December 8, 2025 • 36 mins
In this M&A edition of Market Maker, Anthony and Piers break down Medline’s potential $5 billion IPO, the biggest of the year, and what it signals for private equity exits and the 2026 deals landscape.
They also explore Amazon’s bold entry into the AI chip race with Tranium 3, how it stacks up against Nvidia and Google, and what it means for cloud dominance. Plus, OpenAI hits the panic button as Google’s Gemini and Anthropic’s Claude rapidly gain ground.
(00:00) Medline’s $5B IPO
(08:52) ECM Process
(12:45) Bank Rankings Shuffle
(20:06) Amazon’s AI Chip
(28:09) Cloud Wars Heat Up
(32:04) OpenAI’s Code Red
*****
Final chance to join our in-person Finance Bootcamp happening in London 15-19 December 2025. Use code BOOT500 at checkout to access a £500 discount as a loyal podcast fan!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Hello and welcome back to the Market Maker podcast and welcome
back to the M and A show, which is where we do a bit of an
overview of some of the major deals of the last week and talk
about some other interesting headlines related to the field
of corporate finance. So I'm going to have to
apologise to our researcher Saab, who always does such a
(00:21):
thorough job, but I'm going to completely skip over his deals
of the week just just because A,there's not too much sexy there,
but because what is sexy is one of the biggest lined up IPOs of
the year coming just in the nickof time.
So going to talk about that for a little bit.
(00:43):
Really interesting developments then of what we've seen in the
league tables. We might well have for any of
our long time listeners a special guest or one and only
Stephen Barnett might be coming back for an outlook on the M&A
scene for 2026. So that is coming folks.
(01:04):
Let's get the red carpet out. Get the red carpet out for
Stephen. But we'll have a little touch
here because this latest IPO, we're going to talk about
Medline is of such an order of magnitude that it is going to
shift around the league tables alittle bit in terms of the
sectors that have been driving the ECM part of the investment
banking scene. And then we might talk about AI,
(01:28):
but in the context of everyone'sbeen talking about how Google
has kind of thrown the cat amongst the pigeons in a sense
of Gemini and what's come out. And that's put the red alert on
as a status at open AI, but alsothe chips.
What we were talking about with with Google recently and how
that's impacted NVIDIA last week, Amazon have come forward
(01:52):
with their latest 1. So what's Amazon got to offer
and should NVIDIA be worried? So, yeah, perhaps we can start
with this, this biggest IPO dealof the year, Pierce.
Yeah, Medline, here we come. This could be an absolute
monster where they are, you know, looking to raise $5
(02:15):
billion. I mean, look, just to just to
get the timings right here. It hasn't happened yet, but it's
probably looking looking on for early part of next year
probably. Hopefully January maybe.
But this is an interesting case for a couple of reasons.
Firstly, they've been on the docket for an IPO for a while.
(02:37):
They were the one of the big kind of names on the list that
decided to delay their IPO because of Trump's tariff
policy. And and look, partly that's
because of the line of business that they're in.
So that you know, who are Medline, They're actually a
privately well, they were a family run business run out of
(02:57):
Chicago. We'll talk about the PE deal
they did in 2021 in a minute, but family business business out
of Chicago, they they distributemedical supplies to like
hospitals and clinics. And so we're talking surgical
equipment, you know, gloves, gowns, you know all the rest of
it, right? And they supply that that
(03:18):
globally. So actually they've got 43,000
employees, they operate over 100countries, sales around about
$23 billion, right? So from a tariff perspective,
obviously you've got a fully global business here and it's
about obviously shipping equipment, you know, all around
the planet. And if tariffs were going to
(03:39):
come in and really put the the glue in, in the kind of movement
of goods globally, then that wasgoing to be a really big issue
for them. So they kind of just stepped
back and said, all right, their PE owners, I should say, stepped
back and said, OK, we're not going to IPO now.
We're going to wait. Originally they'd wanted to go
early 2025, but they kind of changed their mind.
(04:03):
So, you know, this is the business, you know, why do they
want to IPO? Because well, it's a private
equity cycle, right? the PE boys came in.
In fact, they're the biggest privately owned US distributor
of medical products. And so they did APE deal back in
2021 and it was acquired for $34billion.
(04:25):
OK, So it was a leveraged buyoutand you had the big guns like
Blackstone, Carlisle, Hellman and Friedman.
I think you had the sovereign wealth funds getting involved
from the Middle East as well, right?
Re obviously 34 billion dollar deal, that's that's a huge check
size. So you had a consortium of PE
(04:45):
firms coming together to, to kind of kind of do that.
But of course private equity companies, their strategy is to
buy, then increase the value of that asset, then sell.
OK. It's very much on a cycle.
The PE world often work on this kind of three to five year idea
where it takes them three to five years to a pay down some of
(05:06):
the debt. So remember, normally this is a
leveraged buy out. So the PE firms will use debt as
well as a little bit of cash to buy the company.
They'll then run the company andin the meantime pay down that
debt so that then when they sellit at the end of the cycle, you
know they've got more of an equity portion and it and it
(05:28):
kind of leverages their returns.OK.
Also at the same time, it's not just about buying the company,
paying down the debt and then selling it.
They can they improve the business, can they grow the
business or can they improve margins and make it more
profitable. And so actually this has been
executed really well. So we've seen the profitability
(05:49):
and the margins of Medline improving through this kind of
PE cycle. So it kind of made sense that it
was ready to go. The problem that the PE industry
have had though is that the IPO market has been closed, right?
Deals have just not been coming because the investor appetite
(06:09):
just hasn't been there. Why interest rates?
First of all, interest rates have been really high.
And look, if you want to do sellbig assets like this, who are
you selling to? Well, investors will often need
to leverage up and take on some debt to finance that purchase.
But if debts really expensive because interest rates are high,
(06:29):
it just puts a whole puts on puts the ice all over that kind
of setup, right. So private equity firms have
been incredibly frustrated and that they haven't been able to
exit through the IPO channel a lot of these investments, which
is bad. It's a bad look for the
(06:50):
investors from the private equity firms.
Investors, you know, they want payouts, they want returns, they
want liquidation events, right? And these PE firms haven't been
able to do it. So this would be an awesome deal
to get over the line, A, becauseof its size, it's absolute
monster, but B, it would be a great kind of signal for these,
(07:13):
particularly the USPE space, to see a really big liquidation
event and see these PE firms being able to return capital
back to their investors. So yeah, it's it's certainly a
headline grabbing moment here. And then I'm assuming that
because of the size of this deal, hence the reason why
(07:37):
there's so many investment bankson the books.
So lead book runners here I've got listed are Goldman Sachs,
Goldman Sachs, Morgan Stanley, Bank of America, JP Morgan, so
all of the big boys. I mean, that's, I mean we'll
look in a second, but I'm assuming that's like 4 of the
(07:58):
top five. Right now, yeah, you're missing
city. It's fourth in the top five.
