Mr Lambo

Options and derivatives

[Mr Lambo @0xmstr]

$TRT

[Mr Lambo @0xmstr]

Hi, I checked Trio-Tech International again today, including the company's current press-release archive and investor-relations filings.

And I think the absence of a new headline is actually telling us exactly what matters next.

As of September 9, Trio-Tech has not announced another AI-GPU order after the July 28 $3.8M order, and the FY2026 annual report has not yet appeared on its investor-relations page. The latest corporate announcement remains the August approval to transfer TRT to the Nasdaq Global Market.

So we're entering what I would call the proof phase of the TRT investment thesis.

The company has already proven that demand exists.

During fiscal Q2, Semiconductor Back-End revenue reached $12.4M, +113% year over year, driven by higher testing volumes. For the first six months, semiconductor revenue reached $23.8M, +88%.

And we know from the subsequent quarter that growth accelerated further.

We also know TRT has announced roughly $14.2M of orders since March for burn-in boards supporting a next-generation AI GPU platform. The company's official 2026 archive confirms the sequence of AI-related orders alongside its automotive semiconductor win, capital raise and Q3 results.

But here's the point I want everyone to understand:

The next $5M order would be good news. Margin expansion could be much bigger news.

Why?

Because TRT has already built a dramatically larger revenue base.

The problem is that gross margin hasn't followed it yet.

In fiscal Q2, gross margin was only 16%, compared with 26% one year earlier. Management attributed much of that compression to a higher proportion of lower-margin, high-volume testing services. (Trio-Tech)

That's the number I now care about most.

Think of $TRT as a machine that's suddenly processing twice as much material.

The first question was:

Can customers fill the machine?

The answer increasingly appears to be yes.

The second question is:

Can TRT make substantially more money from that volume?

That's what the next few quarters need to prove.

And there's another piece of the story I think deserves more attention.

$TRT isn't simply buying exposure to AI with a press release.

This company has been providing semiconductor reliability testing for more than 60 years. Its customers use $TRT to verify components for automotive electronics, computing, communications, industrial electronics and data-storage applications. (Trio-Tech)

AI is potentially adding a powerful new demand engine to an infrastructure that already existed.

That's very different from a tiny company suddenly adding “AI” to its corporate presentation.

$TRT's Q1 management commentary explicitly said the Semiconductor Back-End business had entered final testing services for next-generation high-performance AI devices for a leading AI chip manufacturer. (Trio-Tech)

That wording matters.

We're talking about actual semiconductor testing activity, not an aspirational AI strategy.

Now add another interesting piece.

Nasdaq trading is approaching.

$TRT's transfer to the Nasdaq Global Market is expected later this month.

Again, I don't believe changing exchanges creates fundamental value by itself.

But imagine the combination if the next annual results eventually show:

**higher semiconductor revenue

improving gross margin
continued AI orders
stronger earnings
Nasdaq visibility.**

That's when a tiny company can begin attracting investors who previously never looked at it.

And this is why I think patience matters enormously here.

We're not waiting for some mysterious miracle.

We're waiting for specific, measurable evidence.

My checklist for the FY2026 annual results is now extremely simple:

Revenue: Did the growth continue?

Gross margin: Is 16% beginning to recover?

Operating income: Is scale finally reaching the bottom line?

Cash flow: Is growth funding itself more effectively?

AI pipeline: Does management indicate continued demand beyond the ~$14.2M already announced?

Customer concentration: Is TRT becoming less dependent on one major customer?

If three or four of those improve together, the thesis becomes considerably stronger.

If revenue remains huge but margins and cash generation don't improve, then we need to reassess.

That's investing.

Not falling in love with the stock.

Testing the hypothesis against reality.

There are still real risks: concentrated customers, an unidentified AI-GPU customer, dilution from the roughly $10M April capital raise, lower-margin testing volumes and the inherent volatility of a microcap semiconductor business.

But TRT also has something that many speculative microcaps don't:

actual customers, actual orders and actual revenue growth.

That's why I remain interested.

The story has now moved beyond:

"Can TRT win AI-related business?"

It has.

The question has become:

"Can TRT turn its new scale into meaningful earnings?"

If the answer eventually becomes yes, that's when I think the mathematics of TRT can become genuinely powerful.

So don't be disappointed by a quiet couple of weeks.

Sometimes the most important period in an investment is the space between the exciting announcements and the financial statements that either validate them or expose them.

We're approaching that moment.

Orders → Revenue → Utilization → Margin → Earnings → Recognition.

$TRT has already moved through the beginning of that chain.

Now comes the part that could determine how valuable this company ultimately becomes.

$nvda $NVDS $sndk $nu $mu $te $be $BMNR $SBET $COHU $cohr $meta $trt

[Mr Lambo @0xmstr]

@JonkooTrades $TRT what do you think next earnings?

[Mr Lambo @0xmstr]

$ACMR
Friends, I think ACM Research is entering one of those periods where investors need to separate price action from operating reality.

As of September 3, $ACMR closed at $69.42, down roughly 14% over the prior week and more than 20% over three months. Yet the company’s latest investor-relations page still shows August 7 as the latest press-release date and August 21 as the latest investor presentation. In other words, I do not see a fresh company disclosure showing that the fundamental story has suddenly deteriorated.

That matters because the valuation has changed quite a lot.

At about $69, $ACMR's market capitalization is now approximately $4.8 billion, while enterprise value is closer to $3.7 billion because of ACM's substantial net cash position. The stock is trading around 32× trailing earnings and roughly 32× forward earnings based on current market data.

That is still not “deep value.”

But it is considerably more interesting than the valuation investors were being asked to pay when ACMR traded above $120 earlier this summer.

And importantly, the underlying growth numbers have not changed.

Q2 revenue was $292.9 million, up 36% year over year. ECP revenue grew 168%, advanced packaging grew 153%, and ACM shipped its 2,000th electroplating chamber. Management raised 2026 revenue guidance to $1.125–$1.175 billion, representing roughly 25%–30% annual growth.

So today we have an interesting divergence:

The valuation multiple is compressing while the business is still growing at more than 30%.

That does not automatically mean the stock is cheap. But it improves the equation.

One external development makes ACM's memory exposure more interesting

There has been an important development in the global memory industry.

SK hynix recently said it expects the current memory shortage to persist through 2030, driven by AI and HBM demand, while continuing major capacity investment. The company is also building its $4 billion advanced-packaging and R&D facility in Indiana, with HBM4E production targeted for 2029.

I want to be precise here:

This is not an ACM order announcement.

We should not turn an industry-capex story into fictional ACM revenue.

But it matters because ACM sells equipment into memory manufacturing and advanced packaging, two areas receiving enormous investment because of AI infrastructure.

The stronger and longer the memory-capex cycle becomes, the larger ACM's potential opportunity becomes—provided it continues winning qualifications.

That distinction is crucial.

A good industry does not guarantee a good investment.

But a growing equipment vendor operating inside a strong capex cycle has much better conditions for compounding than one fighting a shrinking market.

There is another fascinating dynamic developing in Korea and China

Reuters recently reported that Samsung and SK hynix have been evaluating Chinese semiconductor-equipment suppliers at their China-based fabs as a hedge against tighter U.S. equipment restrictions.

The reported testing involves another supplier, AMEC—not ACM Research.

But strategically, the story matters for ACM.

For years, one argument against Chinese or China-centered semiconductor-equipment suppliers was:

“Top global chipmakers will never seriously qualify these tools.”

That assumption is becoming harder to defend.

If geopolitical restrictions force global manufacturers to diversify equipment sources inside China, Chinese equipment ecosystems could gain credibility faster than investors previously expected.

Again, I am not claiming those AMEC evaluations translate directly into ACM revenue.

My interpretation is broader:

The barrier preventing Asian semiconductor manufacturers from evaluating alternative equipment suppliers appears to be weakening.

And ACM already has international ambitions.

The company has delivered multiple cleaning systems into a Singapore fab, has announced advanced-packaging orders from customers outside mainland China, and is preparing its Oregon operation to support further global expansion.

That is exactly the direction I want to see.

The balance sheet deserves more attention at today's price

At the end of Q2, ACM had approximately $1 billion of net cash.

Against a market cap below $5 billion, that means a meaningful portion of ACMR's valuation is backed by financial assets rather than simply expectations.

This gives management something extremely valuable during a volatile semiconductor cycle:

time.

ACM can keep investing in:

R&D, Oregon, global support infrastructure, customer evaluations and new products

without relying heavily on external capital.

That reduces financial risk even though business and geopolitical risks remain.

Now let's look at the numbers from a different angle

ACM generated $901 million of revenue in 2025.

The midpoint of current 2026 guidance is roughly $1.15 billion.

That would imply around 28% growth in one year.

Meanwhile, the current enterprise value is approximately $3.7 billion.

Very roughly, that puts the operating business at only a little above 3× this year's expected revenue.

That is a more useful way for me to think about ACMR than the headline trailing P/E, because Q2 GAAP earnings benefited from investment gains.

A 3×-ish enterprise-value-to-sales multiple is not automatically cheap either.

But for a semiconductor-equipment company growing around 25%–30%, with mid-40s gross margins and several emerging product categories, it starts to become genuinely interesting.

That is interpretation—not a price target.

There is also an important historical benchmark coming

Last year, ACM Shanghai disclosed backlog of approximately RMB 9.07 billion, or about $1.27 billion, as of September 29, 2025. That was 34.1% higher year over year.

For comparison, backlog had been roughly RMB 6.76 billion a year earlier.

This is why the next backlog disclosure matters so much.

A company approaching $1.2 billion in annual revenue with backlog around—or potentially above—one year's worth of sales has significant visibility.

But we should wait for the actual figure.

If the next backlog number grows strongly again, it would strengthen the argument that the recent stock correction is primarily valuation and sentiment-driven.

If backlog disappoints materially, then we would need to reconsider that interpretation.

That is the kind of evidence-driven approach I prefer.

And there are still serious risks

We should not become more bullish simply because the stock is cheaper.

The major risks remain.

China concentration is still significant.

U.S.-China export policy can change quickly.

New platforms such as PECVD, Track and panel-level packaging still have to convert evaluations into production orders.

Gross margin fell to 46.0% from 48.5% in Q2, so product mix and pricing need monitoring.

And although ACMR has fallen substantially, the stock is still up dramatically over the last twelve months. Current data shows market capitalization approximately 173% higher than one year ago.

So expectations have not disappeared.

My interpretation today

I think ACMR is becoming more attractive for a very simple reason:

the expectations built into the stock are falling faster than the company's growth rate.

At $120+, investors were paying a very large premium for the possibility that ACM could become a global multi-product equipment company.

Around $69, they are paying considerably less for the same possibility.

And since the peak, the latest company evidence still shows:

36% Q2 revenue growth
168% ECP growth
153% advanced-packaging growth
raised guidance
$1 billion net cash
continued international expansion
Those facts remain intact.
What we need now is the next layer of proof.
Not another analyst upgrade.
Not another green day in the stock.
I want to see:

backlog growth, international repeat orders, successful PECVD/Track qualifications, packaging production ramps and gross margins staying healthy.

If those arrive while ACMR remains around today's valuation, I think the investment case becomes considerably more compelling.

Because eventually stock prices have to reconcile with earnings power.

And right now, ACMR's share price is telling us investors have become much more cautious.

The business, at least based on the latest evidence available, is not sending the same message.

$nvda $NVDS $meta $ORCL $mrvl $mu $sndk $BE $TE $COHU $COHR $rddt $ACMR $AEHR $TRT

[Mr Lambo @0xmstr]

Friends, today's comparison between Trio-Tech International and $AEHR Test Systems is about something slightly different.

We've spent several editions looking at orders, AI exposure and backlog. Today I want to focus on what the market is actually demanding from each company at the current price.

As of September 1, $TRT closed around $9.86, giving it a market capitalization of roughly $100M. AEHR closed at $76.59, worth approximately $2.50B.

That means the market currently values $AEHR at roughly 25 times $TRT.

But here's where the comparison gets fascinating.

$TRT's trailing revenue is already approximately $58.4M. $AEHR generated only about $50M in FY2026.

So we're looking at two companies with surprisingly similar recent revenue bases — but valuations separated by approximately $2.4 billion.

Obviously, there's a reason for that.

The market isn't valuing $AEHR on $50M of revenue

Nobody buying $AEHR at $76 is really paying for the business AEHR was last year.

They're paying for the business investors believe AEHR is becoming.

And there's considerable evidence supporting that belief.

$AEHR's fiscal Q4 produced $18.8M of revenue and adjusted EPS of $0.11, both ahead of expectations. More importantly, bookings and backlog reached record levels.

