AI/Semi Supply Chain Analyst
ex. RISC-V FDN, AI research scientist;

Serenity

Just some notes on $POET AGM and read through on optical markets:

- "The top three laser suppliers control 68% of the market, and they’re completely sold out for the next two years"

$LITE CEO said into 2028, so POET implicitly confirms laser shortage is going into 2029 now.

- NRE with a new customer, building on POET’s interposer for high-power external light source.

This is high confidence $SIVE as laser supplier given the Sept 29, 2025 PR on ELS, and new customer qualification would be material for revenue if it goes into volume ramp. 

- Poet expects Lumilens commercial agreement to scale to over $500 million over the next 5 years.

Lumilens claimed a top-3 hyperscaler was their initial customer (Linkedin OSINT)

- $830M cash on hand on balance sheet (this is more for Poet fundamentals).

- "The entire optical components industry today is facing a severe shortage of critical components."

Reaffirming what we know already regarding optical bottlenecks.

- "Production ramp in the second half of 2026 this year." 

Just production timelines H2. 

In terms of $POET volumes:
- existing capacity around 1 million optical engines per year
- projected demand exiting 2027 around 1 million optical engines per month
- roughly 10x capacity expansion

I don't own Poet, but if management delivers on these projections it directionally looks very positive.

However something to note is that unlike other suppliers that have named hyperscaler customers to to back up capacity revenues like $AAOI: $POET seems more questionable.

But this does look like a very positive outlook for $POET if they match projections.

Nothing is more stupid than Bloomberg copying BofA playbook with KOSPI back in March.

And framing Taiwan’s $TSM ~23.6x forward p/e and $NVDA upstream supply chain as a “bubble”.

Then using the words going into “debt” instead of “margin” + comparing Taiwan’s market to the .com bubble. 

Crap like this is why retail trust in media has dropped off a cliff in the past few years.

Feels like the only thing that hasn’t crashed…

Is memory like $MU, indexes, or large cap semis like Intel so far.

- Photonics from $AXTI to $SIVE down 40%.

- Space from $ASTS and $RKLB down 40% 1M. 

- Popular AI names like $PLTR is down ~35% YTD. 

- Software like $CRM down -40%.

- Bitcoin sub <60k, Ethereum sub <$16k.

Not a fun time with a hawkish fed narrative and potential rate hikes. 

However this does sorta feel overshot due to margin liquidations on less liquid assets compared to mega caps.

But we’ll see what happens, usually fundamentals override liquidity shock in the longer run.

I’m still personally bullish on the AI buildout + upstream AI capex beneficiaries, but 1-2 potential rate hikes certainly don’t help.

Markets should be cheering on domestic champions like $AAOI. 

Since it's ideal to support critical AI infra from laser fab to production in the US, rather than being a bear.

Feels like everyone just outsources transceivers to Asia like Malaysia or Thailand...

With $INTC, $IQE, $XFAB, $MU, $WOLF, $SOI, $SIVE, and others... 

If you haven't noticed by now, they're all critical to US supply chains. And every one of them are getting subsidies for securing Western supply chains. 

Before a major trade was to short developing US/Western equities, then hedge with subsidized foreign ones. 

As seen with the energy/solar firms that went bankrupt, this backfired a lot on US AI infrastructure years later with the power grid. 

I wanted to help change this mindset, since I believe it's very positive sum to invest in building up critical Western supply chains like photonics today. 

Especially if $AAOI hits their $471m/month projections after reshoring their production to America. 

Instead of hoping they fail and calling critical nodes in the supply chains memestocks/bubbles, maybe it's good to change mindsets a bit so we don't see a repeat of the US Solar sector years later. 

US/EU don't just hand out subsidies or CHIPS act grants to anyone.

Curious if anyone's portfolio is green after today's fun day.  

$KORU: -32.06%
$SOXL: -22.98%
$IQE: -13.58%
$DRAM: -12.6%
$AXTI: -12.57%
$FLNC: -12.5%
$AAOI: -11.2%
$SIVE: -11.7%
$TSEM: -10.24%
$SNDK: -12.5%
$MU: -11.8%
Sk Hynix: -12.35%
Samsung: -9.6%
$MRVL: -8.3%
$LITE: -7.6%
$SOI -7.15%
$TSM: -6.1%
$AMD: -6.04%

Feels like anything high-beta or semis had a steep drop.

Very interesting statement today: $MU CEO predicts a multi-decade memory demand cycle driven by humanoid robots.

"Humanoid robots, he says, will require roughly ten times more memory than today’s Level 2+ autonomous vehicles."

"And that demand wave is set to begin before the decade is out."

Something as well as was "Over time, we expect the value of on-device AI combined with pent-up unit replacement demand to drive memory demand growth"

Which is also another trend (Apple Intelligence is currently dog, but I'm sure we'll see innovations with localized/edge AI). 