Yeah. And we can dive into this
actually maybe this is a good segue to talk about, you know,
the the divisional breakdown of when we when we talk about
investment banking, obviously the mine tends to gravitate
(08:19):
towards M&A fees, but ECMDCM andwho are the the leading ones
there and when these bigger firms naturally have much bigger
franchises when they have the ability to take on that type of
business for sure. So the other thing is it's going
to be on the NASDAQ. So in terms of the exchange, so
(08:41):
the NASDAQ Global Select Market under the ticket MDLN.
So, yeah, any other points on that before we move on?
Well, I mean, it probably may beworth just reminding people of
the process here, why it's suddenly coming to the news,
because so the, the process is that first you've got to file a
prospectus with the regulators. And that's what's just happened,
(09:02):
right? And so it's like, wow, OK,
that's your signal that right, these PE firms and Medline, OK,
they're back at the table here looking to get a kind of process
done. You file the prospectus with the
regulators, they check it, they verify that the information in
it is valid and true. Then it's like go time.
Well, what does that mean? Well, it means these banks then
(09:25):
run Well, they they kind of run a roadshow right where they
basically take Medlines kind of management team and they take
them on a road show and sit themin front of, you know, big
investor communities. And it's about telling the
story, you know, why is this an investable opportunity?
You know, talking about your hockey stick charts with amazing
(09:47):
future growth and all the rest of it, getting them excited.
OK. Then hopefully from that process
you'll also get what we call some cornerstone investors,
which will be, you know, hopefully like a small handful
of big, you know, notable investors that are saying yes,
we're in. We want to be a kind of
(10:08):
cornerstone investor. This is great for a process
because if you've got a big nameon the ticket, then obviously
that's reassuring for all the other investors in the community
and perhaps then kind of greasesthe wheels for other investors
to get on board, right? So you might want to try and
find some cornerstone investors,but then you do a book building
(10:29):
process after the roadshow, which is like, hey, right, you
lot, you've heard the story, right?
Who's interested, you know, showus your interest.
And then from so you basically get these kind of bids coming
in, which is then enables and that's why you need these
multiple banks on the ticket if you're trying to sell something
that's going to be worth we think roughly $50 billion.
(10:52):
So that'll be a great flip. They bought it for 32 billion in
2021. They're looking to flip it for
50 here, and obviously the amount of debt in that business
will have reduced as well. So this is going to be a great
return for the PE firms on such a massive scale, right?
But 50 billions a lot. So you need multiple book
runners out there. So you need the big guns tapping
(11:12):
into their respective really large networks.
And so you do the book building,which gives you a gauge on
demand because ultimately you'vegot to price this thing, right?
Come IPO day, you're going to issue shares and sell them.
What's the share price? Well, these banks calculate the
share price from understanding the demand that they've been
(11:36):
able to gauge during the book building process.
They'll often give guidance, right?
So when they're out on the roadshow, they'll guide, right?
We're looking at a range and they'll give a share price range
and that's something that the investors can kind of have a
think about. And then when they put their
bids forward, they kind of placetheir bid somewhere in the
range. And if you're super positive on
(11:58):
this opportunity and you really want it, you're going to bid top
end. If you're lukewarm but
interested, maybe you'll go bottom end, right?
And then you build this book of demand, then they price it and
hopefully they price it correctly because then it's like
right, IPO day, let's now sell these shares to the investment
community that have shown interest at the price we've
(12:20):
decided and great the IP OS done.
That's where the PE firms exit. They sell their shares in the
business to then the other new investors coming in and then it
becomes publicly traded. So that's they've chosen the
NASDAQ and then anybody in the world will be able to trade
these shares in the secondary market on the NASDAQ thereafter.
(12:44):
Yeah, It seems like, seems like this is a significant signal
then on many fronts for deals in2026 because at around $5
billion, this IPO would dwarf the current US leader Venture
Global's $1.75 billion listing in Jan, in January.
Remember beginning of the year, we were all like this is the
(13:05):
year for M&A and then the complete opposite kind of
happened. But as you were just rightly
saying, I mean, I was just looking back on a post that I
did, this was back in four months ago and it was the FT ran
an article talking about privateequity firms flipping assets in
record numbers to themselves in using continuation funds to cash
(13:30):
out for investors. But is it just, is this a PE
bubble? We're recycling this debt as you
were describing. But as you said, they're going
to do very well out of this. And in combination with the
general macro climate, which is indicating at least at this
point that rates in the US are going to come down.
(13:52):
You're going to have a Fed shiftthat's probably going to be more
aligned and more favourable for those more dovish conditions as
well. So what looking quite good
again, or we were, we were hypothetically thinking this
time last year that Trump's going to come in and create that
environment. Whereas this year, that's now a
(14:14):
known thing and a lot of those uncertainties or indeed risks
over Trump have somewhat dissipated, particularly the big
trade debacle. So is this now actually a much
more known fertile setup for 2026 than it was actually
looking back retrospect for 2025?
(14:34):
I mean, yes, you know, without wishing to make the same mistake
twice, yes, it looks good for 2026, but the main thing is
interest rates for these deals. So as long as the Fed are dovish
in 2026 and who knows, does Hassett get the gig and all the
(14:55):
rest of it and do rates come down, then you're talking?
Because I think at the moment, like part of the story about the
IPO deal as they're, sorry, the IPO market being a great winner
from the Trump presidency, you know, half of it was about the
regulatory side, like removing red tape.
And Trump hasn't really got ontothat part of his agenda yet.
(15:16):
He's been, so his docket's been full with his tariff game,
right? So I think in 2026, we'll see a
little bit more of a focus on that from the Trump
administration. If you can couple that with a
rate cutting cycle continuing, that's a pretty, that's a pretty
interesting kind of kind of cocktail to feed through to what
(15:38):
is a huge pent. There's a massive queue of
companies want or PE firms particularly wanting to exit
investments right there. They are queuing up.
So it's all there. It's just ready to come and
maybe that the cork is going to get pulled out of that thing and
we might get a really interesting year of activity on
(16:01):
the IPO market in 2026. So maybe then to finish here, we
know this is going to be, if it goes ahead, the biggest IPO.
So where does that puts it in terms of the sectors and then
the rankings for these book runners?
Yeah, well, on from a sort of, well the top 10 IPOs so far this
(16:21):
year. Let's start there.
I'm actually just looking down the list.
The first the, the US, the firstUS entry on the list, this is
like a crazy impressive is actually 5th.
So there are four bigger IPOs than any you I mean the US is
always top of this league table always.
(16:42):
So they're they're coming in 5thand that was Venture Global and
the ones above that are coming out of APEC like like all of
them, right. So for the US this will put them
back on the top. This will this they will be the
biggest deal, but it's not goingto come in 2025.
So it's a bit academic that, butbut yeah, in terms of sectors as
(17:06):
usual, the technology sector is the largest player, right, As in
the volume of IPO deals, technology is leading the pack,
but healthcare which is currently in fourth.
So if you think about the sectors in 2025 so far,
healthcare are so technology is #1 industrials is number 2,
(17:30):
Financials 3, healthcare for materials five.