Management has guided FY2027 revenue to approximately $130M–$150M.

So using the $140M midpoint, today's $2.50B market capitalization represents roughly:

17.9× forward sales.

Even subtracting $AEHR's substantial cash balance and looking at enterprise value, the multiple remains around 17× forward sales.

That's an extraordinary valuation.

But $AEHR is also attempting something extraordinary.

If it successfully converts its AI processor, silicon-photonics and power-semiconductor opportunities into a business growing well beyond $150M, today's valuation could ultimately prove justified.

The problem is that success is no longer optional.

It's embedded in the share price.

$TRT has almost the opposite problem

$TRT is trading at roughly:

$100M market cap / $58.4M trailing revenue = ~1.7× sales.

Its enterprise value is even lower, around $89M, reflecting the company's net-cash position.

So why doesn't the market pay more?

Because $TRT's profitability remains tiny.

Trailing operating income is only around $0.53M, while trailing net income is roughly $0.35M.

That's the entire debate in one sentence:

$AEHR is being valued on future profitability. $TRT is being punished for present profitability.

And I think that's exactly where the opportunity may lie.

$TRT's next earnings report matters much more than another AI press release

This is perhaps the biggest change in how I'm looking at $TRT now.

I don't need another announcement telling me that AI demand exists.

We already have enough evidence.

$TRT's trailing revenue has increased about 64%, reaching roughly $58M.

We already know the semiconductor segment has been growing dramatically.

We already know $TRT has received repeated Burn-In Board orders related to a next-generation AI GPU platform.

We already know management has taken approximately 104,000 additional square feet of Malaysian capacity.

The next question is no longer:

"Is $TRT growing?"

It clearly is.

The question is:

"When does that growth start falling meaningfully to the bottom line?"

That's the number I would watch above almost everything else in the upcoming results.

$AEHR has already demonstrated the economic model $TRT still needs to prove

$AEHR's advantage isn't merely technological.

It's economic.

When $AEHR sells a FOX-XP system, it isn't just selling a machine.

That installation can create future demand for proprietary WaferPak contactors, additional systems, automation and capacity expansion.

Its lead AI processor customer is already doing exactly that.

A February order alone was worth $14M and expanded the customer's installed FOX-XP production capacity.

Then $AEHR received its record $41M Sonoma production order from its lead hyperscale AI customer, with shipments beginning in FY2027. That customer is simultaneously developing a substantially higher-power next-generation AI accelerator. (Aehr)

And silicon photonics is developing into another important leg of the business.

AEHR has already won a major new silicon-photonics customer that ordered systems for both qualification and high-volume production, not merely R&D.

That's an excellent semiconductor equipment model.

$TRT simply isn't there yet.

But $TRT doesn't need $AEHR economics to work

This is where valuation becomes incredibly important.

Suppose $AEHR eventually produces $200M of revenue.

At today's ~$2.5B market capitalization, investors are still paying more than 12× that hypothetical future revenue.

Now consider $TRT.

$TRT could become a significantly more valuable stock without ever approaching $AEHR's margins or valuation multiple.

If $TRT eventually becomes a $100M-revenue company with only a 10% net margin, the earnings profile becomes completely different.
At $10M of earnings and a 30× multiple, we're talking about approximately $300M of equity value.
That's roughly three times today's capitalization.
And that's without giving TRT anything remotely resembling $AEHR's current sales multiple.
The next phase of $TRT's story is therefore about margins
This is why I think investors should stop obsessing about whether the next Burn-In Board announcement is $3M, $5M or $8M.
Those contracts matter.
But we're approaching the point where the income statement matters more.

And that's potentially important because $TRT is still extraordinarily small.
At approximately $100M, it's firmly a microcap.

$AEHR at $2.5B is already in a completely different institutional universe.
There's also an interesting psychological difference between the stocks now
AEHR has become famous.
Its market capitalization has increased roughly 235% over the past year, even after the stock pulled back from its recent highs.
Investors understand the narrative:
AI processors + extreme power density + burn-in + silicon photonics + FOX + Sonoma.

TRT remains comparatively obscure.

And despite its enormous move from last year's levels, its entire company is still worth less than $100–101M. Its market capitalization has increased more than 300% year over year, so we should also acknowledge that TRT is no longer the completely undiscovered $3 stock it once was.
That's important for us.
The easy rerating has already happened.
The next rerating needs earnings.
This is the key milestone I'm waiting for
Imagine TRT reports a future quarter where semiconductor revenue remains strong — but instead of gross margin hovering around the mid-teens, we suddenly see it moving toward 20%+.
Then imagine the following quarter confirms the trend.
That would tell us something extremely valuable:
the capacity expansion is starting to create operating leverage.

At that point, I'd become materially more confident in assigning TRT a higher long-term valuation.

Until then, it's still a thesis.

A very interesting thesis — but a thesis nonetheless.

What would make me more cautious?

Three things.

If semiconductor revenue begins slowing sharply before margins recover, that's a warning.

If the Malaysian expansion materially increases costs without producing sufficient incremental revenue, that's another.

And if TRT's AI-related business remains heavily dependent on one customer or one GPU platform, I would reduce the valuation multiple I'd be willing to assign the business.

Those risks are real.

AEHR faces customer concentration too, but its enormous order book gives us much greater near-term visibility.

That's why AEHR deserves the quality premium.

My conclusion today

AEHR remains the better company.

TRT remains potentially the better expectation mismatch.

And those aren't contradictory statements.

AEHR is worth approximately $2.5B because investors already believe it is becoming a highly profitable strategic supplier to the AI semiconductor ecosystem.

TRT is worth approximately $100M because investors still aren't convinced that its extraordinary semiconductor growth will translate into extraordinary profits.

So the next chapter of the TRT thesis isn't really about revenue anymore.

It's about proving this equation:

AI demand + expanded capacity + higher utilization = operating leverage.

If TRT proves that, I think the valuation gap with AEHR becomes increasingly difficult to ignore.

If it doesn't, AEHR's premium will remain completely justified.

That's what makes the next set of TRT results particularly important: we're approaching the point where the story needs to become earnings.

$COHR $ACMR $meta $BMNR $SBET $sndk $mu $ORCL $mrvl $rddt $d $be

[Mr Lambo @0xmstr]

$TRT
Guys, I checked Trio-Tech International again today, and there is something important I want to emphasize.

There is no new major AI order announcement since July 28, and there is still no FY2026 10-K.

I'm actually pleased to say that rather than manufacture another bullish headline, because this is exactly how we should approach TRT: follow the evidence, not the excitement.

And the evidence we already have is substantial.

TRT's fiscal Q3 revenue reached $16.5M, +124% year over year. Semiconductor Back-End Solutions generated $13.1M, +141%, while Industrial Electronics grew 76% to $3.4M.

Those aren't normal growth rates.

More importantly, the semiconductor growth was driven by higher reliability-testing volumes, with management specifically pointing to customers developing advanced AI computing and EV automotive-related chips.

Now consider the order sequence we already know:

March → $5.3M
May → +$2.5M
June → +$2.6M
July → +$3.8M

Approximately $14.2M of announced orders associated with a next-generation AI GPU platform.
For perspective, that's almost as much as TRT generated in total company revenue during its entire latest reported quarter.
That is why I'm still paying attention.

But today I want to introduce a concept that I think is essential for understanding where TRT could go next:

The difference between growth and scalable growth.

Anyone can grow revenue temporarily by spending aggressively.

The great businesses eventually reach a point where each additional dollar of revenue produces disproportionately more profit.

$TRT isn't there yet.

And that's precisely the opportunity—and the risk.

During fiscal Q2, for example, revenue reached $15.65M, but gross margin was only 16.0% versus 25.7% a year earlier. Operating income was just $97,000. (Stock Titan)

So despite spectacular revenue growth, the earnings machine hasn't fully switched on.

That's why I believe the next annual report is potentially the most important $TRT financial report we've seen in years.

I don't need another +100% revenue headline.

I want margin improvement.

Imagine $TRT at approximately today's revenue run-rate with a 16% gross margin.

Now imagine the same company eventually returning toward a 20–25% margin as utilization improves.

The difference flowing through operating profit could be enormous relative to $TRT's current earnings base.

That's what operating leverage can do to a small company.

And there's another encouraging piece of evidence buried inside the financials.

Through the latest reported nine-month period, $TRT's operating cash flow improved to approximately $2.3M. The company also acquired the remaining 50% of Trio-Tech Malaysia and subsequently raised approximately $10M of additional equity capital.

That gives management considerably more flexibility to expand.

But it also creates accountability.

They raised shareholders' money.

Now we need to see returns on that capital.

That's the deal.

The image above illustrates the less glamorous—but very important—part of $TRT's business.

This is Trio-Tech HTRB reliability-testing equipment.

The basic idea is beautiful in its simplicity: semiconductor components are deliberately exposed to extreme heat and electrical stress so weak devices fail before they're installed in expensive equipment.

Think about the economics.

A defective $2 chip inside a cheap consumer product is annoying.

A defective high-value semiconductor inside an AI accelerator, EV power system, aerospace platform or industrial system can be extraordinarily expensive.

As semiconductor complexity and value increase, reliability becomes more economically valuable.

That's why $TRT's 60+ years of reliability-testing experience potentially matters much more today than it did a decade ago. The company positions itself as a one-stop provider covering reliability-test equipment, solutions and testing services across automotive, computing, industrial, communications and other semiconductor markets.

And then we have the next near-term catalyst.

September 16 — Nasdaq

$TRT is scheduled to begin trading on the Nasdaq Global Market on September 16, following its voluntary transfer from NYSE American.

I want to keep expectations sensible here.

Nasdaq itself doesn't create earnings.

It doesn't create customers.

And it doesn't magically make $TRT worth twice as much.

But for a tiny technology company, improved visibility among technology-focused and institutional investors can become valuable if the fundamentals continue improving.

That's the key qualification.

So here's how I see $TRT today:

The demand question: increasingly answered.

The revenue-growth question: answered emphatically.

The capacity question: management is investing aggressively.

The margin question: still unresolved.

The earnings question: therefore still unresolved.

The market-recognition question: Nasdaq may help.

And that gives us a very clean investment thesis.

Orders → Revenue → Utilization → Margins → Earnings → Recognition

We're somewhere in the middle of that chain.

If $TRT gets stuck there, the investment may ultimately disappoint.

But if it completes the chain?

That's where things become genuinely interesting.

The biggest mistake we could make now would be assuming success is guaranteed simply because the first part of the thesis worked.

It isn't.

Customer concentration remains significant. The AI GPU customer still hasn't been publicly identified. Margins remain compressed. The April equity raise diluted shareholders. And microcap execution risk is real.

But the opposite mistake would be ignoring what has already changed.

A company this small has produced 124% quarterly revenue growth, repeated AI-related orders and significant semiconductor expansion—and is about to move onto Nasdaq.

That's unusual.

So I'm still very comfortable with the conclusion:

$TRT has earned the right to prove the next stage of the story.

The next annual results should tell us whether all this spectacular growth is finally beginning to produce spectacular economics.

That's what I'm waiting for.

Not another exciting press release.

Profit.

Because when a tiny company combines rapid growth with genuine operating leverage, that's when the mathematics can become extremely powerful.

And if $TRT shows us that transition in the coming quarters, I think the market will eventually have to look at this company very differently.

$nvda $TRT $NVDS $meta $ORCL $ACMR $AEHR $COHU $COHR $sndk $mu $wolf $SILC $aaoi

[Mr Lambo @0xmstr]

Friends, ACM Research closed Friday, August 28, at $74.42, down about 7.5% in one session. That looks dramatic—and after a drop like that, the natural reaction is to ask whether something fundamental has changed.

After reviewing ACM’s latest investor-relations releases, filings, recent analyst activity and Friday’s semiconductor-sector news, I do not see a new company-specific fundamental event that explains a deterioration in ACM’s business.

In fact, ACM’s investor-relations page still shows August 7 as its latest press-release date, with the newer August 21 investor presentation as its most recent major corporate communication.

Friday was also a weak session across parts of technology and semiconductors. Marvell fell more than 8% after investors focused on the timing of future AI-related revenue, while other semiconductor and optical names were dragged lower. The Nasdaq itself finished down around 0.5%.

So I think this is a good moment to talk about something investors often forget:

Price risk and business risk are not the same thing.

$ACMR can fall 8%, 15% or even 25% without ACM Research suddenly selling fewer tools.

And the latest operating evidence remains strong.