Feels like all the industry leaders from $TSM Chairman, $TSLA Elon Musk, to $MU CEO see humanoids as the next major trend so physical AI is probably next. 

I wonder if the world is going to have enough memory. Or if we'll see enough breakthroughs to shrink memory usage.

Trendforce reports that $AMD is actively trying to secure CW laser supply with multiple major procurement orders...

Is probably just the start of the bottleneck?  

There's not much independent capacity in Western supply chains left other than $SIVE or $AAOI and maybe Macom. 

Especially after Lumentum/Coherent got locked up with multi-year agreements with Nvidia. (disclosure, own Sive and aaoi)

Lumentum is already CW laser constrained and is likely buying off Japanese companies like Sumitomo/Furukawa if I had to guess per ER, and those are probably running at max capacity.  

From the Trendforce report, this is:

"to ensure that its future capacity will not be constrained by NVIDIA and other major Cloud Service Providers (CSPs)."

I wouldn't be surprised if other hyperscalers like Amazon, Microsoft, and others saw Nvidia / AMD signing LTAs, and are trying to secure capacity next. 

A lot of it is game theory on not getting choked out by competitors, and looks like AMD is tipping the first domino after Nvidia.

But my opinion is that this just goes and show how invaluable this CW laser chokepoint is and the companies are inside it.

Just in case you’re wondering why indexes + individual names like $SNDK to $MRVL to $LITE are green now.

Trump just cancelled attacks on Iran.

This market is so volatile… 

Fun new information discovery from Poet OSINT community:

Seems likely that $POET / $SIVE are going to power a Top-3 hyperscaler (either Amazon, Microsoft, Google). 

Given a Linkedin update from Ankur Singla (CEO of Lumilens). 

Who stated their customer is one of the top 3 hyperscalers with their post focusing on CPO/NPO. 

With that clue, seems more likely the Sivers CW DFB light source path over other EML suppliers given it's CPO Scale Out/NPO. 

If you don't remember, Sivers is the laser supplier to Poet. And Poet has purchase agreements with Lumilens.

Always fun to find major potential breadcrumbs in the wild before they're officially confirmed. (Disclosure, long Sive)

Yeah… I think all your upstream semi supply chain companies are going much higher.

Goldman now expects a combined $5.3 trillion of capex spending for the four largest hyperscalers $GOOGL / $META / MSFT / $AMZN from 2025 to 2030.

Revised up from $4.5T from Q1 earnings.

“Aggregate capex est. $7.6 trillion between 2026 and 2031.”

And it flows upward to these tiny chokepoints like $SIVE for CPO lasers/ $SOI for Silicon Photonics substrates. Leaderdrive/Harmonic for Humanoids components. 

And so on…

Ai names don’t move in a straight line up, 
but is just the beginning of the next Industrial Revolution as we move from  R&D/compute buildout into commercialization from Agents -> Physical AI -> discovery.

Look who joined team $NBIS. 

5.6% is a pretty massive stake, maybe he realized by now it’s miles better than the dumpster fire that is $IREN.

That being said: Nebius is now up ~3x since I went long last year. 

Weebius has plot armor. Source: Serenity

I wonder how it feels to miss out on the AI supercycle run…

With photonics, memory,
and even Neoclouds like $NBIS.

Because you’re busy bagholding into the endless $6,000,000,000 active ATMs with $IREN. 

Okay... just some more weekend shower thoughts about $XFAB.

I still feel like it could be the next $TSEM, just early stage at a $1.4B MC?

They kinda leapfrogged current gens (which $TSEM are getting volume from) to compete for H2 2027 CPO scale up inflection point ($ASX docs cite Xfab (aka. photonixFAB) as focusing on CPO)

By building out some black magic MTP (transfer printing) architecture for lasers w/ other stuff like TFLN.

Basically next-gen integration IP, they're still behind on yields, sure.

But $NVDA evaluating it for transceivers/switches to see if it can volume ramp. That $NOK sets the specifications/assembly for. (nvidia invested in nokia for this these switches/networking too btw). 

And if their MTP supply chain works... (eg. with Smartphotonics providing lasers, EU players doing assembly).

It basically volume ramps with $NVDA just like why Nvidia signed long term agreements with $TSEM?

Downside risk?

Already below replacement book value, can always go lower yeah, but typically to a certain point. 

Maybe more CHIPS act subsidies next few months from chips act 2. If it doesn't go well there's SiC (152% Y/Y Growth, 195% Y/Y SiC wafer shipment growth)/GaN power semi upside.

Europeans /LLMs will say "oh evaluations doesn't mean it's a future contract!".

This is kinda different since the European Union is behind this effort and $XFAB for soverign photonic supply chains.