OK, in the order, this is deal volume, IPO deal volume by
sector in 2025. If they land this deal, then
that healthcare sector 1, I think that'll push them up 1
notch to 3rd. I think that'll get them up up
above the financials. But yeah, still technology is
(17:53):
out like 25 billion, whereas industrials is about 19 billion,
financials at 16 billion. Healthcare currently sat around
about 12 1/2 billion. So yeah, they're definitely not
going to get out to the technology.
They're not going to hit top spot.
But healthcare generally, as I said in the episode last week,
(18:15):
it's kind of it's been a really bright, bright spot of the
marketplace in the last one month when you've had a little
bit of a tech, big tech rolloverwith NVIDIA down whatever 14% in
the last one month, the healthcare stocks have really
been ramping. And a lot of this is due to the,
it's not just IPO, it's all the M&A activity that's going on in
(18:38):
that sector and now this monsterIPO.
And it's like, wow, OK, this sector's looking really hot and
this is feeding into some share price appreciation.
And if they get this Medline deal away if and it's successful
and if there's large demand and if it goes for top end of the
price range, if, if, if, then itcould really be a great
(19:02):
statement and a great measure ofthe market appetite right at the
start of 2026, which could really set the ball rolling for
the 12 month period ahead. Yeah, when you talk about that
pent up demand as just looking at the stats here, because you
said Apex like the outright kindof runaway story here, but the
(19:24):
US suffering, as you said earlier by the shutdown because
that paused new listings, right?And in the third quarter alone,
there were 64 IPOs in the US raised 15 points or $15.3
billion. So the shutdown when it did
happen a few weeks ago, happenedright at possibly the worst time
(19:45):
if you were doing IPOs. But meaning though that they
were they were running hot. You've had a gap.
Rates are now probably heading lower.
Yeah. So probably going to come back
pretty quickly if Medline gets away, as you said.
Really. Well, cool.
All right, Well, let's let's head on and talk about this
(20:07):
Amazon. Yeah, it's funny.
It's like, as we will go on to describe, Amazon have been
working on this stuff for a longtime.
However, it's funny that they'rethe news breaks just when they
feel like they've got to come tomarket and start talking this
stuff up in the context of we'vehad, you know, this is the third
(20:27):
element or the third leg, so to speak.
It's been NVIDIA, we're talking Google last week, now we're
talking Amazon. So what have Amazon got to bring
to the table? Amazon have got Tranium 3,
that's what they've got. Just slapped that onto the
table, which is their latest in the house, basically AI
(20:49):
accelerator chip. You know, it's a, it's direct
competition for Nvidia's GPUs. And you'll remember that Google,
like it's so interesting with Google and actually Google, I
mentioned the tech sector rolledover a bit.
NVIDIA down 14%. Google's up 13% in the last one
month because of the, the drop of Gemini.
(21:12):
You know, I forget the versions now.
Is it Gemini 3? And it being, wow, hang on, this
thing's amazing. Hang on, isn't this, this is
better than ChatGPT? And really key point, it was
trained on Google's own Tpus, sothey trained it on their own
chips. So now all of a sudden Google's
(21:34):
like, well, hang on, I'm a ChatGPT competitor and I'm an
NVIDIA competitor at the same time right now Amazon are like,
OK, well, OK, I'll see your I'llkind of see your TP use and
I'll, I'll, I'll give you, I'll give you my training 3, right?
Which is their equivalent, right?
(21:55):
So look, this is just we knew itwas always going to happen.
The the big guns turn up at the party and NVIDIA have been
enjoying I think they've got 90%market share, right.
So and they've been and with that have had the luxury of
pricing power. And so their revenues have just
(22:18):
been nothing short of disgustingin terms of revenue growth,
right. But but there's competition.
So Amazon are saying, hey, we'vegot our training three note the
the letters, sorry, the number three there.
So to your point, this isn't just something they've quickly
rushed out. You know, they've been working
on this chip for a while and nowthis is the third generation and
(22:39):
look, they they've trained some of their kind of AI capability
on their chips and that that's that's the proof point, right?
Like Google have, you know, thisis just isn't a a random new
sideline that we've suddenly spun up to compete with NVIDIA.
We've been working on this internally for a long time and
(23:00):
it works because we've trained our own stuff using these chips.
So it's a great kind of sort of validator right out of the gate.
And obviously they won't have a problem with distribution and
they won't have a problem with ramping up manufacturing, you
know, these big boys. So so yeah, this is this is
Amazon coming to the chip table.And yeah, it looks pretty
(23:25):
interesting. Well, yeah, maybe I could give
you some details on the chip in terms of context and then we can
talk about the market fit. Yeah.
So compared to the previous generations of these chips,
because like you said, this is training in three servers and
they're about four times faster and roughly 4 times more energy
(23:46):
efficient. Importantly with early customers
testing indicating cost reductions of up to 50% versus
traditional GPU based setups. And obviously for Amazon, you
know what's critical to that success of that business is
their cloud and subsequently cloud AI for rivals that lean
(24:08):
heavily on NVIDIA and the attempts strategically to
reposition AWS as that market leader.
Not just about right with a business like Amplify me, right,
we're going to rent some Amazon servers.
No, this is the underlying AI hardware stack as well as the
cloud. The entire ecosystem, which
(24:28):
we'll go on to talk about, has many benefits from a user
perspective. Yeah.
And look, look at the end of theday, like if you're a customer
wanting to buy these chips, well, at the moment you can't
because NVIDIA, basically all ofNvidia's supply is getting
(24:50):
bought by the big guns who are bullying themselves to the front
of the queue, right? So you can't get hold of this
stuff. If you can, it's ridiculously
expensive. So if you get another player
coming to the market at half price, like even if it's not
quite as good, I mean, it's better than a double the price
chip that you can't buy anyway. So So there's two points there,
(25:13):
though, because there's two sides to that.
Number one. Well, isn't this really bad for
NVIDIA? Then they're going to lose
market share. Well, fine, NVIDIA stock is down
14%, but it's not like it's collapsed.
The point is here, the market's so big, demand is so far right
stripping supply, there's probably room at this table for
(25:35):
a few big guns. Yes, NVIDIA have enjoyed the
whole table for themselves for the last couple of years, but
there's plenty of space for others to come in and take up
some of that that demand, right?So that's one point to make.
Don't like rush out the door andsell all your NVIDIA stock
immediately, because that's it. The game's up.
(25:57):
So that's one point to make. But yeah, at the end, I think I
do like the energy side of it. Four times more energy efficient
is a really important point. And as all of these guys go
through their generations, then they're going to have to, you
know, bring in some major energykind of efficiencies, if you
like. Yeah.
(26:18):
And you know, I think there's like you, you were just kind of
alluding to, I think there's just different on the enterprise
side. I think there's just different
use cases. So there's your regular company
that's moving things to the cloud and then there's
developers who who need like thepower to, you know, on the on
the layers to drive their development projects.
(26:40):
So if you're an enterprise and you're thinking about, you know,
your cost conscious, so cost perunit of useful compute rather
than thinking about benchmark performance, right?
You know, this is where trainingin three aims to make large
scale AI training agentic AI, all these sort of video
(27:00):
generation materially cheaper torun.