Q2 revenue was $292.9 million, up 36% year over year. More importantly, ECP revenue grew 168% and advanced-packaging revenue grew 153%. ACM shipped its 2,000th ECP chamber, demonstrating that this is becoming a real high-volume manufacturing franchise across logic, memory and 3D packaging—not simply a product sitting in development.

Management also raised full-year 2026 revenue guidance to $1.125–$1.175 billion, implying approximately 25%–30% annual growth. First-half orders increased about 105% year over year, according to the Q2 call summary, giving management meaningful visibility into the second half.

That combination matters:

strong current revenue + strong orders = better visibility into future revenue.

It doesn’t eliminate execution risk, but it is much more valuable than growth based only on optimistic forecasts.

Something new happened this week: Wall Street coverage became more bullish

On August 25, Daiwa Securities initiated coverage of $ACMR with a Buy rating and a $136 price target. That joined a growing group of positive analysts.

Current compiled consensus shows approximately:

8 Buy
1 Hold
0 Sell

with an average 12-month target around $117.75.

I never buy a stock simply because analysts have high targets.

Analysts can absolutely be wrong.

But new institutional coverage is still useful information because it suggests ACM is increasingly becoming a company that larger investors feel they need to understand.

That is another step in $ACMR’s evolution from obscure semiconductor-equipment name to increasingly mainstream growth company.

At Friday’s $74.42 close, that $117.75 consensus target represents roughly 58% potential upside—although obviously there is no guarantee the consensus will prove correct.

I think valuation just became more interesting again

This is where Friday’s decline actually changes something.

At around $74, $ACMR’s market capitalization has fallen to roughly the low-$5 billion range.

Yet ACM ended Q2 with approximately $1.0 billion of net cash.

So investors are effectively paying materially less for the operating business after adjusting for that cash.

And remember what management is building.

ACM’s product footprint now spans:

cleaning, electroplating, furnace, Track, PECVD, stress-free polishing, wafer-level packaging and panel-level packaging.

This is becoming a portfolio company rather than a single-product story.

The latest Tahoe development illustrates this very well.

ACM has expanded Ultra C Tahoe into a platform combining multiple wet-processing functions. It has already been adopted by multiple semiconductor manufacturers, while its monitor-wafer reclaim application is now running in volume production at customer facilities. ACM says the system can reduce sulfuric-acid consumption by as much as 75%.

Think about why that matters commercially.

If one machine can consolidate processes previously performed by separate tools, it potentially offers customers:

lower chemical consumption, fewer wafer transfers, shorter process cycles and higher manufacturing efficiency.

That gives ACM something much stronger than “our equipment works.”

It gives ACM a potential economic argument for customers to adopt its technology.

And semiconductor fabs ultimately care enormously about economics.

Advanced packaging is becoming one of my favorite parts of the ACM story

ACM’s new panel-level ECP platform has now received its first production order, for a 510 × 515 mm system from an existing advanced-packaging customer in mainland China.

At the same time, a new Asian panel manufacturer placed an evaluation order for a smaller version, with delivery expected during Q4 2026.

That combination is exactly what we want to see:

an existing customer progressing to production while a new customer begins evaluation.

This is how an equipment franchise develops.

Evaluation → qualification → production order → repeat orders → broader installed base.

Not every evaluation will convert.

But the pipeline gives ACM multiple shots at success.

PECVD could be another future engine

One product I think deserves more attention is ACM’s PECVD platform.

In April, ACM shipped its first SiCN PECVD system to a leading semiconductor manufacturer after meeting customer-defined specifications in ACM’s laboratory. The tool is now undergoing customer validation for advanced back-end-of-line and packaging applicationss

Today that contributes very little compared with the established cleaning business.

But that is precisely why it is interesting.

If PECVD becomes another meaningful revenue category, ACM does not need its existing businesses to carry all future growth.

This is why product diversification can be so powerful.

There is still one enormous question

International expansion.

ACM has already delivered multiple 300mm cleaning systems to a foundry facility in Singapore, its first deployment to a Singapore fab.

Management is also investing in international infrastructure and highlighting customer evaluations beyond mainland China.

This is the part of the thesis I would watch most carefully over the next 12–24 months.

Because successful international expansion could solve two problems simultaneously.

It would create another revenue growth engine.

And it could reduce the geopolitical discount investors place on ACMR.

That second effect could be extremely important.

But we should not ignore the risks

$ACMR has real risks.

First is China exposure and the possibility of tighter U.S. export restrictions.

Second is valuation. Even after the recent decline, investors are still paying a growth-company multiple, meaning a disappointing quarter can produce violent corrections.

Third is margin pressure. Q2 gross margin was 46.0%, down from the prior-year period, as the business mix changes and component costs remain important.

Fourth is execution risk.

Track, PECVD, panel-level packaging and international evaluations are opportunities—not guaranteed future revenue streams.

That distinction matters.

So how do I interpret Friday’s decline?

I see it very differently from a company announcing that orders have collapsed.

If ACM had cut revenue guidance, lost a major customer, suffered failed qualifications or disclosed export restrictions that materially damaged future sales, then a 7.5% decline would deserve a very different analysis.

But that isn’t what happened Friday based on the information currently available.

Instead, we have a company that recently reported:

Revenue +36%
ECP +168%
Advanced packaging +153%
Orders +105% in H1
Guidance raised
~$1B net cash

while its share price has fallen from a 52-week high of $127.19 to $74.42.

That does not automatically make the stock cheap.

But it certainly makes the risk/reward more interesting than it was at $120+.

And this is the lesson I would give any friend investing in growth stocks:

Great investments rarely travel in straight lines.

Sometimes the business keeps improving while the stock corrects because expectations, positioning or the broader sector change.

Those periods can feel uncomfortable.

But they are exactly when we need to return to fundamentals.

For $ACMR, I would currently focus on five things:

orders, margins, international qualifications, ECP/advanced-packaging growth, and whether Track/PECVD become real revenue contributors.

If those continue moving in the right direction, I am much less concerned about whether $ACMR trades at $74, $80 or $90 next month.

Because the bigger thesis remains this:

ACM is attempting to evolve from a successful specialist into a multi-product global semiconductor-equipment company.

That transformation is far from guaranteed.

But based on the latest evidence, it is still progressing.

And after the recent correction, the market is asking investors to pay considerably less to participate in that possibility.

$nvda $NVDS $mu $sndk $COHU $COHR $ACMR $TRT $BMNR bitcoin:native $eth $meta $ORCL

[Mr Lambo @0xmstr]

Friends, after reviewing ACM Research’s latest filings, August 21 investor presentation, and current valuation, I think the $ACMR story has reached an important stage.

The company has already proved that it can grow fast.

Now it has to prove something harder:

Can ACM become a global semiconductor equipment company rather than mainly a very successful China-focused one?

That distinction could determine what kind of valuation $ACMR deserves over the next several years.

The latest operating numbers remain strong. Q2 revenue reached $292.9 million, up 36% year over year, while management raised full-year 2026 revenue guidance to $1.125–$1.175 billion. ACM’s investor-relations page still shows the August 7 results as its latest earnings release, with the newer August 21 investor presentation now the freshest major corporate update.

But what I find more interesting today is what management is saying about the destination.

ACM has discussed a long-term internal revenue objective of approximately $4 billion, with roughly $1.6 billion potentially coming from overseas markets and $2.4 billion from mainland China. Management has specifically highlighted expansion opportunities across South Korea, Singapore, Malaysia, Vietnam and Thailand.

Pause on that for a second.

If ACM eventually generates $1.5–$1.6 billion outside China, then international revenue alone would be larger than the entire company is today.

That is not a forecast.

It is not something investors should simply assume will happen.

But it tells us where the biggest optionality in ACMR lies.

Why international wins could change the way the market values ACM

One of the biggest reasons investors have historically discounted ACMR is obvious: geopolitical and regulatory exposure.

ACM’s 10-Q continues to detail meaningful risks related to export controls, international trade rules, China exposure and the company’s corporate structure. Those risks are real, and no bullish thesis should pretend otherwise.

But imagine the other side of that equation.

If customers in Korea, Singapore, the U.S. or Southeast Asia begin qualifying ACM equipment in meaningful production environments, then the perception of the company could gradually change from:

“Chinese semiconductor-equipment growth story”

to:

“global equipment challenger with competitive technology.”

Those are two very different investment narratives.

And global qualification would tell us something deeper than revenue diversification.

It would help validate the technology itself.

A major semiconductor manufacturer outside ACM's traditional market does not install production equipment as a favor. Qualification requires process performance, reliability, yield, service, cost advantages and confidence that the supplier can support the tool for years.

That is why I believe international production wins are more valuable than their initial revenue contribution might suggest.

That is my interpretation, not management guidance.

The product portfolio is already becoming broader

ACM is no longer dependent purely on wafer cleaning.

Its current product portfolio includes wet processing, electrochemical plating, thermal processing/furnace systems, Track, PECVD, stress-free polishing and advanced packaging applications. The company’s own website now reflects that increasingly broad process footprint.

This matters because semiconductor equipment economics can be extremely attractive once a supplier establishes trust.

The same customer that originally qualified ACM for cleaning can potentially evaluate:

ECP → furnace → Track → PECVD → advanced packaging.

Every additional qualified process increases the amount of customer capital expenditure that ACM can theoretically address.

And we already have evidence that diversification is working.

Q2 growth was particularly strong in electroplating and advanced packaging, while ACM has also continued to announce new production and evaluation activity for its panel-level packaging tools.

So the question is gradually changing from:

“Does ACM have a good cleaning technology?”

to:

“How many equipment categories can ACM successfully penetrate?”

That is a much bigger opportunity.

The balance sheet gives management room to try

Another reason I remain constructive is financial strength.

ACM exited Q2 with substantial liquidity, and recent valuation data puts enterprise value materially below market capitalization because of the company’s large net cash position. As of August 25, ACMR traded around $79.44, with a market capitalization near $5.5 billion and enterprise value around $4.4 billion.

That cash matters.

It allows ACM to invest aggressively in:

R&D, demo centers, production infrastructure, overseas service teams and customer support.

A weaker balance sheet would make international expansion much more dangerous.

ACM can pursue it from a position of strength.

But valuation now requires discipline

At roughly $79, ACMR is trading around 37–38× trailing earnings depending on the data source and timing. Yahoo Finance recently showed a trailing P/E around 36.6 and a forward multiple in the high 30s.

So investors are no longer being paid simply to discover an ignored company.

A lot of optimism is already embedded in the price.

That is why I would not argue:

“ $ACMR is cheap, therefore it must go higher.”

I would argue something different:

$ACMR can still become much more valuable if earnings eventually catch up with the scale of the opportunity.

There is a big difference.

At these multiples, execution has to remain strong.

One insider detail worth understanding

There have also been recent insider-selling filings, and investors should understand the context rather than either dismissing them or panicking.

ACM’s latest 10-Q states that CFO Mark McKechnie adopted a Rule 10b5-1 trading plan in May allowing the exercise and sale of up to 105,199 shares, with sales beginning August 20.

More importantly for the next few months, CEO Hui Wang adopted his own Rule 10b5-1 plan in June allowing exercise and sale of up to 270,000 shares, beginning September 8, 2026.

This is worth knowing because future CEO sales could generate scary-looking headlines.

But these are pre-arranged trading plans, not necessarily spontaneous signals that management suddenly believes the business is deteriorating.

I would still monitor the selling, especially its scale and frequency. But context matters.

What would make me more bullish from here?

Not another 5% stock move.

Not another analyst price target.

I want evidence that the business itself is de-risking.

Specifically:

meaningful repeat orders outside China, successful Track and PECVD qualifications, continued ECP/packaging momentum, sustained operating profitability and healthy gross margins.

If those pieces arrive together, ACM's addressable market starts becoming much more believable.

And then the $4 billion long-term ambition stops looking like a PowerPoint number and starts looking like an actual strategic pathway.

What could go wrong?

We should keep four risks front and center.

Geopolitics remains number one. Export-control changes can be sudden and difficult to model. Customer concentration remains meaningful. Semiconductor equipment acceptance can create lumpy quarterly revenue. And valuation leaves less room for disappointment after $ACMR’s extraordinary run over the past year.

The stock has gained dramatically even after its recent pullback; Yahoo Finance showed ACMR up roughly 99% year to date as of August 20.

That means volatility should be expected.

But volatility and business deterioration are not the same thing.

My interpretation today

I think $ACMR is entering what I would call the validation phase.

Phase one was proving that ACM could build competitive technology.

Phase two was proving it could scale revenue in China.

Phase three — the one we are entering now — is proving those products can win internationally and across multiple equipment categories.

That is the phase that could create the greatest change in how investors perceive the company.

