mole

Engineer digging through AI infrastructure, photonics and overlooked tech stocks. posting the research as I go. NFA.

[mole @MoleLabs]

@TheodorusAtheos Good point. I think that sivers partnership with seminex is excellent for those orbital data centers opportunities.

[mole @MoleLabs]

Serenity made a great point.

There are “no notable public EU merchant laser suppliers other than $SIVE”

Lumentum, Coherent and AAOI can redirect laser capacity into their own optical modules. Europe has almost no public supplier capable of providing high-power CW DFB lasers for SiPh without competing downstream.

$SIVE fits that slot unusually well.

70/100mW lasers already sampled with pluggable customers
200mW lasers being discussed for NPO
laser arrays for CPO
partner-foundry capacity available now
Glasgow expansion online by end-2027

Marvell’s NPO window: late 2027–2028.

Obviously this doesn’t confirm Sivers is supplying Marvell. But geography + merchant model + product + capacity + timing makes this one hell of a coincidence 🤔

What do you guys think? Am I reading too much into this, or is there a cleaner European candidate than $SIVE?

[Serenity @aleabitoreddit]

$NVDA told JP Morgan... That their revenue could increase more than 100%+ Y/Y unconstrained.

But ended up giving a ~70% growth projection because that's what Nvidia believes they can supply.

FY27 est: ~$401B revenue...
FY28: ~$682B at 70% or >$802B unconstrained.

- Nvidia IR said it gave the updated outlook because it had greater visibility + saw a meaningful gap between Street ests. and its own projections

- Inference is now larger than training after being roughly 50/50 about 18 months ago

I wonder if Nvidia will be the first $10T company... this growth is absurd.

[mole @MoleLabs]

@Joeybks Because companies need all the InP laser supply they can get. The shortage is going to be out of this world.

[mole @MoleLabs]

$AVGO is more than tripling its EML, CW, VCSEL and InP factory buildout YoY, yet Hock Tan still says laser demand is “far surpassing supply.” Are you listening yet anon ??

Six days ago, $SIVE said pluggable customers were still heavily constrained by CW laser supply.

Then tonight, Hock Tan said demand for CW and EML lasers is “far surpassing supply.”

Broadcom is already a major laser supplier, yet it is still expanding its EML, CW, VCSEL and InP factories across the US and Singapore. Management said the buildout is more than tripling YoY!!!and will continue growing over the next two years.

EMLs are mainly used in pluggable transceivers, while CW lasers feed silicon photonics used across pluggables, NPO and eventually CPO.

Multiple architectures are fighting over the same limited pool of qualified laser capacity. $LITE CEO’s comment about InP shortages being worse than memory is starting to sound like an understatement lmao

This makes Sivers’ decision to push its 70mW/100mW CW lasers, expand Glasgow and bring on an outside InP foundry look a lot less speculative.

Sivers said CW supply was constrained, and six days later one of the largest players in the industry basically said the exact same thing while pouring money into capacity.

[mole @MoleLabs]

$SIVE is on the edge of oversold, but it hasn’t earned “bottom” yet.

RSI ~30 after a 37.6% five session drop. It closed at its low on above average volume and lost the 26–31 SEK base.

Watch 20–22. If it reclaims ~27.3 SEK and holds the retest, then it gets interesting.

Will 20–22 hold? We’ll find out tomorrow 🫡

[mole @MoleLabs]

@Classedx Exactly. POET was packaging the lasers, not actually making them. Marvell may have cut POET out, but Celestial’s system still needs CW lasers from somewhere. We don’t know who’s supplying them now, but sivers would make a lot of sense.

[mole @MoleLabs]

Fair concern around cash burn and the gap between a non binding $1.2B pipeline and recurring revenue. One nuance tho, the raise and debt conversion were dilutive, but they also left Sivers debt free with a much stronger balance sheet. 2027 is hopefully when we should start seeing meaningful pipeline to revenue conversion 🤞

[mole @MoleLabs]

real conviction should make you calmer, not more obsessive.

if you need to check your account every 5 minutes, the position is probably too large, your thesis isn’t clear enough, or both.

know why you own it, what would prove you wrong and how long it could realistically take. then go walk, go lift, do literally anything besides staring at every tick lol

[mole @MoleLabs]

@BULLOFBRITAIN LMAO

[mole @MoleLabs]

@retail_mourinho lol much needed closing

[mole @MoleLabs]

@dockery58 @Sofigoodboy Exactly 😭 people act like Sivers can just wake up tomorrow and press the Nasdaq button. There’s an entire process behind a dual listing, and getting the timing right matters.