Because one of the things on Google is it's got that was it
nano banana where you can generate these videos content,
things like that. And actually there's different,
I think the use cases for the for individuals as well as
changing even for my own habits.Whereas ChatGPT 5.1 is quite
(27:22):
quick, whereas Gemini is quite slow.
But the output qualities are quite different.
And actually my behaviour has changed where I'm not actually
now looking for speed, I want quality.
And so you're, you're willing tosacrifice some speed and things
like that, which is quite interesting.
But yeah, and ultimately this whole kind of, I guess the the
(27:43):
terminology from a technical perspective is this vertically
integrated hardware plus the sweetener, the managed
infrastructure part, which is then where it all comes together
for Amazon. Why it's such a beast because it
reduces those supply bottlenecks.
It gives Amazon more control over pricing capacity allocation
(28:04):
than if it was to rely purely ona NVIDIA solution.
Yeah. And the cloud compute race,
there's three players. Basically.
It's Amazon AWS, it's Azure fromMicrosoft, and then it's Google
Cloud, right, who come in very much third.
But actually Microsoft and Google have been really
benefiting more than AWS over this sort of two year period.
(28:29):
And this is kind of AWS kind of finally strategically kind of
positioning themselves to, as you said from that vertical
stack to kind of start to benefit on the cloud side.
So it may be that 2026 is going to see Amazon return to the, you
know, super fast cloud growth trajectory that we've been
(28:50):
seeing from Microsoft and Googleand that we used to see from
Amazon in the past, they've justkind of plateaued a little bit,
but maybe now this is the time that AWS revenue growth will
start to accelerate. One interesting point I saw in
our notes was that talking aboutNVIDIA and Amazon obviously at
the moment is a very large significant customer of NVIDIA.
(29:13):
We talked about Nvidia's concentration risk.
Well, they've was it four or five clients that they've got
which carries the bulk of their revenue.
I think 5 clients is 60% of their revenue.
I think so. So here, more than 10% of
Amazon's capital expenditures godirectly to Nvidia's products.
Yeah. And on the vice versa, Amazon
(29:35):
accounts for roughly 7 1/2% of Nvidia's revenue.
Yeah. So ramping up his own chips, you
know, chipping away at that relationship where NVIDIA is
already exposed to to some degree.
But I know that it's not quite as straightforward as that.
It's not about displacement. As you said, there's much demand
(29:57):
to go around to feed many males here.
And so one of the things that I think looking ahead was talking
about Amazon already saying thatthey're working on Tranium 4I
mean, of course they are like, like we know there's all these
tech firms, there are many the developments happening and the
release cycle is very different.And then next generation AI
accelerator is designing to be compatible with Nvidia's NV Link
(30:22):
Fusion networking tech, which connects chips with AI server
racks a very high bandwidth. So it's a strategy here for a
lot of these companies, maybe even NVIDIA and in itself is
that it's going to give some ground here.
We talked about that kind of somewhat incestuous investment,
(30:43):
circular narrative. But in the end, there there
there is. Better to be together and
working in that way than try andcompete against each other.
I think so that's right. And so that's why this isn't
necessarily a mega bear case forNVIDIA.
(31:04):
Yeah, there's good, you know, these big, big tech, they're all
going to be at the table. I mean, there wasn't going to be
1 winner for for AI, it's. Sort of like we were saying,
like I think yeah, AWSI mean maybe it positions itself is
it's a hybrid solution and you could have access to a business
(31:24):
could have access to Tranium or NVIDIA GP US, as I said,
depending on the necessity of the demand you require from an
output. So you have the same high speed
fabric, let customers mix and match chips right?
And instead of a in what they can access a single architecture
(31:46):
and then start to then pick off the menu and choosing a single
vendor within that mixed within your ecosystem.
It just makes active sense. Depending on what you're what's
most important to you as an individual customer and and
customers are different, so they'll want they'll have
different priorities. Cool.
All right, well, look, one of the one of the things here to
wrap up then conversation was that things aren't looking so
(32:09):
rosy for our our dear fellow SamAltman at Open AI because he's
had a he's had an absolute breakout year, but he's maybe
getting a little bit nervous around the, you know, cutting
the Turkey this Thanksgiving because just why everyone was
kicking back, enjoying a feast. He was declaring code red, which
(32:32):
sounds pretty dramatic because Ithink in my head I hear code
red. I don't know, I think of a blend
of Arnold Schwarzenegger and andalso submarines like the Russian
sort of this is the sort of where my showing my age probably
slightly, but they have a traffic light system, right?
That's how they open AI prioritize the projects in which
(32:56):
their multiple teams are workingon.
So you know, you have yellow, orange, red signalling urgency
across their their team. So yeah, what?
What? Anything to add on what has
actually happened here at Open AI?
Well, in short, they've got competition.
It's like NVIDIA, but now not not in the chip game, but
(33:16):
actually in the chat sort of space, right, the AI agent
space. And so again, ChatGPT, I mean,
they kicked off this whole revolution when they dropped
their their kind of product to the consumer market and it just
waste the fastest growing product in history, right?
But big tech are hunting them down and look, you can't.
(33:42):
In the end, can you compete? I don't know.
But Gemini, for example, over atGoogle, they are just cut.
Not only is there Gemini 3 now arguably better than ChatGPT 5,
user numbers are kind of really catching up.
So actually, if you look at active monthly users on Gemini,
(34:02):
it's gone from 450 million usersearlier in the year to now 6000,
Sorry, in July, it was 450 million users.
In October 650, you know, they are coming.
And if you look in your rear view mirror and you can see
Google gaining ground that, I mean, code red.
I mean, isn't there a code that's even more urgent than
(34:26):
red? Because I put that one on.
So, yeah, I mean, look, they need to just double down here.
They're feeling a lot of pressure.
But yeah. And the other one is Anthropic
because look, Gemini's thought, why is this now the best one to
use if you're a individual consumer, maybe.
And then on the sort of enterprise side, you've got
(34:47):
Anthropic and their one is called Clawed and their and
their model is kind of graining gaining a lot of traction
amongst the business, you know, customer community.
And that's a roading open a eyesearning lead there, right.
So they've kind of got these twokind of competitors that are
gaining ground and someone was like, right, stop, because
(35:09):
they're working on some other stuff, right?
Open AI, they're working on someother gadgets and some wearables
and all this kind of other stuffthat kind of fits in and around
their ChatGPT. But Sam's basically gone.
Right stop, stop, everything down, tools on everything.
Everyone's all in back on ChatGPT here because we need to
(35:30):
accelerate the kind of evolutionof our product because the
others are catching it. It's always tough as being on
top NVIDIA and open AI. Yeah, well, I tell you,
especially for open AI when you've got the hyper scale tech
giants like Google in your race,that's that's tough, tough to
(35:56):
compete. All right.
Well, look, we'll wrap it up there.
Any points that you that you have to add or questions,
anything like that, feel free toto drop or say a question if
you're listening on Spotify or on YouTube.
Otherwise, I think we've got oneor two episodes left before we
get to the end of the year. So stay tuned and thank you,
(36:18):
Piers. Yeah, have a good week ahead.