So when friends ask me what I am watching now, my answer is simple:

Not whether $ACMR goes from $79 to $85 next week.

I am watching whether ACM can go from roughly a $1.1 billion semiconductor-equipment company today toward a diversified global platform over the next several years.

If management can execute that transformation, the upside in the business could be far more important than the short-term movement in the stock.

That outcome is not guaranteed.

But the pathway is becoming clearer.

$nvda $nvds $meta $googl $cohr $cohu $orcl $bmnr $btc $eth $wolf $mram $dram $sndk $mu $be $te $aaoi $silc $ACMR

[Mr Lambo @0xmstr]

Guys, today's update on Trio-Tech International is a little different.

There hasn't been another major commercial announcement since the $3.8M AI-GPU order on July 28, and I don't want to manufacture excitement where there isn't new information. Trio-Tech's own 2026 news archive still shows the August 19 Nasdaq approval as its latest announcement.

But something interesting has quietly appeared around the company:

institutional ownership is beginning to deserve our attention.

Recent regulatory filings show that quantitative investment firm Renaissance Technologies increased its $TRT position by 162,848 shares during Q1, taking its reported holding to 325,526 shares, or roughly 3.6% of the company at the time of the filing.

That does not mean Renaissance has suddenly endorsed our investment thesis. Quant funds buy stocks for many reasons, and institutional filings are backward-looking.

But it's worth putting beside another development:

$TRT is moving to Nasdaq.

On August 19, Trio-Tech confirmed that its shares have been approved for the Nasdaq Global Market, with trading expected to begin on September 16, 2026, while NYSE American trading is scheduled to end after September 15. The ticker remains TRT.

Management itself explained why this matters: it believes Nasdaq better aligns TRT's public-market identity with its technology business and growth strategy.

I think we need to understand this development correctly.

Nasdaq doesn't make $TRT more valuable.

But Nasdaq can potentially make $TRT more visible.

And visibility becomes much more interesting when the underlying company is simultaneously changing.

Remember the fundamental numbers.

Fiscal Q3 revenue reached approximately $16.5M, an extraordinary 124% increase year over year.

Meanwhile, $TRT has announced approximately $14.2M of orders since March associated with burn-in boards for a next-generation AI GPU platform.

And management raised approximately $10M of fresh capital in April specifically while expanding capacity around AI and automotive semiconductor opportunities.

So look at the sequence developing:

AI demand → repeat orders → revenue growth → capacity investment → Nasdaq → potentially broader investor awareness.

That's becoming a much more coherent story.

And the image above illustrates why I continue to like TRT's position in the semiconductor ecosystem.

This is Trio-Tech's HTRB reliability-testing equipment.

The company isn't trying to design the next GPU.

It's helping ensure increasingly sophisticated semiconductor devices survive the extreme electrical and thermal conditions they're expected to operate under.

As chips become more powerful, expensive and mission-critical, failure becomes more expensive too.

That makes reliability testing an interesting picks-and-shovels business behind AI, automotive and power semiconductors.

But now comes the important part.

I don't think orders are the biggest question anymore.

We've already seen evidence of demand.

We've already seen spectacular revenue growth.

What TRT needs to prove next is:

profitability.

Revenue growing 100%+ is exciting.

Revenue growing 100%+ while gross margin eventually recovers is potentially transformative.

Those are two very different businesses.

TRT's expansion has pressured margins, and that's why the forthcoming FY2026 annual report matters enormously.

As of August 26, Trio-Tech's investor-relations page still lists the May 14 Q3 10-Q as its latest quarterly report. The FY2026 annual filing has not yet appeared there.

When it arrives, I'm going straight past the headline revenue number.

I want to know:

Did Q4 maintain the semiconductor growth trajectory?

Did gross margin finally begin recovering?

Are the AI orders translating into revenue efficiently?

How much cash is the expansion consuming?

What does management say about the next wave of AI, CPU and automotive demand?

Because if we eventually get high revenue growth + improving margins, $TRT's earnings can potentially grow much faster than its sales.

That's operating leverage.

And that's where small companies can become very interesting investments.

There are still important risks.

$TRT's customer concentration remains substantial.

The AI-GPU customer is still not publicly identified, so we should continue ignoring internet speculation claiming it must be Nvidia, AMD or another specific company.

The April capital raise diluted shareholders.

And TRT remains a small company where one large customer or program can materially change quarterly results.

Those aren't reasons to abandon the thesis.

They're the variables we need to monitor to make sure we're right.

That's an important distinction.

I don't want us to become emotionally attached to $TRT .

I want us to become attached to the evidence.

And right now, the evidence says something genuinely unusual has happened to this company during FY2026.

The next annual report should tell us whether that transformation is beginning to reach the bottom line.

That's the moment I'm waiting for.

Not $12.

Not $15.

Not $20.

Margins and earnings.

If those begin following the revenue trajectory, then we can start discussing what a fundamentally different TRT might actually be worth.

Until then, patience.

The best small-cap stories rarely move in a straight line.

But when business growth, institutional awareness and market visibility begin improving simultaneously, I pay very close attention.

And right now, $TRT has earned that attention.

Trio-Tech Investor Relations · Official Nasdaq listing announcement

$nvda $mu $sndk $cohu $mrvl $cohr $be $te $Orcl $meta $googl $acmr $aehr $bmnr $sbet

[Mr Lambo @0xmstr]

Friends, this time I want to focus on one of the biggest differences between Trio-Tech International (TRT) and Aehr Test Systems (AEHR):

$AEHR gives investors visibility. $TRT gives investors optionality.

Those are not the same thing, and the market prices them very differently.

Aehr finished fiscal 2026 with $80.6 million of reported backlog and roughly $100.6 million of effective backlog, while management guided fiscal 2027 revenue to $130–150 million. Then, on August 12, AEHR announced another $22 million follow-on production order from its lead AI processor customer for wafer-level burn-in systems.

That is excellent visibility.

For a semiconductor equipment company, this is close to the ideal setup: a large installed base, repeat production orders, proprietary hardware, and customers scaling real programs rather than just evaluating technology.

$TRT does not have that level of disclosed backlog visibility today.

And this is precisely why $AEHR deserves a premium valuation.

AEHR is becoming easier to underwrite

A year ago, AEHR still carried some uncertainty around the pace of SiC demand and whether its diversification into AI processors, silicon photonics and other markets would become material.

Now we have harder evidence.

Aehr reported record quarterly bookings of $60.7 million, effective backlog above $100 million, and management expects FY2027 revenue to be 2.6x to 3.0x FY2026 revenue.

Then the company followed that with the $22 million AI processor order.

That matters because repeat orders are much more valuable than a first design win.

A first order tells you the customer is interested.

A follow-on production order tells you the customer is using the equipment and expanding capacity.

That is one of the strongest signals in semiconductor equipment.

AEHR also received a follow-on order from its lead silicon photonics customer in August, supporting continued expansion of next-generation photonic integrated circuit production.

So Aehr is increasingly building a pattern:

qualification → production → expansion → repeat order.

That is exactly the pattern investors want to see.

This is where TRT is still less proven

TRT's latest reported results were spectacular on the top line.

Fiscal Q3 revenue reached $16.5 million, up 124% year over year, while Semiconductor Back-End Solutions revenue reached $13.1 million, up 141%.

But TRT's disclosure model is different.

We don't have a clean $100 million backlog number that lets us map future revenue with the same confidence.

So TRT requires more interpretation.

That makes it riskier.

But also potentially more interesting.

Because the market tends to pay up once visibility becomes obvious.

The biggest gains in microcaps often occur between "something may be happening" and "the numbers now prove it."

TRT appears to be somewhere in that middle zone.

The most important TRT number may not be revenue

For me, the number that deserves the most attention right now is actually gross margin.

TRT's Q3 gross margin was only 16%, down from 27% a year earlier, because the company's growth has been driven heavily by lower-margin, high-volume testing services.

This is the central weakness in the TRT thesis.

Revenue can double and shareholders can still be disappointed if incremental revenue generates almost no incremental profit.

And we can see that tension clearly.

For the six months ended December 2025, TRT's Semiconductor Back-End Solutions revenue increased from $12.7 million to $23.8 million, yet segment operating income fell from $274,000 to only $88,000.

That is not operating leverage.

Not yet.

So if someone tells you TRT is simply "AEHR at a tiny valuation," I think that's incomplete.

AEHR has already demonstrated much better economics.

TRT has demonstrated demand.

Now it has to prove that demand can become profit.

But TRT's current expansion tells us management expects more

TRT leased approximately 104,000 square feet of additional space in Penang, Malaysia, with the lease effective June 1, 2026. (Comisión de Bolsa y Valores)

For a microcap company, that is a meaningful expansion.

TRT also raised approximately $10 million gross in April through the issuance of roughly 1.05 million shares at $9.50.

The bear interpretation is straightforward:

$TRT diluted shareholders and expanded capacity while margins were weak.

That's a legitimate concern.

But the bull interpretation is equally important:

Management is preparing for a volume level that the old infrastructure could not support.

If the new capacity fills, the economics may look much better once utilization improves.

If it doesn't, the company will have added fixed costs and dilution without enough return.

That is why TRT remains a higher-risk investment than AEHR.

AEHR's valuation is now the opposite problem

The market has already recognized AEHR's success.

AEHR closed August 21 at approximately $101.73 after trading as high as $147.40 earlier in August.

With roughly 32.5 million shares outstanding, even around $100 per share the company is still valued in the multi-billion-dollar range.

Compare that with $130–150 million of expected FY2027 revenue.

AEHR is still trading at a very substantial forward revenue multiple.

That means investors are paying today for many years of growth.

Again, that may prove justified.

But AEHR now has a different risk:

execution may be excellent and the stock can still disappoint if expectations are even higher.

TRT faces almost the reverse.

Expectations remain comparatively low.

This is the "cost of certainty"

Investors love certainty.

But certainty is expensive.

AEHR's $100+ million effective backlog gives us confidence.

Its repeat AI orders give us confidence.

Its silicon photonics expansion gives us confidence.

Its 18%–22% targeted non-GAAP net margin gives us confidence.

And the stock price reflects that confidence.

TRT offers much less certainty.

But uncertainty is exactly why the potential rerating can be larger.

If TRT eventually reports:

sustained semiconductor revenue above today's run rate,
improving utilization,
gross-margin recovery,
repeat AI-related orders,
and stronger operating income,

then the market won't simply be reacting to higher revenue.

It will be reacting to greater certainty that the business model is working.

That usually deserves a higher multiple.

One risk with TRT deserves special attention

TRT's customer concentration has become significant.

For the six months ended December 31, 2025, one customer represented 41.2% of total revenue, while two others represented 12.9% and 10.3%.

That is substantial.

So some of TRT's exceptional growth may be tied to a relatively small number of customers.

This is why I want to see more than just revenue growth.

I want to see:

customer diversification + repeat orders + margin improvement.

That combination would dramatically strengthen the thesis.

Until then, TRT should be treated as a concentrated execution story, not a fully proven compounder.

The investment comparison is becoming clearer

If I wanted the business with the strongest current evidence, I would choose AEHR.

It has proprietary systems, much stronger margins, enormous backlog relative to historical revenue, and repeat orders from AI and silicon photonics customers.

If I wanted the business where a relatively small improvement in perception could change the valuation dramatically, I would choose TRT.

That distinction matters.

AEHR is increasingly a question of:

How big can this already-successful platform become?

TRT is still a question of:

Can this rapidly growing semiconductor operation become structurally profitable?

And because TRT starts from such a small base, the reward for answering that second question positively could be substantial.

My current view

AEHR remains the superior company today.

I would not pretend otherwise.

But investing isn't simply about choosing the best company.

It's about comparing quality, expectations and price.

AEHR gives us much greater certainty — but we pay heavily for it.

$TRT gives us much less certainty — but the market is also asking us to pay far less for the possibility that the business is entering a structurally higher growth phase.

That is why I still find TRT so compelling as a long-term asymmetric idea.

The real catalyst is not another AI headline.

The real catalyst is the moment the market can say:

"TRT's growth is repeatable, diversified and profitable."

If that moment arrives, the valuation conversation changes completely.

And that is the part of the story I think is still underappreciated today.

[Mr Lambo @0xmstr]

Friends, $ACMR Research published a new investor presentation on August 21, and for me this is more important than another day of stock-price volatility because management put much more structure around what the company is actually trying to become.

The headline is ambitious: ACM now lays out an internal long-term target of more than $4 billion in annual revenue. That is not formal guidance, and the company explicitly says it is an internal planning target rather than a forecast. But the framework behind it is worth studying. ACM estimates that its current portfolio addresses roughly a $22 billion serviceable market, spanning cleaning, plating, furnace, PECVD, track and advanced packaging.