Not your typical company + hyperscaler evaluation, since $NVDA wants to be nice to Europe's regulators. They'd prob be pissed if nvidia just stayed in US/Taiwan/China.

So if they can make this MTP black magic work with mass production, feels almost for sure nvidia/nokia volume ramp on some tiny $1.4B silicon photonics foundry or at least throw them a bone with smaller contracts. 

In terms of timelines, maybe just a months early since it volume ramps H2 2027/H1 2028 (which happens to be in line with CPO scale up timelines)...

Or just unknown because they named their project something stupid like photonixfab?

Like XFAB Photonics would have been better? so institutions/screeners can connect the dots when looking at CPO silicon photonic foundry players?

Automotive should also coming out of a slump medium term, sped up by self-driving (TSM Chairmain comments yesterday said ai automotive was TSM's growth vector alongside robotics). So their core business also should pick up speed too medium term. 

Obviously markets/europeans want a "Nvidia signs $2B+ contract, XFab volume ramping 2027!" 

But by then it will be a $9B+ company and you miss out on all the upside. And especially since everyone analyst/institution is blind to volume expectations for these.... 

Normally don't invest in companies in evaluation stages, but this just seems very de-risked by EU sovereignty + Gov backing, and you have Nvidia + Nokia there for volumes if they can make the IP work. 

I think markets are probably missing something here...  there's almost 0 value being assigned to being CPO exposure in Europe as their long term upside.

Honestly this decade from 2020-2030 might be the most goated in human history.

- Scaling massive reusable rockets for orbital compute with $RKLB to $SPCX.

- on the cusp of ASI and recursive learning with Anthropic and OpenAI 

- Backflipping Boston Dynamics and Unitree humanoids to replace the human workforce

- Star Wars laser beams from $EOS.ASX to AI DC lasers like $LITE.

- Waymo and $TSLA self driving cars everywhere in urban cities

- and we get industry Quantum commercialization end of decade

This is kinda crazy to be an investor in this timeframe. Feels like every movie from the Star Wars Death Star to Skynet is coming to life.

What’s next?

Everyone is out there making life changing returns on $RDDT.

Fully leveraged on the wrong $SPCE (Virgin Galactic) ticker.

Instead of $SPCX, which is yet to launch. 

I guess this just goes to show how much retail demand there is for SpaceX’s IPO. 

There was interesting research published called "Democratization of Retail Trading".

That did a study on 1.6 Million $RDDT WSB comments. 

and found: 

1. "WSB outperformed almost all investment banks at detecting top-performing stocks."

2. "Their average returns compete with the best investment banks and outperform them in certain cases."

Their conclusion? 

"We conclude that WSB may indeed constitute a freely accessible, valuable source of investment advice."

I do find WSB is really early to names like $RKLB, $HOOD, and others, but often get timing extremely wrong (with options). 

I think X is where all the alpha is at nowadays.

Had some more time go through $SIVE earnings transcript, it’s very bullish:

- $JBL pluggable partnership led to more optical transceiver requests for $SIVE  

So maybe Innolight/Eoptolink and other large players are my guess. 

- More laser capacity on top of Win Semi with more partners being developed. “When the timing is right, we will bring those details to the market.”

CEO said it’s not just Win/Glasgow. I’m already confident in Win Semi scaling capacity, given they’re critical in SpaceX / $AVGO supply chains already. 

But this derisks those capacity ramp even more. Just not publicly disclosed yet.

-  tremendous executive credibility and experience with U.S. markets, as well as strong M&A experience

Flagged M&A in regards to new board members, which we guessed based on their backgrounds.

- “Production orders are imminent from our lead SATCOM”… 

So that’s volume ramp for space 

- U.S dual listing progressing smoothly

No exact timeline, my personal guess was around late Q3 or Q4. Probably after June board meeting, they’ll announce timing since that’s when new board members come in.

- “Viewing the ecosystem vendors as competitors is the wrong way to go about it in supercycles where demand far outstrips supply.”

Too much demand in photonics

- “Over the last five months, there's been a rapid increase in the Photonics pipeline as well”

Basically 77% growth pipeline came from photonics (which validated thesis about cpo/pluggable growth vectors for sivers)

TLDR:

We moved from “can sivers get customers this small and can they compete with $LITE?”

To execution eg. “how much can sivers even produce to feed into each supercycle” as they’re volume ramping while demand > supply.”

To me that’s very positive if anything they make gets bought. 

Also there’s likely to be a lot of TAM expansion in the photonics space post-M&A as well as more ongoing hyperscaler supplier qualifications hinted. 

Again, revenue pipeline surged 77% in just their quarter (5m) largely from photonics… over the entire company’s history. 

As CPO scales up h2 2027 onwards, I’m expecting the revenue numbers to look like an exponential curve.