Hello and welcome back to the Market Maker podcast and welcome
back to the M and A show, which is where we do a bit of an
overview of some of the major deals of the last week and talk
about some other interesting headlines related to the field
of corporate finance. So I'm going to have to
apologise to our researcher Saab, who always does such a
(00:21):
thorough job, but I'm going to completely skip over his deals
of the week just just because A,there's not too much sexy there,
but because what is sexy is one of the biggest lined up IPOs of
the year coming just in the nickof time.
So going to talk about that for a little bit.
(00:43):
Really interesting developments then of what we've seen in the
league tables. We might well have for any of
our long time listeners a special guest or one and only
Stephen Barnett might be coming back for an outlook on the M&A
scene for 2026. So that is coming folks.
(01:04):
Let's get the red carpet out. Get the red carpet out for
Stephen. But we'll have a little touch
here because this latest IPO, we're going to talk about
Medline is of such an order of magnitude that it is going to
shift around the league tables alittle bit in terms of the
sectors that have been driving the ECM part of the investment
banking scene. And then we might talk about AI,
(01:28):
but in the context of everyone'sbeen talking about how Google
has kind of thrown the cat amongst the pigeons in a sense
of Gemini and what's come out. And that's put the red alert on
as a status at open AI, but alsothe chips.
What we were talking about with with Google recently and how
that's impacted NVIDIA last week, Amazon have come forward
(01:52):
with their latest 1. So what's Amazon got to offer
and should NVIDIA be worried? So, yeah, perhaps we can start
with this, this biggest IPO dealof the year, Pierce.
Yeah, Medline, here we come. This could be an absolute
monster where they are, you know, looking to raise $5
(02:15):
billion. I mean, look, just to just to
get the timings right here. It hasn't happened yet, but it's
probably looking looking on for early part of next year
probably. Hopefully January maybe.
But this is an interesting case for a couple of reasons.
Firstly, they've been on the docket for an IPO for a while.
(02:37):
They were the one of the big kind of names on the list that
decided to delay their IPO because of Trump's tariff
policy. And and look, partly that's
because of the line of business that they're in.
So that you know, who are Medline, They're actually a
privately well, they were a family run business run out of
(02:57):
Chicago. We'll talk about the PE deal
they did in 2021 in a minute, but family business business out
of Chicago, they they distributemedical supplies to like
hospitals and clinics. And so we're talking surgical
equipment, you know, gloves, gowns, you know all the rest of
it, right? And they supply that that
(03:18):
globally. So actually they've got 43,000
employees, they operate over 100countries, sales around about
$23 billion, right? So from a tariff perspective,
obviously you've got a fully global business here and it's
about obviously shipping equipment, you know, all around
the planet. And if tariffs were going to
(03:39):
come in and really put the the glue in, in the kind of movement
of goods globally, then that wasgoing to be a really big issue
for them. So they kind of just stepped
back and said, all right, their PE owners, I should say, stepped
back and said, OK, we're not going to IPO now.
We're going to wait. Originally they'd wanted to go
early 2025, but they kind of changed their mind.
(04:03):
So, you know, this is the business, you know, why do they
want to IPO? Because well, it's a private
equity cycle, right? the PE boys came in.
In fact, they're the biggest privately owned US distributor
of medical products. And so they did APE deal back in
2021 and it was acquired for $34billion.
(04:25):
OK, So it was a leveraged buyoutand you had the big guns like
Blackstone, Carlisle, Hellman and Friedman.
I think you had the sovereign wealth funds getting involved
from the Middle East as well, right?
Re obviously 34 billion dollar deal, that's that's a huge check
size. So you had a consortium of PE
(04:45):
firms coming together to, to kind of kind of do that.
But of course private equity companies, their strategy is to
buy, then increase the value of that asset, then sell.
OK. It's very much on a cycle.
The PE world often work on this kind of three to five year idea
where it takes them three to five years to a pay down some of
(05:06):
the debt. So remember, normally this is a
leveraged buy out. So the PE firms will use debt as
well as a little bit of cash to buy the company.
They'll then run the company andin the meantime pay down that
debt so that then when they sellit at the end of the cycle, you
know they've got more of an equity portion and it and it
(05:28):
kind of leverages their returns.OK.
Also at the same time, it's not just about buying the company,
paying down the debt and then selling it.
They can they improve the business, can they grow the
business or can they improve margins and make it more
profitable. And so actually this has been
executed really well. So we've seen the profitability
(05:49):
and the margins of Medline improving through this kind of
PE cycle. So it kind of made sense that it
was ready to go. The problem that the PE industry
have had though is that the IPO market has been closed, right?
Deals have just not been coming because the investor appetite
(06:09):
just hasn't been there. Why interest rates?
First of all, interest rates have been really high.
And look, if you want to do sellbig assets like this, who are
you selling to? Well, investors will often need
to leverage up and take on some debt to finance that purchase.
But if debts really expensive because interest rates are high,
(06:29):
it just puts a whole puts on puts the ice all over that kind
of setup, right. So private equity firms have
been incredibly frustrated and that they haven't been able to
exit through the IPO channel a lot of these investments, which
is bad. It's a bad look for the
(06:50):
investors from the private equity firms.
Investors, you know, they want payouts, they want returns, they
want liquidation events, right? And these PE firms haven't been
able to do it. So this would be an awesome deal
to get over the line, A, becauseof its size, it's absolute
monster, but B, it would be a great kind of signal for these,
(07:13):
particularly the USPE space, to see a really big liquidation
event and see these PE firms being able to return capital
back to their investors. So yeah, it's it's certainly a
headline grabbing moment here. And then I'm assuming that
because of the size of this deal, hence the reason why
(07:37):
there's so many investment bankson the books.
So lead book runners here I've got listed are Goldman Sachs,
Goldman Sachs, Morgan Stanley, Bank of America, JP Morgan, so
all of the big boys. I mean, that's, I mean we'll
look in a second, but I'm assuming that's like 4 of the
(07:58):
top five. Right now, yeah, you're missing
city. It's fourth in the top five.
Yeah. And we can dive into this
actually maybe this is a good segue to talk about, you know,
the the divisional breakdown of when we when we talk about
investment banking, obviously the mine tends to gravitate
(08:19):
towards M&A fees, but ECMDCM andwho are the the leading ones
there and when these bigger firms naturally have much bigger
franchises when they have the ability to take on that type of
business for sure. So the other thing is it's going
to be on the NASDAQ. So in terms of the exchange, so
(08:41):
the NASDAQ Global Select Market under the ticket MDLN.
So, yeah, any other points on that before we move on?
Well, I mean, it probably may beworth just reminding people of
the process here, why it's suddenly coming to the news,
because so the, the process is that first you've got to file a
prospectus with the regulators. And that's what's just happened,
(09:02):
right? And so it's like, wow, OK,
that's your signal that right, these PE firms and Medline, OK,
they're back at the table here looking to get a kind of process
done. You file the prospectus with the
regulators, they check it, they verify that the information in
it is valid and true. Then it's like go time.
