[music]

Bloomberg Audio Studios podcasts radio

news.

[music]

Welcome to the Daybreak Asia podcast.

I'm Doug Krner. In the US session, we

had a powerful rally in semiconductor

[music] stocks. This really reflected

the view that the recent selling that we

have seen in this group was simply

overdone. The Philadelphia semiconductor

index jumped by more than 8% and on top

of that other areas in the AI trade

rallied as well. Take for example

Microsoft up 15% on strong cloud growth

and Oracle shares were higher by 8%.

That was after the company expanded its

Gemini AI partnership with Google. And

some more tech earnings after the

closing bell. We heard from Amazon and

Apple. And we're going to take a closer

look now at the market action with

Daniel Newman. He is the CEO of Futurum

Group. Daniel, joining from here in New

York City, thank you so much for being

here. Give me your sense of what you've

seen play out over the last 24 hours.

>> Well, Doug, I think we we are seeing a

few different things at the same time.

Structurally, we believe that the AI

trade, the AI buildout is very much

intact. We've had a very good set of

earnings through this cycle so far. more

very strong earnings delivered this week

from the leaders in AI, the leading

hyperscalers as well as the

semiconductor companies. But at the same

time, you saw a market go from, you

know, all-time highs or near all-time

highs to year lows in less than 30 days.

We saw the Cosby uh in Korea almost on

the daily uh you know, [clears throat]

setting off trip uh you know, setting

off trip wires and and having to to stop

the market on a day-to-day. And then of

course here in the US you know we had

that hard fall and people were kind of

wondering what's going on and obviously

you know there was a lot of deleveraging

going on there was there's been a lot of

leverage into the market of course in

Korea that was a thing but it was also

it's also a very significant thing here

in the United States. Um I think what we

saw yesterday um was a lot of the

capitulation that the market wanted to

see and then you got some short covering

and you now saw a huge powerful rally

that was based on getting some good

results from Microsoft getting some good

results from uh you know today from

Amazon. uh seeing these companies that

are committing these big dollars in

capex feeling confident that they are

spending wisely and building an ROI

strategy for their AI gave the market a

chance to rally back higher.

>> So Dan, I'm sure you're aware that we

had some news today on the hedge fund

called Situational Awareness. The fund

sold the bulk of its public equity

holdings to Citadel. And I think there's

maybe a little bit of irony here in

terms of the firm's name. situational

awareness had been investing in

companies tied to the boom in AI. So

Dan, I'm wondering whether you think

that we're going to get more stories

like this, which is to say hedge funds

being caught, maybe leverage is a factor

and then the significant pain to their

positions that is inflicted as a result.

>> Well, I think we've gotten to the point

where we risk too much leverage in the

market everywhere. You know, this is

what's happened with these, you know,

three and four times leveraged ETFs on

memory. This is what happened to Ashen

Brener. He was, you know, heavily

levered to make the kind of profits that

he was making. And there's always

opportunists in the market. In this

case, it was Citadel. And Citadel was

able to basically buy situational

awarenesses positions at the bottom and

immediately get the kind of rally we got

today. But I think with the market

between the Fed news this week, the

concern about potential interest rate

hikes, the uncertainty about in

inflation, the macro overall, the China

competition for AI with the United

States, I think there is a lot to be

said about the market as a whole and

whether it has anything to do with how

resilient uh or robust the AI trade is.

What we believe though, Doug, is there's

somewhere between 10 and 12 trillion

dollars of cumulative capex that's going

to be spent on AI between now and 2030.

And companies like Nvidia and Micron,

which were, you know, sold heavily,

especially Micron, the memory names, in

this last 30 days, these companies are

trading at extremely cheap valuations

based on their long-term agreements and

the long-term profit expectations that

they have. So I think this shakeout was

probably healthy to delever some of what

was going on in the market, but I think

it also pro provides a good environment

for some of these stocks that have gone

down it feels like for a month straight

to start going back up. So I hear your

optimism when it comes to the buildout

of AI focused on you know the chip

sector and and the hardware

manufacturers but I'm curious about the

hyperscalers whether you expect the

continued growth in data center activity

to be concentrated in these big names

whether it's an Amazon and I want to get

your take on the results that we heard

from Amazon after the bell names like

Meta names like Alphabet names like

Microsoft where we're going to see

continued concentration

among the hyperscalers or will we

gradually see an expansion in data

center buildout to go beyond these these

big companies?

