
Don't forget to share it with a friend who might also enjoy this episode.

Josh:
Ejaz and I, cool story. We were walking down the street last night after dinner

Josh:
and two people walked up to us.

Josh:
They were like, hey, you guys are those podcast guys. You host the show.

Josh:
And we were like, yeah, we do. Like, cool. Thanks for watching.

Josh:
So that's like so cool whenever that happens. And that is because of you sharing

Josh:
with your friends, letting everyone else know that the show exists.

Josh:
And if you enjoyed it, you know, don't forget. Thumbs up. You could subscribe.

Josh:
You could give us a five-star review on your favorite podcast platform.

Josh:
Any parting thoughts, EJs?

Ejaaz:
Yeah. Homework for you guys. If you see us on the street, don't be shy.

Ejaaz:
Come and say hi. We want to meet you guys.

Josh:
Don't be shy. Say hi.

Ejaaz:
Yeah. It's lovely to meet you guys. We've now met people. We've now met listeners,

Ejaaz:
in New York and we've met them in San Francisco.

Ejaaz:
I have people calling in from Europe to my family talking about these random

Ejaaz:
guys that yap about AI. Turns out it's us. Like it's really cool to see the

Ejaaz:
momentum that we're getting here.

Ejaaz:
And it's all thanks to you guys. So if you're one of these people that care

Ejaaz:
passionately about what we talk about and tune in every day, thank you so much.

Ejaaz:
And if you're not, turn on notifications, please. Subscribe to us.

Ejaaz:
We bring the best news, hot, fresh out the oven, every single day,

Ejaaz:
or rather four times a week.

Ejaaz:
And we would love to hear from you. Leave us a comment, DM us on X.

Ejaaz:
And yeah, I think that's it. Thank you so much for listening.

Josh:
See you next time.