Well, what does that mean? Well, it means these banks then
(09:25):
run Well, they they kind of run a roadshow right where they
basically take Medlines kind of management team and they take
them on a road show and sit themin front of, you know, big
investor communities. And it's about telling the
story, you know, why is this an investable opportunity?
You know, talking about your hockey stick charts with amazing
(09:47):
future growth and all the rest of it, getting them excited.
OK. Then hopefully from that process
you'll also get what we call some cornerstone investors,
which will be, you know, hopefully like a small handful
of big, you know, notable investors that are saying yes,
we're in. We want to be a kind of
(10:08):
cornerstone investor. This is great for a process
because if you've got a big nameon the ticket, then obviously
that's reassuring for all the other investors in the community
and perhaps then kind of greasesthe wheels for other investors
to get on board, right? So you might want to try and
find some cornerstone investors,but then you do a book building
(10:29):
process after the roadshow, which is like, hey, right, you
lot, you've heard the story, right?
Who's interested, you know, showus your interest.
And then from so you basically get these kind of bids coming
in, which is then enables and that's why you need these
multiple banks on the ticket if you're trying to sell something
that's going to be worth we think roughly $50 billion.
(10:52):
So that'll be a great flip. They bought it for 32 billion in
2021. They're looking to flip it for
50 here, and obviously the amount of debt in that business
will have reduced as well. So this is going to be a great
return for the PE firms on such a massive scale, right?
But 50 billions a lot. So you need multiple book
runners out there. So you need the big guns tapping
(11:12):
into their respective really large networks.
And so you do the book building,which gives you a gauge on
demand because ultimately you'vegot to price this thing, right?
Come IPO day, you're going to issue shares and sell them.
What's the share price? Well, these banks calculate the
share price from understanding the demand that they've been
(11:36):
able to gauge during the book building process.
They'll often give guidance, right?
So when they're out on the roadshow, they'll guide, right?
We're looking at a range and they'll give a share price range
and that's something that the investors can kind of have a
think about. And then when they put their
bids forward, they kind of placetheir bid somewhere in the
range. And if you're super positive on
(11:58):
this opportunity and you really want it, you're going to bid top
end. If you're lukewarm but
interested, maybe you'll go bottom end, right?
And then you build this book of demand, then they price it and
hopefully they price it correctly because then it's like
right, IPO day, let's now sell these shares to the investment
community that have shown interest at the price we've
(12:20):
decided and great the IP OS done.
That's where the PE firms exit. They sell their shares in the
business to then the other new investors coming in and then it
becomes publicly traded. So that's they've chosen the
NASDAQ and then anybody in the world will be able to trade
these shares in the secondary market on the NASDAQ thereafter.
(12:44):
Yeah, It seems like, seems like this is a significant signal
then on many fronts for deals in2026 because at around $5
billion, this IPO would dwarf the current US leader Venture
Global's $1.75 billion listing in Jan, in January.
Remember beginning of the year, we were all like this is the
(13:05):
year for M&A and then the complete opposite kind of
happened. But as you were just rightly
saying, I mean, I was just looking back on a post that I
did, this was back in four months ago and it was the FT ran
an article talking about privateequity firms flipping assets in
record numbers to themselves in using continuation funds to cash
(13:30):
out for investors. But is it just, is this a PE
bubble? We're recycling this debt as you
were describing. But as you said, they're going
to do very well out of this. And in combination with the
general macro climate, which is indicating at least at this
point that rates in the US are going to come down.
(13:52):
You're going to have a Fed shiftthat's probably going to be more
aligned and more favourable for those more dovish conditions as
well. So what looking quite good
again, or we were, we were hypothetically thinking this
time last year that Trump's going to come in and create that
environment. Whereas this year, that's now a
(14:14):
known thing and a lot of those uncertainties or indeed risks
over Trump have somewhat dissipated, particularly the big
trade debacle. So is this now actually a much
more known fertile setup for 2026 than it was actually
looking back retrospect for 2025?
(14:34):
I mean, yes, you know, without wishing to make the same mistake
twice, yes, it looks good for 2026, but the main thing is
interest rates for these deals. So as long as the Fed are dovish
in 2026 and who knows, does Hassett get the gig and all the
(14:55):
rest of it and do rates come down, then you're talking?
Because I think at the moment, like part of the story about the
IPO deal as they're, sorry, the IPO market being a great winner
from the Trump presidency, you know, half of it was about the
regulatory side, like removing red tape.
And Trump hasn't really got ontothat part of his agenda yet.
(15:16):
He's been, so his docket's been full with his tariff game,
right? So I think in 2026, we'll see a
little bit more of a focus on that from the Trump
administration. If you can couple that with a
rate cutting cycle continuing, that's a pretty, that's a pretty
interesting kind of kind of cocktail to feed through to what
(15:38):
is a huge pent. There's a massive queue of
companies want or PE firms particularly wanting to exit
investments right there. They are queuing up.
So it's all there. It's just ready to come and
maybe that the cork is going to get pulled out of that thing and
we might get a really interesting year of activity on
(16:01):
the IPO market in 2026. So maybe then to finish here, we
know this is going to be, if it goes ahead, the biggest IPO.
So where does that puts it in terms of the sectors and then
the rankings for these book runners?
Yeah, well, on from a sort of, well the top 10 IPOs so far this
(16:21):
year. Let's start there.
I'm actually just looking down the list.
The first the, the US, the firstUS entry on the list, this is
like a crazy impressive is actually 5th.
So there are four bigger IPOs than any you I mean the US is
always top of this league table always.
(16:42):
So they're they're coming in 5thand that was Venture Global and
the ones above that are coming out of APEC like like all of
them, right. So for the US this will put them
back on the top. This will this they will be the
biggest deal, but it's not goingto come in 2025.
So it's a bit academic that, butbut yeah, in terms of sectors as
(17:06):
usual, the technology sector is the largest player, right, As in
the volume of IPO deals, technology is leading the pack,
but healthcare which is currently in fourth.
So if you think about the sectors in 2025 so far,
healthcare are so technology is #1 industrials is number 2,
(17:30):
Financials 3, healthcare for materials five.
OK, in the order, this is deal volume, IPO deal volume by
sector in 2025. If they land this deal, then
that healthcare sector 1, I think that'll push them up 1
notch to 3rd. I think that'll get them up up
above the financials. But yeah, still technology is
(17:53):
out like 25 billion, whereas industrials is about 19 billion,
financials at 16 billion. Healthcare currently sat around
about 12 1/2 billion. So yeah, they're definitely not
going to get out to the technology.
They're not going to hit top spot.
But healthcare generally, as I said in the episode last week,
(18:15):
it's kind of it's been a really bright, bright spot of the
marketplace in the last one month when you've had a little
bit of a tech, big tech rolloverwith NVIDIA down whatever 14% in
the last one month, the healthcare stocks have really
been ramping. And a lot of this is due to the,
it's not just IPO, it's all the M&A activity that's going on in
(18:38):
that sector and now this monsterIPO.
And it's like, wow, OK, this sector's looking really hot and
this is feeding into some share price appreciation.