>> Yeah, I I think you're going to see some

augmentation to these bigger companies.

I think these hyperscalers will remain

the largest. They had such a market

advantage coming out. We've seen new

entrance popular names in the market

like Nebius, Cororeweave, iron that are

all building and adding capacity as part

of this AI buildout and boom. But the

hyperscalers have so much uh operational

leverage. They have so much size and

scale. They have the customers in many

cases like in Google's case, they have a

full stack of infrastructure. Amazon and

Microsoft are following suit. Microsoft

has the benefit of the largest

enterprise distribution on the planet

and AWS had the benefit of being the

largest cloud coming into this uh AI

pivot. Uh I think over the next handful

of years what we know for sure is we are

constrained. And so as long as we are

constrained, we will see opportunistic

uh you know new companies entering the

market to compete into these spaces and

you will see these companies actually

partnering with the likes of Google,

Amazon, uh Microsoft, Oracle. So I think

for the next handful of years it's more

of a situation of and than or. But I

think that you know the reason the AI

trade as a whole will continue to have

concern is not so much about to your

point about the the hardware and the and

the capex and the buildout. These

companies Amazon validated and

reaffirmed today that they actually

raised their capex guidance to 220

billion and still can't build enough

compute. Um Google raised its capex

number significantly a week ago after

having 82% growth in its cloud business.

These companies are going to partner.

They're going to expand. They're going

to get access to every bit of compute

they can because what we found out is as

models are changing very quickly. And we

saw the Kimmy incident a year ago. We

had the DeepSeek incident. We're finding

models are not the moat in itself.

The moat is compute and having enough

compute to support this boom.

>> To that point, Deepseek is building an

AI data center in Inner Mongolia. I

mean, do we have to be concerned about I

hear what you're saying in terms of

where the LLM kind of part of the story

fits in and the deepseek moment or the

Kimmy K3 moment in China, but I'm

wondering whether or not we have to be

cognizant of the fact that China is

going to be pretty aggressive in

building out the compute infrastructure,

the building out data centers, and

whether or not they could do it at a

lower cost and what level of risk that

represents to American firms.

Look, this is the most important. It's

an existential battle for global

economic leadership, winning AI. This is

what people have to understand when

they, you know, are are having this

debate about the US and China. They're

having debate about whether the US

provides its leading edge technology uh

from an Nvidia or, you know, whether we

allow ASML to ship their most advanced

machines. China

absolutely is a formidable competitor

and of course they play by different

rules. We've seen that with Kimmy and

others where there's you know suspicion

that Kimmy is trained largely by using

advanced techniques to distill and do

more than just distill u claw anthropic.

And so this is going to be the front in

which we wage the next multiple decades

of economic leadership war. And so yes,

absolutely it's important that we are

concerned. It's important that we build

the right rules, that we provide US

companies the opportunity to do business

in China where it makes sense. But what

we don't do is arm the largest

competitor for economic leadership with

all of our innovation, all of our tools

without understanding the consequences.

>> I want to ask about Apple and then I

have one other question as it relates to

breaking news on Anthropic. What did you

make of Apple's results after the bell?

Um, Apple did fine. I mean, look, the

way Tim Cook was talking about input

costs, his his desire for Chinese

memory, uh, to be able to enable them,

you would have thought they were going

to miss their their profits. I mean,

again, Apple's an operation, you know,

is an operating machine. They they tend

to make their numbers. They very rarely

miss. Uh, the market's selling it

because Apple's been a safe haven for

the last month. It's run to all-time

highs. There's still some questions

around the company's AI strategy. You

know, iPhone sales look decent. Mac

sales were good, but you have to

remember a lot of the the the revenues

are based on pricing right now. Um, not

just the actual unit sales. And and then

of course, you know, China was down. So,

you always have to look at that number.

That's an important one for Apple. But

they did make their their EPS number.

