Graylume

looking

[CM X @CM_X_CM]

$ANY at 14mill marketcap with Joel Block as CEO(Ex CFO $HUT)

I decided to make it a bigger bag and added a bunch

Model 1: Straight EV per MW
This is the fastest sanity check. If 53 MW is worth $225 million, that is about $4.25 million per MW.

Applying that same multiple to the 100 MW pipeline implies about $425 million of incremental value if those megawatts are monetized at the same quality and certainty.

That gives an indicative platform value of roughly $650 million for 53 MW + 100 MW together, before adjusting for execution risk, time to delivery, or financing dilution.

Model 2: Discounted staged rollout
This model assumes the 53 MW is the near-term cash-flow asset and the 100 MW pipeline is delivered in phases, so you discount the pipeline more heavily. If you value the 53 MW at the full $225 million and then apply a 35% to 50% haircut to the 100 MW pipeline for timing and execution risk, the pipeline might be worth roughly $213 million to $276 million instead of $425 million.
That produces a combined enterprise value range of about $438 million to $501 million.

This is the most conservative model and is usually the right one if capex, permits, or tenanting are still uncertain.

Model 3: AI colocation / option value
This model assumes the 53 MW signed contract is the floor and the 100 MW pipeline has option value because AI inference and edge colocation demand often rerates capacity once tenants are secured. The 53 MW still starts at $225 million, but the 100 MW pipeline can arguably deserve a premium multiple if the market believes it can be converted into sticky AI capacity rather than generic hosting.

Using a 1.25x to 1.75x premium on the base $4.25 million per MW for the pipeline gives roughly $531 million to $744 million for the 100 MW upside block alone.

That yields a total platform value of about $756 million to $969 million, which is the bull-case frame if the company executes like a real AI infrastructure platform rather than a miner with spare power.

$keel $dgxx $bgde $slnh $hive $shaz $aib

[Graylume @Graylume1]

Q1 2026 inflection: Revenue €1.495M (–7% YoY due to timing), but EBITDA +€0.330M (+22.1% margin) — first clear positive quarter. Operating cash flow turned positive. Cash €9.1M, net cash €6.59M, equity ratio 90.8%. 2026 guidance: both revenue and EBITDA expected to rise. $MODU

[Graylume @Graylume1]

Defense & aerospace: radiation-tolerant lasers with ESA flight heritage (SMOS, Galileo), high-brightness pumps, rangefinding, and LIDAR.

EU strategic autonomy push + rising defense photonics budgets are direct tailwinds for a European fab with proven qualifications.

[Graylume @Graylume1]

@stckpkr I think Modulight has bottomed here and it's go time soon

[Graylume @Graylume1]

You found Modulight?

They have their own semiconductor laser chip fab in Tampere, Finland. 62,000 sq. ft. Fully vertically integrated. $50M invested. ISO5 cleanrooms.

What do they build?

Lasers from UV to 3000+ nm. Fiber lasers, VCSELs, DBRs, solid-state. Applications:

- Quantum computing
- Semiconductor metrology (wafer inspection)
- Flow cytometry, diagnostics, defense

[Graylume @Graylume1]

Solid take on $MRN

If you're scanning European photonics plays, also check Modulight $MODU.HE (Finland).

Core business right now is medical lasers, oncology photoimmunotherapy, ophthalmology, flow cytometry.

But they also have serious AI/quantum upside:

Own one of the few fully vertically integrated III-V semiconductor laser fabs in the EU (Tampere cleanroom).

Supply precision narrow-linewidth, stable lasers for quantum computing & sensing (key enabler for future photonic/quantum AI acceleration).

Their ML6600 platform already integrates AI-driven beam alignment, predictive maintenance, and cloud-based treatment optimization.

Tiny ~€45-50M market cap vs much larger AI optics peers. Undervalued moat in a strategic EU photonics play.

[Graylume @Graylume1]

Hey, check out Modulight right next door in the $SIVE neighbour country Finland.