What makes the roadmap interesting is how management divides that potential. The target assumes roughly $2.5 billion from mainland China and approximately $1.5 billion from the rest of the world. The non-China piece includes internal share targets of about 15% in cleaning and ECP, 10% in furnace, and 8% in PECVD and track. Those numbers are aspirations, not achievements—but they tell us exactly where management believes the next major leg of growth must come from.

And that leads to what I think is the most important point today:

ACMR’s future valuation may depend less on China growth and more on whether international qualification becomes repeat production business.

The new presentation says ACM currently has multiple evaluations underway with major customers in the U.S., Europe and Asia, while it is expanding sales and service teams across the U.S., Europe, Korea and Southeast Asia. It also highlights multiple advanced-packaging orders from customers in North America and Asia.

That is encouraging because international adoption has always been the missing proof point in the ACMR story.

The company also reiterated that its Oregon facility is expected to begin operations in the second half of 2026, with the Q2 summary specifically pointing to a planned Q4 opening. The site includes production capability and a clean-room environment intended to support global customers and advanced R&D.

Again, this does not guarantee major U.S. revenue.

But strategically, it matters. Semiconductor equipment vendors win trust through local process support, demonstrations, field service and qualification work. ACM is building the infrastructure required to compete on those terms.

There is another part of the new presentation that deserves attention: product breadth.

Management says ACM now addresses approximately 95% of cleaning process steps, but cleaning is increasingly only one part of the story. The company continues to highlight ECP, furnace, Track, PECVD and new panel-level packaging systems. It also says Track and PECVD have a “solid evaluation pipeline.”

The revenue mix is already showing this transition. In the first half of 2025, cleaning represented about 73% of revenue. In the first half of 2026, it was only 49%. ECP, furnace and other technologies rose to 41%, while advanced packaging and other revenue reached 11%.

That is a significant structural change.

ACM is not simply selling more of the same tool.

It is increasing the amount of semiconductor capital spending it can potentially capture from each customer.

And the operating performance continues to support that expansion. Q2 revenue was $292.9 million, up 36%, while GAAP operating income increased 56.9% to $49.7 million. Non-GAAP operating income reached $56.3 million, or 19.2% of revenue. Non-GAAP EPS was $0.61 versus $0.55 a year earlier.

The caveat remains gross margin. GAAP gross margin declined to 46.0% from 48.5% a year earlier. That is still within ACM’s historical target range, but it deserves watching as newer equipment categories become a larger part of sales.

The balance sheet is another source of strength. ACM’s August presentation shows cash and interest-bearing time deposits rising to roughly $1.36 billion at the end of Q2, giving management considerable flexibility to fund international expansion, R&D and capacity.

Now, one fresh development that we should not ignore: insider selling.

ACMR’s SEC filing page shows new Form 4 and Form 144 filings on August 18–21. CFO Mark McKechnie sold roughly $4.26 million of shares on August 20 after exercising options; importantly, the transactions were conducted under a Rule 10b5-1 plan adopted in May, meaning they were prearranged rather than an impulsive decision after the latest quarter. Director Haiping Dun also exercised and sold shares under a trading plan earlier in the week.

I would classify that as something to monitor, not automatically a bearish signal.

Executives sell stock for many reasons, especially after large gains and option exercises. What would concern me more would be repeated discretionary selling by several insiders combined with weakening fundamentals. We do not have that evidence today.

The stock itself remains volatile. ACMR traded roughly around $78–79 at the end of August 21, after touching an all-time high above $127 in June.

At roughly that price, my interpretation is that ACMR is no longer obviously cheap, but it also does not require the $4 billion target to succeed in order to justify further upside.

The crucial question is whether revenue can continue compounding at a strong rate while operating margins remain healthy and international sales gradually become material.

So this is how I would explain ACMR to a friend today:

The $4 billion target is not the thesis. The pathway toward it is.

Management is trying to turn ACM from a largely China-centered cleaning specialist into a global, multi-product semiconductor equipment platform.

We now have real evidence of the first transformation: cleaning has fallen from 73% to 49% of first-half revenue as other businesses expand.

We have early evidence of the second: international evaluations, North American and Asian packaging orders, a Singapore shipment and the Oregon facility.

What we still need is the strongest evidence of all:

repeat, meaningful production revenue outside mainland China.

If that arrives, I think the market will have to evaluate ACMR very differently.

That is why I remain constructive—but disciplined.

The exciting part of this story is not a stock chart or a $4 billion slide.

It is the possibility that ACM is quietly building several new revenue engines at the same time.

And if even a few of those engines reach scale, the company several years from now could look very different from the one producing roughly $1 billion of revenue today.

[Mr Lambo @0xmstr]

Guys, we finally have a genuinely new development at Trio-Tech International — and I think it deserves more attention than a normal corporate announcement.

On August 19, Trio-Tech announced that it has been approved to move its stock listing from NYSE American to the Nasdaq Global Market. Trading on NYSE American is expected to finish on September 15, with TRT beginning trading on Nasdaq on approximately September 16, 2026, keeping the same $TRT ticker.

First, an important clarification: the SEC filing uses Item 3.01, whose title includes the word “delisting.” This is not TRT being kicked off an exchange. It is a voluntary exchange transfer after Nasdaq approved the company for listing on its Global Market.

Why do I like this?

Because Nasdaq is the natural home of technology and semiconductor companies. Management explicitly said it believes the move can broaden awareness among institutional investors, improve TRT’s visibility and encourage greater market participation over time.

Will moving to Nasdaq suddenly double TRT’s value?

Of course not.

But investing in microcaps is often about watching several small pieces fall into place.

And look at the sequence developing here.

TRT entered FY2026 with Q1 revenue growth of 58%. Q2 accelerated to 82%. Then Q3 revenue reached approximately $16.5 million, up 124% year over year.

At the same time, the company has accumulated approximately $14.2 million of announced orders since March for burn-in boards supporting a next-generation AI GPU platform. The most recent was another $3.8 million order announced July 28.

And management isn’t sitting still waiting for demand.

TRT is expanding testing and manufacturing capability while explicitly identifying AI and automotive semiconductor applications as high-growth end markets. In announcing the Nasdaq move, CEO S.W. Yong again said both Semiconductor Back-End Solutions and Industrial Electronics are growing rapidly.

Now combine those developments:

AI orders → rapidly rising revenue → capacity expansion → stronger capital-markets profile.

That’s a much more interesting sequence than simply saying, “TRT is an AI stock.”

And there’s another subtle point.

TRT remains tiny. Recent market data puts its market capitalization at only around $115 million.

At that size, institutional awareness matters.

Many professional investors will never spend serious time researching an obscure NYSE American microcap. A Nasdaq Global Market listing doesn’t guarantee institutional buying, but it potentially puts TRT in front of a larger technology-focused investment audience.

That’s why I see this move as strategically positive rather than fundamentally transformative on its own.

The fundamentals still have to do the heavy lifting.

And that’s where we need to remain disciplined.

TRT’s biggest unresolved problem is still profitability.

The business has produced spectacular revenue growth, but margins have been compressed during this expansion phase. Current trailing metrics still show operating margins around only 1%, which tells us very clearly that revenue growth hasn’t yet translated into the earnings power we ultimately need.

So the bull case isn’t:

“TRT moved to Nasdaq, therefore the stock goes up.”

That’s superficial thinking.

The real bull case is:

If TRT can maintain this much larger semiconductor revenue base, fill the capacity it is building and recover margins, earnings could grow dramatically faster than revenue from today’s very low base.

That’s operating leverage.

And THAT is the part of the story that could eventually change TRT’s valuation.

There are still meaningful risks.

Customer concentration remains high. The AI-GPU customer remains undisclosed. The April equity raise created dilution. Margins need to improve. And because TRT is such a small company, losing one significant program could materially affect results.

We should never hide those facts simply because we own the shares.

But equally, we shouldn’t ignore what’s happening.

A company that generated only $36.5M of revenue in all of FY2025 has already produced a dramatically different revenue trajectory during FY2026.

It has secured repeated AI-related orders.

It is expanding semiconductor capacity.

Management continues talking about AI, CPUs, EV and automotive semiconductor demand.

And now it has secured approval to move onto the Nasdaq Global Market.

None of those facts individually makes TRT a great investment.

Together, however, they’re beginning to describe a company moving into a different stage of its life.

The next piece is the one I’m most excited to see: FY2026 annual results.

TRT’s fiscal year ended June 30, and as of August 23 the company’s investor-relations page still shows the May Q3 filing as its latest quarterly financial report and FY2025 as its latest annual report.

When FY2026 numbers arrive, forget the first five minutes of stock-price movement.

Look at Q4 semiconductor revenue, gross margin, operating income, cash generation and management’s comments about the AI pipeline.

If revenue remains strong and margins finally begin moving upward, that will be far more important to me than the Nasdaq announcement.

Because then we could potentially have something very powerful happening simultaneously:

A rapidly growing semiconductor business + AI exposure + improving profitability + increased investor visibility.

That’s the combination I’m waiting for.

So my verdict after this week’s news?

The TRT thesis has strengthened again — modestly, but genuinely.

Nasdaq isn’t the destination.

It’s another piece of infrastructure around a business that now has to prove just how big it can become.

Orders → Revenue → Utilization → Margins → Earnings → Recognition.

The first stages are already happening.

Now let’s see whether TRT can complete the chain.

$nvds $nvda $mu $sndk $cohr $cohu $mstr $bmnr $bmnp $mstx $acmr $wolf $meta $orcl

[Mr Lambo @0xmstr]

$TRT vs. AEHR — Part 4: The Question Nobody Should Ignore — Customer Concentration

Friends, today I want to look at a risk that can easily get lost when semiconductor stocks are growing quickly: who is actually generating that growth?

This is particularly important when comparing Trio-Tech International with Aehr Test Systems, because both companies can be transformed by a handful of large programs.

And strangely enough, I think this comparison reveals both AEHR’s greatest strength and one reason TRT’s tiny size could become an advantage.

AEHR: spectacular visibility, but concentrated visibility

AEHR enters fiscal 2027 in an enviable position.

It finished FY2026 with only $50.0M of revenue, yet its effective backlog reached $100.6M. Management is guiding FY2027 revenue to $130M–$150M, or roughly 160%–200% growth, with non-GAAP net income targeted at 18%–22% of revenue.

Those are extraordinary numbers.

But there is another number investors should know:

AEHR’s ten largest customers represented 82% of FY2026 revenue.

Two individual customers each accounted for at least 10% of revenue.

That’s meaningful concentration.

It doesn’t invalidate the investment case — specialized semiconductor equipment companies often begin major product cycles with concentrated customers — but it means investors shouldn’t treat all $100M of backlog as though it were diversified recurring SaaS revenue.

A production delay at one major customer can matter.

A capex reduction can matter.

A platform change can matter.

But AEHR is doing exactly what we want to see

The encouraging development is that the concentration story appears to be improving qualitatively.

AEHR isn’t relying solely on one application anymore.

Its $41M production order from its lead hyperscale AI customer validates Sonoma package-level burn-in for very-high-power AI processors. That customer is also developing a substantially higher-power next-generation accelerator.

Meanwhile, silicon photonics continues moving from qualification toward actual production.

On August 4, AEHR announced another FOX-XP production order from its lead silicon-photonics customer, with shipment expected during the first half of calendar 2027.

And SiC hasn’t disappeared either: AEHR disclosed more than $8M of new silicon-carbide wafer-level burn-in orders in July.

So the important evolution is:

one technology → several technologies

and potentially:

one dominant growth customer → several meaningful production customers.

If that happens, AEHR becomes a substantially better business.

Now look at TRT through the same lens

TRT’s situation is almost the mirror image.

Its semiconductor operation is dramatically smaller and less technologically differentiated than AEHR’s today.

But because TRT starts from such a small revenue base, a customer that would barely move the needle for a larger semiconductor company can completely change TRT’s financial profile.

The latest reported numbers demonstrate this beautifully.

For the nine months ended March 31, TRT revenue increased from $25.8M to $47.7M — about 85% year over year.

In Q3 alone, revenue jumped from $7.4M to $16.5M — approximately 124%.

That isn’t a theoretical AI TAM slide.

That’s revenue appearing in the financial statements.

And this is why I think TRT deserves more attention than its size suggests.

But here’s TRT’s biggest problem

Growth isn’t enough.

For those nine months, TRT produced $7.64M of gross profit on $47.67M of revenue, implying a gross margin of only about 16%.