It's notable that they made it off the

tariff income that they had returned.

Otherwise, they would have been in the

midpoint of at the midpoint of their

guidance last quarter as opposed to

beating it. Uh it was a good quarter. It

was a tidy quarter. Apple is the safe

haven of the Mag 7 right now. Um and

investors have have been backing it. I'm

wondering right now though with this

rotation if this is it wasn't so much

Apple's bad print. it was more dollars

are rotating back into this this AI boom

with the Microsoft rally with the Amazon

rally. Nvidia has been oversold. So you

might start to see some of the error

come out of that Apple balloon.

>> You know, as I'm listening to you and

I'm thinking about the rate of change

that we have seen in artificial

intelligence, I'm wondering about the

next generation of handheld devices that

is going to be custommade for this AI uh

system that we are describing. And I

want to get your view on anthropic

because tonight in the US we learned

that a couple of the company's AI models

breached three different organizations

in cyber security testing. Something

went wrong. It went ary and this is only

a little more than a week after open

disclosed a similar incident. And I'm

wondering whether you are concerned

about the rate of change, particularly

on the LLM side, the model side, that

things are moving a little too quickly

for your comfort level and that we need

to try to find a way to modulate this,

or are you okay, and maybe you're of the

belief that the industry can regulate

itself and get this stuff under control?

What's your feeling on that? Well, as a

footnote on your comment about the

future of devices, [clears throat] that

is Apple's existential risk is if a new

device or platform

was to disrupt Apple's

position. Now, going back to the

anthropic and open AI situation, this is

really interesting. You know, you have

two companies that have been sort of

screaming from the rooftop that AI needs

more regulation.

Um, we've had this kind of back and

forth about can the Frontier models and

these closed source win. We've had

companies like Nvidia coming out loudly

to support more open-source globally.

Um, and you know, you have these

jailbreak sort of rogue agents coming

out and part of me almost wonders like

how does this happen? How do two

different trillion dollar companies have

agents doing things that they're not

monitoring them? They're not watching

them. It's wild to me that this took

days on the Open AI one. We don't know

what's going to happen with the

anthropic one yet. With the OpenAI one,

these went on for days before they were

even discovered. Um, so we have

technology and a rate of change that's,

you know, that's exponential right now.

And the problem is policy makers and and

and regulation

can never keep up. So expecting those

things to happen and get ahead is is is

it's never happened in history and now

we have the most important arms race

between the east and the west especially

US and China where we're going to say

let's slow down. I just don't see that

happening. I do think we need more

framework. I do think we need policies

that quickly have models evaluated

before they're sent out in mass. And I

do think some of that's being worked on.

I I think that these systems and some of

these rogue events, they will continue

to happen and we do need to, you know,

pay better attention, but I put a lot of

the blame on this on Anthropic and on

Open AI. They're building new products.

They're shipping them out. They should

be watching that.

>> Dan, we'll leave it there. Thank you so

very much. He is Daniel Newman, the CEO

of Futurum Group, joining from New York

City here on [music] the Daybreak Asia

podcast.

>> [music]

>> Welcome back to the Daybreak Asia

podcast. I'm Doug Krishnner. In New York

trading, US listed shares in Silicon

Motion picked up 24%. This Taiwan-based

semiconductor device maker reported

second quarter results above

expectations. Now, Silicon Motion also

gave a revenue forecast much stronger

than expected. The CFO of Silicon Motion

is Jason Sai. He spoke with Bloomberg TV

host David Inglace and Ivon Man.

>> Jason, it's great to have you back on

earnings and and post earnings certainly

as as what the market was liking here

was the guidance for the third quarter.

Um and of course these blowout earnings

from your side of things in the second.

Um what does it tell you about where you

can still gain market share and really

how sustainable this momentum is now?