[mole @MoleLabs]

Vickram said $SIVE could be ready for a US dual listing by spring 2027.

“you can’t rush the pot to boil.”

imo those were the most important lines from the Redeye interview. Ready does not mean they automatically list that spring. It means the groundwork should be complete and Sivers can choose the moment when a listing actually makes sense.

guys, this is not some little box management checks because shareholders are impatient. Sivers gets one first impression with a much larger US investor base. Listing before the product ramps are visible could waste that opportunity.

and like it or not, Sivers is executing. Product/HW revenue grew 18% YoY FX-adjusted while the business shifts from one-off NRE work toward repeatable products across three separate supercycles: AI infrastructure, space/SATCOM and defense.

they’re moving as fast as they responsibly can. spring 2027 is now the checkpoint.

[mole @MoleLabs]

@crux_capital_ Fully agree with you. We are in the pre super cycle era.

[mole @MoleLabs]

@h3llcall @Sofigoodboy Brother no need to spread negativity. I wish you the best on your investing journey.

[mole @MoleLabs]

@Sofigoodboy Lmao marvell didn’t name sivers. But it did name Europe and there aren’t many independent European InP laser suppliers operating their own fab….

[mole @MoleLabs]

@Matrix_B0SS @SiversSemicond Agreed. Sivers could definitely improve their US engagement/reach.

[mole @MoleLabs]

@Sgotteepee @Sofigoodboy It’s on the “Redeye Nordic growth website”.

[mole @MoleLabs]

@BananaPrinciple it removes friction: more US funds can own it, liquidity and visibility improve, and analyst coverage becomes more likely.

[mole @MoleLabs]

$SIVE has not guided to a Nasdaq listing in H1 2027.

it has guided to being Nasdaq-ready by H1 2027. the actual listing still depends on market conditions, investor demand and business momentum.

those are very different things.

Nasdaq says its exchange review generally takes 4–6 weeks. so the long timeline is not about filling out listing forms. my read is that Sivers is using 2026 to repair the financial foundation and let the operating story catch up.

the 2025 audit flagged significant financial-control deficiencies and material uncertainty around going concern. 2024/25 were restated, and then Q2 2026 corrected the 2025 comparative figures again.

walking into an SEC review and a US institutional roadshow with messy historical numbers, negative EBITDA and a $1.2bn pipeline that is still largely non-binding would be a weak way to launch.

H1 2027 gives them a chance to arrive with:

• a complete FY26 audit under the new CFO
• roughly SEK825m of gross fresh capital and a debt-free balance sheet
• LiDAR and ALLSPACE revenue ramping
• Tachyon/FWA and Jabil moving from beta builds toward production orders

they have already spent roughly SEK25.5m on dual-listing work across 2025 and H1 2026. that doesn’t look like an abandoned plan. it looks like expensive preparation for a listing they only want to execute when the company is ready.

the merger question has a similar answer.

Sivers already tried the byNordic Photonics de-SPAC transaction in 2024 and walked away when the small-cap/SPAC environment made the deal unattractive.

today Jabil, POET, O-Net, GlobalFoundries and the foundry partners give Sivers much of the scale and vertical-integration benefit without paying an acquisition premium, taking integration risk, issuing more shares or losing its neutrality as a merchant laser supplier.

so yes, an actual H2 2027 Nasdaq listing is later than i wanted.

but i no longer think the timeline automatically means management is asleep. the rational sequence is:

clean reporting → FY26 audit → production orders → Nasdaq from a stronger position.

[mole @MoleLabs]

@jlin1012 Not a lot of people realize this !!

[mole @MoleLabs]

$DELL just dropped a ridiculous quarter

EPS: $7.04 vs $4.92 expected
Revenue: $47B vs $44.5B expected

and somehow the bigger number is the guide: Dell now expects $74B in FY27 AI server revenue, up from $60B.

AI capex is still accelerating 🚀🚀