Net income was just $419K.

This tells us something extremely important:

TRT has already demonstrated demand. It has not yet demonstrated operating leverage.

That’s the next hurdle.

AEHR’s FY2027 guidance implies the possibility of roughly $23M–$33M of non-GAAP net income if it hits its $130M–$150M revenue range and 18%–22% target margin.

TRT isn’t remotely close to that profitability profile today.

So AEHR deserves a much higher multiple.

No argument from me.

Yet this is exactly where the TRT opportunity lies

Imagine TRT eventually generating $70M–$80M of annual revenue but remaining at roughly today’s profitability.

That wouldn’t excite me much.

Now imagine something different:

TRT’s new capacity gets utilized.

AI GPU burn-in programs become repeat production orders.

Testing volume increases.

Fixed costs are absorbed over a larger revenue base.

And higher-value burn-in boards/equipment become a larger percentage of the mix.

Then the income statement could look dramatically different.

This is my investment scenario, not company guidance.

But mathematically, that’s where TRT’s asymmetry comes from.

TRT doesn’t need another 100% revenue-growth year forever.

It needs today’s extraordinary growth to start producing substantially more profit per incremental dollar of revenue.

Balance sheets change the risk equation too

There’s another reason I’m willing to give TRT time.

At March 31, TRT had approximately $13.0M of cash plus $2.6M of short-term deposits.

That was before considering the subsequent equity financing we’ve discussed previously.

AEHR is much better capitalized, ending May with approximately $116.5M of cash, cash equivalents and restricted cash.

AEHR clearly wins here in absolute financial firepower.

But TRT doesn’t appear to be trying to fund an AI expansion from a distressed balance sheet.

That distinction matters enormously for a microcap.

So which concentration risk would I rather own?

This is where things get interesting.

AEHR offers:

far greater visibility + proprietary technology + higher potential margins + much higher expectations.

TRT offers:

less visibility + lower margins + less proven differentiation + dramatically lower expectations.

That sounds like an easy victory for AEHR.

Until we add one more variable:

starting size.

A $10M incremental program is meaningful for AEHR.

A $10M incremental program can be transformative for TRT.

That’s why I keep describing TRT as an asymmetric situation rather than simply a “cheaper AEHR.”

It isn’t a cheaper AEHR.

It’s a different business at a much earlier stage of recognition.

What would make me substantially more bullish on TRT?

Not another press release containing the word “AI.”

I’d want evidence of repeatability.

The strongest possible sequence would be something like:

AI customer #1 → repeat order → next-generation program → customer #2 → customer #3.

If that happens while gross margins begin recovering, the investment thesis changes dramatically.

At that point, TRT would no longer merely be benefiting from one semiconductor upcycle.

It would be demonstrating that it has become a valuable supplier inside several advanced-semiconductor production chains.

That’s when I would expect investors to reconsider the valuation multiple.

AEHR is already beginning that diversification process

This is exactly what makes AEHR so impressive today.

AI processors are ramping.

Silicon photonics is moving into manufacturing scale-up.

SiC customers are ordering again.

Management is pursuing GaN and memory/HBM opportunities as well.

AEHR therefore gives us something useful:

a roadmap for what successful diversification in semiconductor burn-in can look like.

TRT doesn’t need to copy AEHR’s products.

But strategically, it needs to follow the same principle:

more programs, more customers, more semiconductor exposure — without allowing one customer to determine the company’s destiny.

My conclusion

Today, AEHR is unquestionably the more mature semiconductor investment.

Its $100.6M effective backlog provides exceptional visibility, and its proprietary platforms give it economic characteristics TRT simply doesn’t have yet.

But the market knows this.

What fascinates me about TRT is that the financial statements are already showing a major transformation before the company has demonstrated mature profitability.

Revenue nearly doubled over nine months.

The semiconductor business is driving the change.

AI-related opportunities are appearing.

Capacity is expanding.

Yet profits remain tiny.

Some investors will see that last sentence as the reason to avoid TRT.

I see it as the central question that could create the opportunity.

If margins never improve, today’s skepticism will have been justified.

But if TRT turns today’s revenue expansion into operating leverage while adding additional AI/advanced-semiconductor customers, then the company we are valuing two years from now could look very different from the company represented by today’s trailing earnings.

That’s why $AEHR remains the higher-quality investment today, while TRT remains, in my view, the more interesting “prove it and rerate” opportunity.

And with a microcap like TRT, sometimes you don’t need everything to go perfectly.

You need a few important things to go right before everyone else realizes they’re going right.

$nvda $nvds $bmnr $mram $dram $cohu $cohr $acmr $wolf $mdrna $md

[Mr Lambo @0xmstr]

Guys, today’s Trio-Tech International update is actually interesting precisely because there hasn’t been another major press release since July 28.

I checked both Trio-Tech’s current 2026 news archive and the latest SEC filings. As of August 20, the most recent major commercial announcement remains the $3.8 million additional order for burn-in boards supporting a next-generation AI GPU platform.

That may sound uneventful, but it gives us an opportunity to look at $TRT the way I think serious investors should:

not by asking what happened to the stock this week, but by asking what the next financial statements need to prove.

Let’s remember where we stand.

The AI GPU order sequence announced by TRT has been:

March: $5.3M
May: $2.5M
June: $2.6M
July: $3.8M

That’s approximately $14.2 million in announced orders for one next-generation AI GPU platform in only a few months.

I think the repetition is more important than the headline number.

A single large order tells us a customer is interested.

Four successive orders begin to tell us something about customer validation, product relevance and potentially recurring demand.

And management’s language is worth paying attention to. CEO S.W. Yong said after the July order that TRT continues aligning its capacity, workforce and technical capabilities with demand across AI GPU, CPU and automotive semiconductor markets.

Those three words—capacity, workforce and technical capabilities—matter.

Why?

Because TRT is trying to move from winning orders to building an organization capable of servicing substantially more business.

We can already see that transition physically.

The company has leased another approximately 104,000 square feet in Perai, Penang, Malaysia, to add semiconductor-testing capacity and support AI-related demand, including customers from North America and Europe operating in Southeast Asia.

The image above is Trio-Tech’s own HTRB reliability-testing equipment.

This is the part of the semiconductor industry that doesn’t receive the same attention as GPUs themselves—but it is essential. High-value semiconductor devices must survive extreme electrical and thermal conditions, and reliability testing is designed to identify weaknesses before those components enter mission-critical applications.

That’s why I increasingly think the easiest way to understand TRT is this:

TRT doesn’t have to predict which AI application wins. It needs semiconductor complexity to keep increasing.

More powerful GPUs.

More sophisticated CPUs.

More demanding automotive power semiconductors.

More expensive devices where failure becomes increasingly unacceptable.

All of those trends increase the importance of reliability testing.

And the financial evidence that something is changing inside TRT is already substantial.

Fiscal Q3 revenue reached approximately $16.5M, up 124% year over year.

Semiconductor Back-End Solutions alone generated approximately $13.1M, up 141% year over year.

But here’s the part I think we should focus on now.

The easy phase of the thesis is finished.

We already know TRT can grow revenue dramatically.

Now TRT needs to prove it can make that revenue meaningfully profitable.

During the first six months of FY2026, revenue jumped 69% to $31.2M, but gross margin fell from roughly 25% to 16%. Operating income was only about $143,000, and net income attributable to common shareholders was about $203,000.

That’s the contradiction at the heart of TRT today:

fantastic top-line growth, but profitability hasn’t caught up yet.

I don’t necessarily see that as bad news.

I see it as the single biggest variable we now have to monitor.

Because imagine what happens if TRT maintains a much larger revenue base while utilization of its expanded infrastructure increases.

Fixed costs get distributed across more business.

New capacity becomes productive.

Operational efficiency improves.

And even a modest recovery in gross margin can have an outsized effect on operating profit because TRT is starting from such a low profitability base.

That’s operating leverage.

And in my opinion, that’s where the truly interesting TRT investment thesis begins.

But we should also be intellectually disciplined.

There are real risks.

The customer behind the AI GPU program remains undisclosed. We should therefore not assume it is Nvidia, $AMD or anyone else without confirmation.

Customer concentration remains important.

TRT diluted shareholders with the approximately $10M equity raise in April.

Margins could remain structurally lower if the newer business mix simply isn’t as profitable.

And because TRT is a very small company, losing a major program could materially change the numbers.

These aren’t minor details. They’re exactly what we need to test against the bullish thesis.

So here’s my framework going forward:

Orders → Revenue → Utilization → Margin → Earnings → Cash flow

TRT has already delivered compelling evidence on the first two.

The next three are where shareholder value can really be created.

And this brings us to what I consider the next major catalyst.

TRT’s fiscal year ended June 30, 2026. As of August 20, neither the company’s investor-relations page nor the SEC shows the FY2026 10-K yet; the latest quarterly report remains the March 31 Q3 10-Q.

That means the next annual report is unusually important.

When it arrives, forget the daily share-price reaction for a moment.

Look for these five things:

1. Q4 semiconductor revenue — Did the extraordinary growth continue?

2. Gross margin — Is there finally evidence that profitability is beginning to follow scale?

3. AI order conversion — Are those announced orders flowing into recognized revenue?

4. Malaysia expansion — How quickly is additional capacity becoming productive?

5. Forward demand — Does management continue talking confidently about AI GPU, CPU and automotive programs?

If several of those move positively together, then we’re no longer merely discussing a promising microcap.

We may be watching a genuine earnings inflection.

That’s why I remain excited about TRT—but deliberately patient.

Anyone can get excited when a small stock moves 20%.

The more interesting skill is recognizing when the business underneath the ticker is becoming substantially larger than it used to be.

Revenue +124%.

Semiconductor Back-End +141%.

Four AI-GPU-related order announcements totaling ~$14.2M.

104,000 additional square feet of Malaysian capacity.

Those are facts.

Now we need the next set of facts to tell us whether scale turns into profit.

If margins begin improving while this level of semiconductor demand persists, I think the TRT story enters a completely different phase.

Until then:

No hype.

No imaginary customer names.

No ridiculous price targets.

Just evidence.

And right now, the evidence still gives us very good reasons to keep watching closely.

$aehr $acmr $cohu $cohr $sndk $mu $wolf $poet $silc

[Mr Lambo @0xmstr]

Friends, ACM Research has already shown us that it can grow. The more interesting question now is whether it can turn that growth into something much larger: a genuine multi-product semiconductor equipment platform.

The latest quarter gave us a strong starting point. Q2 revenue reached $292.9 million, up 36% year over year, while GAAP operating income rose 56.9% to $49.7 million. Non-GAAP operating income was $56.3 million, or 19.2% of revenue. Management also raised 2026 revenue guidance to $1.125–$1.175 billion.

But today I want to focus on something different:

$ACMR is openly planning for a business that could eventually exceed $4 billion in annual revenue.

That is not official guidance, and management is careful to label it as an internal long-term target rather than a forecast. Still, the logic behind the target is worth understanding. ACM estimates that its current product portfolio addresses roughly $22 billion of annual serviceable market opportunity across cleaning, plating, furnace, PECVD, track and advanced packaging.

The company’s internal roadmap effectively assumes two businesses developing at the same time.

The first is mainland China, where ACM believes it can ultimately generate around $2.5 billion from categories including cleaning, ECP, furnace, PECVD, track and packaging.

The second is the rest of the world, where its internal long-term target is approximately $1.5 billion.

That second number is the one I would watch most closely.

Why?

Because if ACM remains primarily a China semiconductor-equipment supplier, the company can still grow significantly—but investors will probably continue assigning a geopolitical discount to the stock.

If ACM proves that its technology can win repeat production business in North America, Singapore, Korea, Taiwan and other international markets, the investment thesis changes materially.

And the company is building the infrastructure to attempt exactly that. Its Oregon facility, including a clean-room and R&D/demo capability, is planned to become operational in Q4 2026.

That facility by itself does not guarantee international orders.

But it matters because semiconductor customers often need local demonstration, process development, service and technical support before they trust a new equipment supplier with mission-critical production steps.

So this is where I distinguish fact from interpretation.

The fact is that ACM is opening U.S. infrastructure and has already announced international evaluation and production activity.

My interpretation is that this could become one of the most important catalysts for the stock over the next several years, because international validation could help narrow the valuation discount investors have historically applied to ACMR.

There is also a powerful product-diversification trend developing underneath the headline revenue numbers.

In Q2 2025, cleaning represented about 72% of revenue.

One year later, it represented only 45%.

ECP, furnace and other technologies increased to 44% of revenue, while advanced packaging and other businesses reached another 11%.

This is an important evolution.