>> Yeah, thanks for having me on the show

again. I really appreciate it. Um, look,

I think if you take a look at our core

business over the course of the last

many years, you know, we've been very

consumer ccentric PCs and smartphones

and obviously we've seen all the data

points around the weakness in these

markets driven by high high cost of

memory and storage. We've been gaining a

lot of share here. We've been laying the

foundation for a lot of share gain with

the flash makers, the module makers. And

so we've far exceeded any um, you know,

the the end market results because of

the share gain because we're upgrading

to higherend products. And so that

business, the core business continues to

execute extremely well in a difficult

environment. I think what investors have

really liked is diversifying our

business. Um our automotive uh fair eye

solutions have done extremely well.

That's scaling incredibly well. Our

enterprise boot drive business is

scaling incredibly well and that's going

that's becoming a bigger and bigger

portion of our business. And then you

know our enterprise controller business

is you know was zero a year ago and you

know exiting this year we expect that to

be 5 to 10% of our business and that's

started to scale this second quarter. So

we're we're really seeing a a really

nice diversification of our business

playing out. We're confident that we can

grow our business in excess of 100% this

year. So from you know 800 plus million

last year to you know 100% of that.

There's more than 100% of that this

year. Gross margins are improving.

operating leverage is coming in along

with it. So overall driven by strength

in the core business, expanding

opportunities, improving profitability

and and responsible expense management.

Um you know I think we're we're really

excited by not just what the rest of

this year holds but certainly going into

27 and 28.

>> I mean things look fantastic on all

fronts and I think the fact that you you

know you outperformed your your

guidance. I think my my first question

there then would be you've given

guidance out. Um I mean how how visible

is it? I mean it's it's great to have a

number out there but I just want to get

a sense really of underlying demand

where it's [snorts] it seems companies

are giving out guidance and just

outperforming what they give out.

You know for us there was a lot of

things that happened intraquarter right

you know demand spikes more than

expected our customers are seeing better

sell through our customers are upgrading

the components that they were ordering

all these things are happening

intraquarter you know we've got

inventory that can support that type of

dynamic movement but you know as some of

these new businesses scale new customers

scale sometimes early on they're not

necessarily as confident on kind of the

scale of that um ramp and as that ramp

happens they become more confident on

what demand looks like and they upsize

the orders and so yes I think you know

we we always try to give a balanced view

on what guidance uses a realistic view

on kind of what we think we can do but

certainly um you know the hope is that

you know the the customers are seeing

better demand and I think we're

certainly seeing that through the share

gains and through the the advancements

into new markets, new businesses, and

new customers. One of the things that we

really are proud of is a lot of these

new businesses, the visibility and the

durability is is great. These are

products, our solutions are going into

products that are, you know, in cars and

base stations and in data centers that

are, you know, part of devices that are

tens of thousands, hundreds of

thousands, millions of dollars. And so,

you know, that type of visibility, that

type of design win has [clears throat]

better visibility. It's has higher

barriers to entry, gives us longerterm

understanding as to kind of how we need

to source, what we need to source, what

inventory we need to have. So, it really

helps drive that long-term visibility

for us and gives us confidence on, you

know, not just second quarter, third

quarter doing well, but next two years,

three years doing well.

the again don't get me wrong that's a

fantastic problem to have. Um what are

your conversations like now with uh with

your suppliers? I just want to

understand how you're managing your

future capacity as most demand comes

online.

>> Yeah, you know I think it's well known

that you know TSMC and and a lot of the

substrate uh folks are tight on

capacity. Substrate availability is

tight. We're fortunate that we're one of

the top customers of TSMC and we've been

a great they've been a great partner of

ours. We've been a great partner of them

for 20 plus years. So, you know, they've

always been very helpful for us in being

able to expand our capacity, expand our

allocation, expand our our wafer

availability um in real time. And so,

working with them closely to ensure that

we've got a good understanding as to

kind of what the base case looks like

versus kind of what upside looks like.

and being having that ability to flex

with them. Uh substrates, you know, as

you guys all know, has been tight. Um

our customers have been willing to work

with us to find alternative substrates

that help ease some of those supply

constraints. We've got, you know, a

tier, you know, just best-in-class

operations team that have been on this

since the very early days. and so have

been ahead of making sure that as

substrate availability becomes tighter

and tighter, we've got alternatives that

we can scale with our customers. So,

we're we're confident that even as the

business more than doubles this year, um

we've got the capacity to support that.