And if they get this Medline deal away if and it's successful
and if there's large demand and if it goes for top end of the
price range, if, if, if, then itcould really be a great
(19:02):
statement and a great measure ofthe market appetite right at the
start of 2026, which could really set the ball rolling for
the 12 month period ahead. Yeah, when you talk about that
pent up demand as just looking at the stats here, because you
said Apex like the outright kindof runaway story here, but the
(19:24):
US suffering, as you said earlier by the shutdown because
that paused new listings, right?And in the third quarter alone,
there were 64 IPOs in the US raised 15 points or $15.3
billion. So the shutdown when it did
happen a few weeks ago, happenedright at possibly the worst time
(19:45):
if you were doing IPOs. But meaning though that they
were they were running hot. You've had a gap.
Rates are now probably heading lower.
Yeah. So probably going to come back
pretty quickly if Medline gets away, as you said.
Really. Well, cool.
All right, Well, let's let's head on and talk about this
(20:07):
Amazon. Yeah, it's funny.
It's like, as we will go on to describe, Amazon have been
working on this stuff for a longtime.
However, it's funny that they'rethe news breaks just when they
feel like they've got to come tomarket and start talking this
stuff up in the context of we'vehad, you know, this is the third
(20:27):
element or the third leg, so to speak.
It's been NVIDIA, we're talking Google last week, now we're
talking Amazon. So what have Amazon got to bring
to the table? Amazon have got Tranium 3,
that's what they've got. Just slapped that onto the
table, which is their latest in the house, basically AI
(20:49):
accelerator chip. You know, it's a, it's direct
competition for Nvidia's GPUs. And you'll remember that Google,
like it's so interesting with Google and actually Google, I
mentioned the tech sector rolledover a bit.
NVIDIA down 14%. Google's up 13% in the last one
month because of the, the drop of Gemini.
(21:12):
You know, I forget the versions now.
Is it Gemini 3? And it being, wow, hang on, this
thing's amazing. Hang on, isn't this, this is
better than ChatGPT? And really key point, it was
trained on Google's own Tpus, sothey trained it on their own
chips. So now all of a sudden Google's
(21:34):
like, well, hang on, I'm a ChatGPT competitor and I'm an
NVIDIA competitor at the same time right now Amazon are like,
OK, well, OK, I'll see your I'llkind of see your TP use and
I'll, I'll, I'll give you, I'll give you my training 3, right?
Which is their equivalent, right?
(21:55):
So look, this is just we knew itwas always going to happen.
The the big guns turn up at the party and NVIDIA have been
enjoying I think they've got 90%market share, right.
So and they've been and with that have had the luxury of
pricing power. And so their revenues have just
(22:18):
been nothing short of disgustingin terms of revenue growth,
right. But but there's competition.
So Amazon are saying, hey, we'vegot our training three note the
the letters, sorry, the number three there.
So to your point, this isn't just something they've quickly
rushed out. You know, they've been working
on this chip for a while and nowthis is the third generation and
(22:39):
look, they they've trained some of their kind of AI capability
on their chips and that that's that's the proof point, right?
Like Google have, you know, thisis just isn't a a random new
sideline that we've suddenly spun up to compete with NVIDIA.
We've been working on this internally for a long time and
(23:00):
it works because we've trained our own stuff using these chips.
So it's a great kind of sort of validator right out of the gate.
And obviously they won't have a problem with distribution and
they won't have a problem with ramping up manufacturing, you
know, these big boys. So so yeah, this is this is
Amazon coming to the chip table.And yeah, it looks pretty
(23:25):
interesting. Well, yeah, maybe I could give
you some details on the chip in terms of context and then we can
talk about the market fit. Yeah.
So compared to the previous generations of these chips,
because like you said, this is training in three servers and
they're about four times faster and roughly 4 times more energy
(23:46):
efficient. Importantly with early customers
testing indicating cost reductions of up to 50% versus
traditional GPU based setups. And obviously for Amazon, you
know what's critical to that success of that business is
their cloud and subsequently cloud AI for rivals that lean
(24:08):
heavily on NVIDIA and the attempts strategically to
reposition AWS as that market leader.
Not just about right with a business like Amplify me, right,
we're going to rent some Amazon servers.
No, this is the underlying AI hardware stack as well as the
cloud. The entire ecosystem, which
(24:28):
we'll go on to talk about, has many benefits from a user
perspective. Yeah.
And look, look at the end of theday, like if you're a customer
wanting to buy these chips, well, at the moment you can't
because NVIDIA, basically all ofNvidia's supply is getting
(24:50):
bought by the big guns who are bullying themselves to the front
of the queue, right? So you can't get hold of this
stuff. If you can, it's ridiculously
expensive. So if you get another player
coming to the market at half price, like even if it's not
quite as good, I mean, it's better than a double the price
chip that you can't buy anyway. So So there's two points there,
(25:13):
though, because there's two sides to that.
Number one. Well, isn't this really bad for
NVIDIA? Then they're going to lose
market share. Well, fine, NVIDIA stock is down
14%, but it's not like it's collapsed.
The point is here, the market's so big, demand is so far right
stripping supply, there's probably room at this table for
(25:35):
a few big guns. Yes, NVIDIA have enjoyed the
whole table for themselves for the last couple of years, but
there's plenty of space for others to come in and take up
some of that that demand, right?So that's one point to make.
Don't like rush out the door andsell all your NVIDIA stock
immediately, because that's it. The game's up.
(25:57):
So that's one point to make. But yeah, at the end, I think I
do like the energy side of it. Four times more energy efficient
is a really important point. And as all of these guys go
through their generations, then they're going to have to, you
know, bring in some major energykind of efficiencies, if you
like. Yeah.
(26:18):
And you know, I think there's like you, you were just kind of
alluding to, I think there's just different on the enterprise
side. I think there's just different
use cases. So there's your regular company
that's moving things to the cloud and then there's
developers who who need like thepower to, you know, on the on
the layers to drive their development projects.
(26:40):
So if you're an enterprise and you're thinking about, you know,
your cost conscious, so cost perunit of useful compute rather
than thinking about benchmark performance, right?
You know, this is where trainingin three aims to make large
scale AI training agentic AI, all these sort of video
(27:00):
generation materially cheaper torun.
Because one of the things on Google is it's got that was it
nano banana where you can generate these videos content,
things like that. And actually there's different,
I think the use cases for the for individuals as well as
changing even for my own habits.Whereas ChatGPT 5.1 is quite
(27:22):
quick, whereas Gemini is quite slow.
But the output qualities are quite different.
And actually my behaviour has changed where I'm not actually
now looking for speed, I want quality.
And so you're, you're willing tosacrifice some speed and things
like that, which is quite interesting.
But yeah, and ultimately this whole kind of, I guess the the
(27:43):
terminology from a technical perspective is this vertically
integrated hardware plus the sweetener, the managed
infrastructure part, which is then where it all comes together
for Amazon. Why it's such a beast because it
reduces those supply bottlenecks.
It gives Amazon more control over pricing capacity allocation
(28:04):
than if it was to rely purely ona NVIDIA solution.
Yeah. And the cloud compute race,
there's three players. Basically.
It's Amazon AWS, it's Azure fromMicrosoft, and then it's Google
Cloud, right, who come in very much third.