One of the very few companies left in Europe with a fully owned, vertically integrated semiconductor laser fab in Tampere.

EU Chips Act 2.0 is coming, real in-house laser chip production could be a serious edge for the handful of players who still have it.

[Graylume @Graylume1]

@aleabitoreddit Nailed on $SIVE!

Check Modulight ($MODU). Ask yourself: how many genuine in-house semiconductor laser fabs are actually left in Europe?

EU Chips Act 2.0 is coming. This could give a real boost to the few companies that already have their own fab, IP, and European base.

[Graylume @Graylume1]

Solid write-up on $SIVE and the InP laser momentum.

Quick one for you, take a look at Modulight. Then ask yourself: how many true in-house semiconductor laser fabs are actually left in Europe? EU Chips Act 2.0 is coming. That could be a real tailwind for the few players who already have the fab, the IP, and the European base.

[Graylume @Graylume1]

Fair point! you're right that I used several “coulds”, and the medical business is still clearly the main revenue driver today.

That said, the strategic asset is real: Modulight operates one of the very few fully vertically integrated semiconductor laser fabs in Europe, right here in Tampere (62,000 sq ft facility with in-house epitaxy, chip fabrication, and full system integration).

They already have a Fortune 500 quantum computing customer (prototype orders delivered in 2025) and are active in areas where photonics is becoming a critical bottleneck for AI infrastructure and quantum systems. In that context, the EU Chips Act 2.0 and Europe’s push for photonic sovereignty give this kind of local capability extra relevance.

I’m not claiming they’re the next big semi winner yet, just that the setup is more interesting than a pure medical laser story, and it deserves objective scrutiny rather than immediate dismissal.

[Graylume @Graylume1]

Fair point, and I respect that you’re generally against pumping.

That said, I do think there’s a noticeable Finnish tendency to be overly cautious or even dismissive when our own scandinavian deep-tech companies start gaining attention, especially compared to how aggressively other European countries promote their semiconductor and photonics names. It’s not about blind pumping imo, but about whether we give our own innovators a fair chance to be taken seriously before they fly under the radar for too long.

Modulight is a good example of why this matters. The company runs one of the very few fully vertically integrated semiconductor laser fabs in Europe, located in Tampere. They control the whole stack, from epitaxial growth and chip fabrication to complete laser systems and their own cloud platform.

In today’s AI-driven world with growing concerns over semiconductor supply chains and sovereignty, this kind of independent photonics capability is strategically valuable for Finland. If they successfully scale their Pay-Per-Treatment model and expand into quantum computing and defense, Modulight could develop into a meaningful high-tech exporter and contributor to national technological resilience.

[Graylume @Graylume1]

An interesting perspective.

One does wonder whether this instinctive aversion to “pumpers” when it comes to Finnish deep-tech names is a characteristically Finnish trait — a cultural reflex to quietly undervalue our own innovators rather than actively support them while the rest of Europe hunts for the next bottleneck play.

Genuinely curious to hear your thoughts.

[Graylume @Graylume1]

@Fibonazzo Finnish FinTwit is waking up to the whole AI bottleneck mania in Europe, while their own Finnish AI/photonics stocks like $MODU, $CANATU, $NOK have barely received any real attention from the Finns yet.

[Graylume @Graylume1]

Excellent and timely analysis.

I am particularly grateful that you included Modulight among your picks. It is very satisfying to finally see the company receiving the well-deserved attention from the FinTwit.

Your overview of Finland’s deep-tech strengths and the positive tailwinds from the EU Chips Act 2.0 is both insightful and highly relevant.

[Graylume @Graylume1]

@Ren_aramb You have to check Modulight $MODU

[Graylume @Graylume1]

@valuephilos Check Modulight $MODU

[Graylume @Graylume1]

@Mercantio @emilhartela Emerging high-tech customers in quantum computing (won a Fortune 500 quantum customer already in 2022) and other advanced applications like semiconductors.

$MODU