>> Uh you you talked about these these ramp

ups that you've had obviously with

Nvidia. I think you have one coming up

with Google as well. Um can you tell us

a bit more about I mean it seems like

you're seeing a lot of growth on that

front here.

Should I be worried about concentration

risk I if in any event that you know we

start to see volumes start to to to fade

a little bit or that you know this ramp

up is is is actually going to be delayed

anyway.

>> Yeah. So a lot of these programs that

you're referring to, you know, coming in

later this year going into 27, driving

growth 27 and 28th as well, but these

are, you know, certainly some examples

that folks have talked about, but behind

the scenes, it's a much broader range of

customers and markets, uh, solutions

that we're supporting, right? Um, you

know, we've got a telco company for one

of the largest telco equipment companies

out there work using our b our boot

drives for their base stations that'll

start scaling in the back half of the

year. We've got virtually every

automaker around the world, whether

it's, you know, traditional internal

combustion engine, whether it's battery

electric, whether it's autonomous, what

have you, right? We have virtually every

automaker using our solutions as well.

So, you know, point being is that a lot

of this growing business is scaling

across a very broad range of customers.

Certainly, people like to talk about the

largest GPU company or the largest

search company and those are, you know,

big names to be associated with and

we're certainly very proud of the wins

we have there, but if you just kind of

double click a little bit beyond that,

you know, the the the breath and scale

of other customers is pretty fantastic

as well. So, it's a really diversified

business, diversified end markets, a

diver diversified customer base that we

have. And so, we're not um we're not

concerned that there's going to be an

outsized contribution from one customer

or one product or one vertical that's

going to have issues down the road. And

that's, you know, for our business,

having the diversification helps us kind

of work through the cyclicality of

certain end markets in any given period,

right? Having that wide range of end

markets helps smooth that out.

>> Yeah. And and certainly, I mean, as you

say, you're shifting beyond just the

consumer-driven side of things, but I

guess it is still a headwind for Silicon

Motion is, of course, um the potential

slowdown what we're seeing in the PC

market, smartphone market. I think

during the analyst call, you guys were

asked about the China uh outlook there.

What is the demand picture looking like

to you on that front here? And what what

do you think is going to be the outlook

moving forward?

Yeah, we're certainly seeing that,

right? I mean, I think certainly on

especially on the low end in China, it's

been an incredibly difficult market

there. Um, we don't really do a whole

lot on the low end. We're more

mainstream to premium side. So, we and

that's where a lot of the handset OEMs

are focusing more of their attention on

a low-end handset. The cost of memory,

the cost of DRAM and NAND far exceeds

kind of the affordability of that

product. So where you're seeing the

biggest hit in unit volume and demand is

on the low end. Mainstream premium ones

are still holding up a little bit better

and that's where we see um most of our

business and certainly the share gains

that we've talked about here is doing

wonders. We're going to be you know

almost doubling our smartphone related

business this year, consumer and IoT

related business this year. Our client

SSDs again similar dynamic. low end is

where you're seeing a lot more of the

pressure. We're not so much at the low

end. We're on the mainstream to higher

end. Um we're we you know first half

this year we've grown 40% yearon year

and certainly that's on in the face of a

very challenging PC market. And so

overall I think the story that we've

been talking about the foundation that

we've been building in share gains has

really be continued to pay dividends.

We're seeing flash makers being less

interested in the space, more interested

in outsourcing, more interested in

working with us, and that's helping

drive a lot of opportunities. That was

Jason Sai, CFO at Silicon Motion,

speaking with Bloomberg TV host David

Inglace and Ivon Man, bringing you their

conversation here on the Daybreak Asia

podcast.

Thanks for listening to today's episode

of the Bloomberg Daybreak Asia Edition

podcast. Each weekday we look at the

stories [music] shaping markets,

finance, and geopolitics in the

Asia-Pacific. You can find us on Apple,

Spotify, the Bloomberg Podcast YouTube

channel, or anywhere else you listen.

Join us again tomorrow for insight on

the market moves from Hong Kong to

Singapore [music]

and Australia. I'm Doug Krer, and this

is Bloomberg.

[music]