But actually Microsoft and Google have been really
benefiting more than AWS over this sort of two year period.
(28:29):
And this is kind of AWS kind of finally strategically kind of
positioning themselves to, as you said from that vertical
stack to kind of start to benefit on the cloud side.
So it may be that 2026 is going to see Amazon return to the, you
know, super fast cloud growth trajectory that we've been
(28:50):
seeing from Microsoft and Googleand that we used to see from
Amazon in the past, they've justkind of plateaued a little bit,
but maybe now this is the time that AWS revenue growth will
start to accelerate. One interesting point I saw in
our notes was that talking aboutNVIDIA and Amazon obviously at
the moment is a very large significant customer of NVIDIA.
(29:13):
We talked about Nvidia's concentration risk.
Well, they've was it four or five clients that they've got
which carries the bulk of their revenue.
I think 5 clients is 60% of their revenue.
I think so. So here, more than 10% of
Amazon's capital expenditures godirectly to Nvidia's products.
Yeah. And on the vice versa, Amazon
(29:35):
accounts for roughly 7 1/2% of Nvidia's revenue.
Yeah. So ramping up his own chips, you
know, chipping away at that relationship where NVIDIA is
already exposed to to some degree.
But I know that it's not quite as straightforward as that.
It's not about displacement. As you said, there's much demand
(29:57):
to go around to feed many males here.
And so one of the things that I think looking ahead was talking
about Amazon already saying thatthey're working on Tranium 4I
mean, of course they are like, like we know there's all these
tech firms, there are many the developments happening and the
release cycle is very different.And then next generation AI
accelerator is designing to be compatible with Nvidia's NV Link
(30:22):
Fusion networking tech, which connects chips with AI server
racks a very high bandwidth. So it's a strategy here for a
lot of these companies, maybe even NVIDIA and in itself is
that it's going to give some ground here.
We talked about that kind of somewhat incestuous investment,
(30:43):
circular narrative. But in the end, there there
there is. Better to be together and
working in that way than try andcompete against each other.
I think so that's right. And so that's why this isn't
necessarily a mega bear case forNVIDIA.
(31:04):
Yeah, there's good, you know, these big, big tech, they're all
going to be at the table. I mean, there wasn't going to be
1 winner for for AI, it's. Sort of like we were saying,
like I think yeah, AWSI mean maybe it positions itself is
it's a hybrid solution and you could have access to a business
(31:24):
could have access to Tranium or NVIDIA GP US, as I said,
depending on the necessity of the demand you require from an
output. So you have the same high speed
fabric, let customers mix and match chips right?
And instead of a in what they can access a single architecture
(31:46):
and then start to then pick off the menu and choosing a single
vendor within that mixed within your ecosystem.
It just makes active sense. Depending on what you're what's
most important to you as an individual customer and and
customers are different, so they'll want they'll have
different priorities. Cool.
All right, well, look, one of the one of the things here to
wrap up then conversation was that things aren't looking so
(32:09):
rosy for our our dear fellow SamAltman at Open AI because he's
had a he's had an absolute breakout year, but he's maybe
getting a little bit nervous around the, you know, cutting
the Turkey this Thanksgiving because just why everyone was
kicking back, enjoying a feast. He was declaring code red, which
(32:32):
sounds pretty dramatic because Ithink in my head I hear code
red. I don't know, I think of a blend
of Arnold Schwarzenegger and andalso submarines like the Russian
sort of this is the sort of where my showing my age probably
slightly, but they have a traffic light system, right?
That's how they open AI prioritize the projects in which
(32:56):
their multiple teams are workingon.
So you know, you have yellow, orange, red signalling urgency
across their their team. So yeah, what?
What? Anything to add on what has
actually happened here at Open AI?
Well, in short, they've got competition.
It's like NVIDIA, but now not not in the chip game, but
(33:16):
actually in the chat sort of space, right, the AI agent
space. And so again, ChatGPT, I mean,
they kicked off this whole revolution when they dropped
their their kind of product to the consumer market and it just
waste the fastest growing product in history, right?
But big tech are hunting them down and look, you can't.
(33:42):
In the end, can you compete? I don't know.
But Gemini, for example, over atGoogle, they are just cut.
Not only is there Gemini 3 now arguably better than ChatGPT 5,
user numbers are kind of really catching up.
So actually, if you look at active monthly users on Gemini,
(34:02):
it's gone from 450 million usersearlier in the year to now 6000,
Sorry, in July, it was 450 million users.
In October 650, you know, they are coming.
And if you look in your rear view mirror and you can see
Google gaining ground that, I mean, code red.
I mean, isn't there a code that's even more urgent than
(34:26):
red? Because I put that one on.
So, yeah, I mean, look, they need to just double down here.
They're feeling a lot of pressure.
But yeah. And the other one is Anthropic
because look, Gemini's thought, why is this now the best one to
use if you're a individual consumer, maybe.
And then on the sort of enterprise side, you've got
(34:47):
Anthropic and their one is called Clawed and their and
their model is kind of graining gaining a lot of traction
amongst the business, you know, customer community.
And that's a roading open a eyesearning lead there, right.
So they've kind of got these twokind of competitors that are
gaining ground and someone was like, right, stop, because
(35:09):
they're working on some other stuff, right?
Open AI, they're working on someother gadgets and some wearables
and all this kind of other stuffthat kind of fits in and around
their ChatGPT. But Sam's basically gone.
Right stop, stop, everything down, tools on everything.
Everyone's all in back on ChatGPT here because we need to
(35:30):
accelerate the kind of evolutionof our product because the
others are catching it. It's always tough as being on
top NVIDIA and open AI. Yeah, well, I tell you,
especially for open AI when you've got the hyper scale tech
giants like Google in your race,that's that's tough, tough to
(35:56):
compete. All right.
Well, look, we'll wrap it up there.
Any points that you that you have to add or questions,
anything like that, feel free toto drop or say a question if
you're listening on Spotify or on YouTube.
Otherwise, I think we've got oneor two episodes left before we
get to the end of the year. So stay tuned and thank you,
(36:18):
Piers. Yeah, have a good week ahead.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
My Favorite Murder with Karen Kilgariff and Georgia Hardstark
My Favorite Murder is a true crime comedy podcast hosted by Karen Kilgariff and Georgia Hardstark. Each week, Karen and Georgia share compelling true crimes and hometown stories from friends and listeners. Since MFM launched in January of 2016, Karen and Georgia have shared their lifelong interest in true crime and have covered stories of infamous serial killers like the Night Stalker, mysterious cold cases, captivating cults, incredible survivor stories and important events from history like the Tulsa race massacre of 1921. My Favorite Murder is part of the Exactly Right podcast network that provides a platform for bold, creative voices to bring to life provocative, entertaining and relatable stories for audiences everywhere. The Exactly Right roster of podcasts covers a variety of topics including historic true crime, comedic interviews and news, science, pop culture and more. Podcasts on the network include Buried Bones with Kate Winkler Dawson and Paul Holes, That's Messed Up: An SVU Podcast, This Podcast Will Kill You, Bananas and more.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com