The goal is not for cleaning to become less successful. The goal is for several new businesses to grow around it.

That is exactly how a company can move from $1 billion of revenue toward $2 billion, $3 billion and potentially beyond.

And ACM has a very strong balance sheet with which to finance that expansion. At June 30, the company reported roughly $969 million of cash and cash equivalents, plus about $365 million of short-term time deposits and $105 million of short-term investments.

That financial strength gives ACM room to fund R&D, manufacturing capacity, demo facilities and international expansion without operating from a position of weakness.

Now let’s talk valuation.

ACMR is currently around $80.42, with a market capitalization of approximately $5.8 billion and a trailing P/E around 38×.

That means the market has already started to recognize ACM’s progress.

This is not the same setup as when ACMR traded at very low multiples and investors were paying almost nothing for future product expansion.

Today, expectations are higher.

That creates an important discipline for us as investors:

We should no longer judge ACMR only by whether revenue grows. We should judge whether the company is growing into its valuation.

The good news is that there are several ways it can do that.

Revenue growth remains strong. New product categories are becoming meaningful. The addressable market is expanding. International infrastructure is being built. And management is increasingly discussing ACM as a global equipment platform rather than simply a Chinese cleaning-equipment supplier.

But there are risks we should respect.

Gross margin declined to 46.0% from 48.5% a year ago. Customer concentration remains meaningful. Equipment acceptance timing can shift revenue between quarters. And international trade restrictions remain one of the biggest unpredictable factors surrounding the company. Management explicitly says those uncertainties are incorporated into the current guidance.

There is also a valuation risk.

At around 38× trailing earnings, a bad quarter could hurt the stock much more than it would have when ACMR traded at a much cheaper multiple.

But that does not make the story less interesting.

It simply means the investment thesis is evolving.

A few years ago, the question was:

“Can ACM become a meaningful semiconductor-equipment company?”

Today, I think the better question is:

“Can ACM become a meaningful global semiconductor-equipment company?”

If the answer turns out to be yes, then $1.1 billion of annual revenue may eventually look like an early chapter rather than the destination.

And this is why I still find ACMR compelling.

Not because management has written $4 billion on a presentation slide.

But because the business is gradually building the ingredients that could make a much larger company possible:

broader products, deeper customer penetration, international expansion and a strong balance sheet.

The next evidence I want to see is very specific:

repeat international production orders.

If those start appearing consistently, the ACMR story could move into a very different phase.

$nvds $nvda $sndk $mu $cohr $cohu $wolf $trt $aehr $poet $orcl $meta

[Mr Lambo @0xmstr]

Friends, $ACMR Research just delivered another strong quarter, but this time I want to focus on something deeper than the headline numbers.

Yes, Q2 revenue reached $292.9 million, up 36% year over year. Yes, GAAP operating income increased almost 57% to $49.7 million. And yes, management raised its 2026 revenue outlook to approximately 25%–30% growth, or about $1.125–$1.175 billion. Those are excellent figures.
But what really caught my attention is this:

$ACMR is becoming much less dependent on cleaning equipment.

In the first half of 2025, cleaning represented about 73% of revenue. In the first half of 2026, that fell to roughly 49%.

Meanwhile, ECP, furnace and other technologies grew to around 41% of revenue, and advanced packaging/other rose to about 11%.

That is a major strategic change.

For years, investors could reasonably describe ACM Research as primarily a cleaning-equipment company with some additional products.

That description is starting to become outdated.

ACM is increasingly selling a broader collection of semiconductor tools across electroplating, furnace, track, PECVD, advanced packaging, stress-free polishing and wet processing. The company estimates that its current portfolio addresses roughly $21 billion of annual wafer-fab-equipment opportunity.

And the growth rates inside those newer businesses are remarkable.

In Q2, management said ECP-related revenue grew 168% year over year, while advanced packaging grew 153%.

That is the type of growth I want to see from a company trying to move from being a niche supplier toward becoming a broader semiconductor-equipment platform.

Another milestone deserves attention.

ACM shipped its 2,000th electroplating chamber during the quarter. These systems are already being used in logic, memory and 3D packaging production.

Two thousand chambers may sound like just a statistic, but semiconductor equipment qualification is difficult.

Customers do not casually put equipment into high-volume semiconductor production.

So increasing installed base creates something valuable: process experience, customer references, service relationships and credibility when ACM introduces the next product.

This creates a potential flywheel:

More installed tools → more customer confidence → easier qualification of new products → larger addressable revenue per customer.

That last part is my interpretation, not management guidance—but it is one of the reasons I find $ACMR increasingly interesting.

Now look at what is happening in advanced packaging.

ACM has received the first production order for its Ultra ECP ap-p horizontal panel electroplating system, scheduled for delivery in the first half of 2027, along with an evaluation order from a new Asian panel manufacturer scheduled for Q4 2026.

This matters because ACM is moving from “interesting technology” toward something much more valuable:

customer qualification and production adoption.

The system supports panel sizes up to 600 × 600 mm and plating processes used in pillars, bumps and redistribution layers—important manufacturing steps in advanced packaging.

There is another underappreciated development.

The Ultra C Tahoe platform is evolving from a cleaning product into a multi-process wet-processing platform capable of etching, cleaning, drying and monitor-wafer reclaim applications.

ACM says the Tahoe Recycle process is already running in volume production at customer facilities, while the platform can reduce sulfuric-acid consumption by as much as 75%.

Again, the principle is important:

ACM is trying to make one platform useful across more semiconductor process steps.

And then there is the geographic question.

This remains the biggest issue in the investment thesis.

ACM's SEC filings showed that substantially all revenue was still generated from mainland Chinese customers earlier this year. Customer concentration also remains significant: in Q1, three customers represented approximately 46% of revenue.

That means China exposure, export restrictions and geopolitical risk cannot simply be ignored.

But there is one encouraging development to watch very closely.

Management says its Oregon operation is planned to open in Q4 2026, while the company continues expanding engagements with global customers.

If ACM can eventually prove that its tools can win meaningful production orders outside China, that could be one of the most important valuation catalysts in the entire story.

Now we need to talk about price.

ACMR is around $81 per share today, giving the company a market capitalization of roughly $5.8 billion and a trailing P/E around 39×.

So this is no longer the extremely cheap ACMR many investors discovered several years ago.

At this valuation, investors are already paying for substantial future growth.

That means execution matters.

Gross margin, for example, declined to 46.0% from 48.5% a year earlier.

And ACM still carries some unusual balance-sheet characteristics for a semiconductor equipment company: at March 31 it had roughly $738 million of inventory and about $527 million of net receivables, reflecting long qualification and acceptance cycles.

Those numbers deserve monitoring.

But ACM also finished Q2 with approximately $1 billion of net cash, giving it considerable financial flexibility to fund R&D, manufacturing expansion and global growth.

So my view today is slightly different from a year ago.

The question is no longer simply:

“Can ACMR keep growing?”

It clearly can.

The much more interesting question is:

“How many semiconductor equipment categories can ACM become genuinely competitive in?”

Because if cleaning remains strong while ECP, furnace, track, PECVD and advanced packaging become meaningful businesses, the company could eventually look much more like a diversified equipment supplier than the niche player investors originally knew.

Management continues to target $4 billion of long-term revenue.

That target is ambitious and absolutely not guaranteed.

But $ACMR is now approaching $1.2 billion of annual revenue while several newer product categories are growing triple digits.

That is why I continue to find the story compelling.

Not because the stock has gone up.

Not because AI makes every semiconductor stock attractive.

But because the business itself is becoming broader, more established and harder to dismiss.

The next big proof point will not just be another revenue beat.

It will be whether ACM can convert its new tools—and especially its international evaluations—into repeat production orders.

That is what I’ll be watching next.

$nvda $mu $nu $sndk $cohu $cohr $acmr $trt $nbis $nvds $aaoi $orcl $rddt

[Mr Lambo @0xmstr]

$TRT vs. $AEHR — Part 1: The Valuation Gap

Friends, I want to start this series with what I think is the most important concept when comparing Trio-Tech International TRT with Aehr Test Systems AEHR:

$AEHR is currently the better recognized semiconductor growth story. TRT may be the more interesting asymmetry.

These are not identical businesses. AEHR is much more of a proprietary semiconductor test-equipment/platform company, particularly around wafer-level and package-level burn-in. TRT combines semiconductor back-end testing and burn-in services with equipment and other industrial-electronics activities. That distinction matters because AEHR deserves structurally higher margins and probably a higher valuation multiple.

But the difference in market perception is enormous.

AEHR has already been discovered

AEHR's latest numbers are genuinely impressive. Fiscal 2026 revenue was only $50.0M, but management is guiding fiscal 2027 revenue to $130M-$150M — roughly 160%-200% growth. Even more importantly, effective backlog has reached about $100.6M, following record quarterly bookings of $60.7M. Management expects non-GAAP net income of 18%-22% of revenue.

That is an exceptional inflection.

AI processors, silicon photonics, SiC, GaN and potentially memory/HBM give AEHR several ways to win. This is exactly why the market has rewarded the company so aggressively.

But remember an important investing principle:

A fantastic company and a fantastic stock are not necessarily the same thing at every valuation.

By mid-July, AEHR shares had already risen roughly 257% year-to-date and 387% over twelve months. (Barron's)

The market is no longer waiting to discover AEHR's growth story. It is actively paying for it.

And that's where $TRT becomes fascinating.

$TRT is much earlier in its recognition cycle

Look beneath TRT's tiny-company appearance and something interesting is happening.

For the quarter ended March 31, TRT generated $16.51M of revenue versus $7.38M one year earlier — an increase of roughly 124%.

Even more important, Semiconductor Back-end Solutions revenue increased from $5.43M to $13.08M, approximately 141% growth.

For the first nine months of fiscal 2026, total revenue reached $47.67M versus $25.80M a year earlier. (SEC)

Think about that for a moment.

TRT generated almost as much revenue in nine months as AEHR generated during its entire fiscal 2026.

Yet investors still perceive the two companies completely differently.

That doesn't automatically mean TRT deserves AEHR's valuation. It doesn't.

But it does tell us where the potential rerating opportunity could come from.

The hidden catch: margins

Here's where we have to be disciplined rather than promotional.

TRT's growth has come with substantial margin compression.

Its Semiconductor Back-end Solutions gross margin fell from 26.5% to 15.1% in the latest reported quarter. For the nine-month period, consolidated gross margin fell from 25.3% to 16.0%. Management explicitly says that newer final-testing services carry lower margins and expects margins to remain below historical levels as these services become a larger part of the business. (SEC)

AEHR therefore deserves a premium.

If AEHR actually delivers $130M-$150M of FY2027 revenue and converts 18%-22% of revenue into non-GAAP net income, it could become an extremely profitable semiconductor equipment company. (Aehr)

TRT isn't there.

But here's what interests me:

TRT doesn't need to become AEHR to generate an excellent investment return.

It only needs to become more valuable than the market currently assumes.

TRT's balance sheet gives the story time

At March 31, before considering the subsequent capital raise, TRT reported approximately $13.0M cash, another $2.6M in short-term deposits, and only about $515K of bank loans.

Total assets were $44.7M against liabilities of $12.3M. (SEC)

And shortly afterward TRT raised approximately $10M of gross equity capital as it prepared to expand capacity. The company also leased roughly 104,000 square feet of additional Malaysian space as part of that expansion. (Triotech)

Dilution isn't something shareholders should celebrate automatically.

But there is a major difference between raising money because your business is failing and raising money because customers are creating more demand than your existing infrastructure was designed to handle.

My interpretation today is much closer to the second scenario.

The comparison I keep coming back to

Imagine two restaurants.

Restaurant A already has a Michelin star, a waiting list and investors know exactly how good it is.

Restaurant B has suddenly doubled its customers, is taking over the building next door, and most people haven't noticed yet.

Restaurant A may still be the better restaurant.

But Restaurant B might produce the larger change in valuation.

That's essentially what attracts me to TRT.

AEHR has demonstrated what happens when semiconductor burn-in becomes strategically important to AI infrastructure.

TRT is showing early evidence that the same secular forces are beginning to materially change its own business.

And because TRT starts from such a small base, relatively modest contracts can transform its financial profile.

What would make the TRT thesis really powerful?

For me, the next stage isn't simply more revenue.

It's operating leverage.

TRT generated $47.7M in nine-month revenue but only about $62K of operating income, because gross margins have compressed while the company scales.

So imagine what happens if the company eventually combines:

$70M-$100M+ revenue + better utilization of expanded capacity + a richer product/service mix + recovering margins.

That's not management guidance. That's my investment scenario, and it shouldn't be confused with a forecast from the company.

If margins remain around 15%-16% indefinitely, TRT deserves a relatively modest multiple.

If revenue keeps growing and margins eventually migrate toward 20%-25%, the earnings power changes dramatically.

And if TRT develops more proprietary, higher-margin burn-in products alongside its testing services, the market could eventually start valuing it less like a generic outsourced testing company and more like a specialized semiconductor infrastructure business.

That's the rerating I'm interested in.

One reason $AEHR still deserves respect

We shouldn't build the TRT thesis by pretending AEHR has weaknesses it doesn't have.

AEHR currently has something TRT lacks:

visibility.

Its effective backlog of roughly $100M already represents about twice AEHR's entire FY2026 revenue. And management has explicitly guided to $130M-$150M next year.

That's extraordinary visibility.

TRT hasn't provided anything comparable yet.

AEHR also historically carries customer-concentration risk — its five largest customers represented 77% of FY2025 revenue — although diversification into AI processors, silicon photonics, GaN and other applications should help over time.

TRT has its own concentration and execution risks, plus its much smaller size means individual contracts can move results dramatically in either direction.

So AEHR is currently the more proven growth story.

TRT is the more speculative rerating story.

And those are different investments.

Why I'm increasingly interested in TRT

If I had to summarize today's comparison in one sentence:

AEHR shows us what the market is willing to pay when semiconductor burn-in becomes a recognized AI growth story; TRT gives us the opportunity to ask what could happen before that recognition is complete.

That is the asymmetry.

I'm not arguing that TRT should trade at AEHR's valuation tomorrow.

I'm arguing something subtler:

The valuation gap does not need to disappear for TRT shareholders to do extremely well. It merely needs to narrow.

And if TRT continues turning AI and advanced-semiconductor demand into real orders, revenue growth and eventually stronger profits, the market may gradually stop thinking of it as an obscure micro-cap industrial company.

That's when things could get very interesting.

$sndk $mu $wolf $cohr $cohu $nvda $nvds $acmr $bmnr $eth $btc $aaoi $nbis $meta $googl

[Mr Lambo @0xmstr]

Guys, I want to revisit Trio-Tech International because the latest development reinforces what I believe is the most important part of this investment story: the AI-related orders are becoming repetitive rather than looking like a one-off event.

On July 28, Trio-Tech announced another $3.8 million in orders for burn-in boards supporting a next-generation AI GPU platform. This takes the cumulative orders associated with this platform to approximately $14.2 million since March.

For a huge semiconductor company, $14.2M would barely move the needle. For $TRT , context changes everything. The company's entire fiscal Q3 revenue was $16.5M, which itself represented extraordinary 124% year-over-year growth.

Think about that for a moment: the AI GPU orders announced since March alone are approaching the size of an entire recent quarter of company revenue.

But the sequence matters even more than the number.

March brought roughly $5.3M. Then came approximately $2.5M, another $2.6M, and most recently another $3.8M.

That repetition is exactly what we should want to see. A first order can be qualification. A second can still be experimental. Repeated orders begin to suggest that TRT may be becoming embedded in an ongoing production program.

And this connects directly to the company's expansion strategy.

$TRT has taken approximately 104,000 additional square feet in Malaysia, increasing capacity for semiconductor testing and specifically positioning itself for AI-related demand from North American and European customers operating in Southeast Asia.

This is where the investment thesis gets interesting.

AI doesn't only require companies designing GPUs. Every generation of high-performance chips becomes more powerful, hotter, more complex and enormously expensive. Those chips need reliability screening and burn-in testing before being deployed into data centers.

TRT sells the picks and shovels behind that process.

The picture above shows the kind of Trio-Tech reliability-testing equipment involved in stressing semiconductor components under high temperatures and electrical loads. That's the less glamorous part of the AI ecosystem—but potentially a very valuable one.

There is still one number I am watching extremely carefully: margin.

TRT's growth has been spectacular, but profitability hasn't yet followed revenue at the same speed. During the first six months of FY2026, revenue increased 69% to $31.2M, while gross margin fell from about 25% to 16%.

That's the central question now.

If TRT can combine today's dramatically higher revenue base with better utilization of the new Malaysian capacity and gradually recover margins, the earnings profile could look very different from the historical company.

But if revenue explodes while margins remain permanently compressed, the investment becomes much less attractive.

There is another risk we should acknowledge: this is still a small company with meaningful customer concentration. We also do not know the identity of the AI GPU customer, so claims online connecting these orders specifically to $AMD, Nvidia or another chip designer should be treated as speculation unless TRT confirms it. That distinction matters.

So I wouldn't describe TRT as a guaranteed AI winner.

I would describe it differently:

TRT is giving us increasingly strong evidence that something significant is happening inside the business.

Revenue +124%.

Semiconductor back-end demand accelerating.

AI GPU orders now around $14.2M since March.

Multiple repeat orders.

104,000 sq. ft. of additional Malaysian capacity.

And management investing to handle substantially more semiconductor volume.

The market will obsess over the stock price from one day to another.

I'm much more interested in watching orders → revenue → capacity utilization → margins → earnings.

If that chain develops the way we hope, the stock price eventually has to respond to the economics of the business.

For now, my conclusion hasn't changed:

The TRT thesis is getting stronger.

Not because of hype.

Because the evidence keeps accumulating.

$nvda $nvds $mu $sndk $acmr $cohu $cohr
$aehr $dram $mram $wolf $orcl $meta $googl
$tsla $spcx

[Mr Lambo @0xmstr]

The most important takeaway from $ACMR Research’s latest quarter is not simply that revenue grew 36% year over year to $292.9 million. It is that the composition of the business is changing in a way that could matter a lot more over time. In Q2 2025, cleaning equipment represented roughly 72% of revenue. In Q2 2026, that dropped to about 45%, while ECP/furnace/other technologies rose to 44% and advanced packaging/other increased to 11%. In other words, ACM is becoming a much more diversified semiconductor-equipment platform.

That diversification is important because it reduces dependence on one product family and increases the amount of capital spending ACM can address inside each customer fab. The business is no longer only about cleaning tools. ACM now spans cleaning, electroplating, furnace, track, PECVD, stress-free polishing and wafer- and panel-level packaging equipment. (ACM Research, Inc.)

The earnings quality was also strong. Q2 GAAP operating income reached $49.7 million, up 56.9%, faster than revenue growth, while non-GAAP operating income rose 35.8% to $56.3 million. Non-GAAP EPS was $0.61 versus $0.55 a year earlier. Gross margin did decline to 46.0% from 48.5%, so this was not a perfect quarter, but the company still demonstrated meaningful operating leverage.

Then we get to the part I find particularly interesting: advanced packaging.

$ACMR has received the first production order for its Ultra ECP ap-p horizontal panel electroplating system from an existing customer, plus an evaluation order from a new Asian panel manufacturer. The production system is scheduled for delivery in the first half of 2027, while the evaluation tool should ship in Q4 2026. ACM says the platform targets copper deposition across pillar, bump and redistribution-layer processes for large-panel advanced packaging. (ACM Research, Inc.)

Why does this matter?

Because advanced packaging is becoming increasingly important for AI, high-performance computing and 2.5D/3D semiconductor integration. ACM had already announced earlier this year advanced-packaging orders from customers in Singapore, North America and elsewhere outside mainland China. These latest orders add another piece of evidence that the company is attempting to build a genuinely international customer base rather than relying indefinitely on domestic Chinese semiconductor spending. (ACM Research, Inc.)

There is another development that should not be ignored: the Ultra C Tahoe platform is expanding beyond cleaning into wet etch and wafer-reclaim applications. ACM says the platform has already been adopted by multiple leading semiconductor manufacturers, with its Tahoe Recycle process now running in volume production at customer facilities. The company also says the architecture can reduce sulfuric-acid consumption by as much as 75%. (ACM Research, Inc.)

That may sound like a technical detail, but this is exactly how semiconductor-equipment companies deepen customer relationships: one qualified platform expands into more process steps, becomes more useful inside the fab, and potentially generates more revenue per customer.

And the financial outlook remains constructive. ACM raised its 2026 revenue guidance to $1.125–$1.175 billion, versus the previous $1.08–$1.175 billion range. Management specifically notes that this outlook already incorporates its current assessment of trade policy, customer spending scenarios, supply-chain constraints and timing of first-tool acceptances. (ACM Research, Inc.)

At around $79 per share, ACMR currently has a market capitalization near $5.7 billion and trades at roughly 38× trailing earnings. That is no longer a bargain-bin valuation, so execution now matters more.

But this is the distinction I would make:

The stock has become more expensive because the business has become more credible.

The bullish case is no longer simply “China semiconductor spending will stay strong.”

The stronger case is becoming:

ACM is growing quickly, broadening its product portfolio, entering advanced packaging, winning international orders, and increasing the number of semiconductor process steps it can address.

There are still real risks. Gross margins softened this quarter. Export controls and US-China regulation remain significant uncertainties. Customer concentration and the timing of equipment acceptance can make quarterly results volatile. And at today’s valuation, disappointment would probably be punished more severely than when ACMR traded at much lower multiples. ACM itself continues to highlight international trade policy and first-tool acceptance timing as factors affecting the outlook. (ACM Research, Inc.)

But when I look at ACM today, I see a company whose opportunity set is expanding faster than its original narrative.

That is what I want to see in a long-term growth company.

Not just higher revenue.

More products. More customers. More geographies. More applications.

If ACM keeps executing across those four dimensions, the company we are evaluating a few years from now could be considerably larger and more diversified than the ACM Research most investors first discovered.

$aehr $mu $sndk $nvda $nvds $cohu $cohr $orcl $mu $dram $mram $wolf $eth $btc $ACMR

[Mr Lambo @0xmstr]

Guys, I think ACM Research is reaching the point where we need to stop thinking of it as a “small semiconductor equipment company” and start looking carefully at what the numbers are actually telling us.

Yesterday $ACMR reported Q2 2026 revenue of $292.9 million, versus $215.4 million a year ago.

That is roughly 36% year-over-year growth.

For a semiconductor equipment company already approaching a billion dollars in annual sales, that is significant.

But there is something I find even more interesting.

Management raised the bottom end of its 2026 revenue guidance. The previous range was $1.08–$1.175 billion. It is now:

$1.125–$1.175 BILLION.

Think about what that means.

We are halfway through the year, management has more visibility into customer spending, export restrictions, supply-chain conditions and tool acceptances — and instead of becoming more cautious, they have effectively narrowed expectations toward the upper part of the previous range.

That matters.

ACMR generated about $782 million of revenue in 2024 and approximately $906 million in 2025. If they reach the midpoint of the new guidance, around $1.15 billion, the company will have increased annual revenue by almost 50% in just two years.

This isn’t a story based purely on AI excitement or a future product that may or may not arrive.

They are selling real semiconductor manufacturing equipment today.

And this is the part of the ACMR thesis that I think many investors still underestimate.

ACM started with advanced wafer-cleaning technology, but the company has been steadily expanding into a much broader equipment platform: cleaning, plating, furnace, track and other semiconductor manufacturing processes.

Every successful new tool creates another opportunity to capture more spending from the same semiconductor fabs.

That’s how a relatively small equipment company can potentially become a much larger one.

There is also another strategic piece worth watching carefully: international expansion.

For years, the obvious criticism of ACMR has been its dependence on China.

That’s a legitimate risk. We shouldn’t pretend otherwise.

US-China trade restrictions, customer concentration, geopolitical tensions and the timing of equipment acceptance can all create volatility.

But management has been deliberately trying to build business outside mainland China.

If ACM can prove that its equipment is competitive not only with Chinese fabs but also with major semiconductor manufacturers elsewhere in Asia and globally, the market may eventually have to reconsider what valuation multiple this company deserves.

And that’s why I don’t think the ACMR story is simply:

“Will next quarter beat estimates?”

The more important question is:

Can ACM evolve from a successful China-focused equipment supplier into a global semiconductor equipment platform?

If the answer eventually becomes yes, today’s revenue numbers could look surprisingly small several years from now.

Of course, nothing moves in a straight line. There will be weak quarters, export-control headlines and probably some brutal corrections in the stock.

But businesses ultimately create value by selling more products, gaining customers and increasing earnings power.

Right now, ACMR continues to move in that direction.

Q2 revenue: +36% YoY.
2026 guidance: raised.
Product portfolio: expanding.
International opportunity: developing.

That’s why I remain very interested in $ACMR.

Don’t just watch the stock price.

Watch the business.

Because if the business keeps compounding, eventually the market has to pay attention.

$nvda $ACMR $cohu $cohr $sndk $mu