Nicolas (@snmart) header
Nicolas (@snmart) profile picture

Deep dives + asymmetric bets | AI infra | Small caps | Building $10k → $1M.

Sharing early-stage ideas
Joined May 2025
Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)Photo by Nicolas (@snmart)
Nicolas retweeted
$BE bulls, genuine question. At $173, $BE is trading at: > Market cap: $51B > Enterprise value: $54B > 2026 revenue guidance: $4.4 - 4.6B > Q3 revenue growth: +138% > EV/2026 sales: 12x > EV/2026 adj. EBITDA: 51x > YTD free cash flow: $42M Im genuinely curious what $BE bulls think Do you consider the stock cheap at $173?
37
10
237
73,316
Me then, me now. Oh well. It worked out. It will work out
1,328
1,886
616
24,633
8,475,969
Nicolas retweeted
I think people are overlooking some really interesting things in $AMZN Q3 results. - AWS revenue +39.3% YoY, accelerating for the 6th consecutive quarter. - AWS backlog reached $578B, more than doubling YoY. - Bedrock token volume grew nearly 17x YoY. - Fortune 500 spending on Bedrock increased almost 7x. - Revenue commitments for Trainium and Graviton each increased more than 20x YoY. Inference already accounts for well over half of AWS AI revenue, and Jassy confirmed that most Bedrock workloads run on Amazon's own Trainium chips. So Amazon isnt just renting out computing capacity. It's building its own chips, running its own AI platform and capturing more of the economics. And despite all the investment, AWS still delivered a 38.7% operating margin. Of course, the $220B capex plan and negative free cash flow are real concerns. But these results suggest that AI demand is already translating into substantial revenue growth, not just future promises. Very impressive quarter from $AMZN
1
1
369
Nicolas retweeted
$AXTI Q3 numbers are out. - Revenue $73.4M, a record. Up 54% from Q2 and 162% from a year ago - Gross margin 56.1% vs 45.0% last quarter and 22.4% a year ago The CEO says they've already passed the goal of doubling InP capacity in 2026, and the 2026 InP revenue opportunity has more than quadrupled. Cash and equivalents sit at $485M, and yesterday's Tongmei financing adds up to RMB 3.67B more for InP capacity. Waiting for the call: Q4 number, InP split, export permits... $AXTI $AAOI $LITE $COHR $IQE
3
11
1,502
Nicolas retweeted
Samsung just reported some insane Q3 numbers > Revenue: $146B > Operating profit: $80B, up +780% YoY > Semiconductor operating profit: $79B Samsung says its HBM supply is already sold out for 2027. Management also expects AI driven memory demand to outpace industry supply by an even wider margin next year, with tight conditions potentially lasting through 2028. $SNDK $MU $SKHY
1
3
404
A lot of questions this week on funding, so here's how a GPU typically gets funded at $IREN. First, the customer signs a multi-year take-or-pay contract and agrees to a prepayment. On our recent 3-year contracts, prepayments have been 45% to 55% of the associated GPU capex. Second, financing is committed, secured by the contracted GPUs. On the Microsoft contract we secured an A-rated facility at around 6%, which together with prepayments covers 96% of GPU capex. More recently, for a non-investment grade counterparty at Mackenzie, we secured financing at 9% fixed, which together with prepayments covers more than the full cost of the GPUs. That's the GPU side. The data centers are a separate, long-term asset base, and we own the ones our GPUs run in. Construction has been funded from cash to date, with no debt against these multi-billion dollar assets. As they come online with contracted customers, they can support financing in their own right. The processes behind the next facilities are continuing, and new paper will cost what the market charges. With GPU paybacks of around two years on our recent 3-year contracts, a wider spread barely moves the economics. There's more customer demand than we can build for. As we noted at our recent earnings, 2026 is largely sold out, recent 3-year contracts are at more than double the prior year's pricing, and demand for anything we can bring online in 2027 covers it several times over, with prepayments a focus in every negotiation. Conversations on 2028 capacity are already underway. Meanwhile announced capacity keeps slipping across the industry, which is the point: signing deals isn't the bottleneck in this market, bringing GPUs online is, and that's our focus.
379
555
190
4,179
730,979
RT @AbelAvellan: American operators hold ~1,000 MHz of low and mid band spectrum, tunable on the @AST_SpaceMobile network. With our curren…
210
68
Nicolas retweeted
A merchant InP foundry just reached a 47% operating margin. That's got to be read through for $IQE I included some more of LandMark's financials in the quoted post. A merchant epi supplier grows the precise semiconductor layers that become the active part of lasers, modulators and detectors. The factories carry heavy fixed costs, so utilization can change the economics very quickly. That has been the central issue for $IQE. It owns valuable InP capacity, but weak utilization has kept the financials far below the quality of the underlying assets. Now the demand signals are improving. > $IQE recently got a multi-year agreement with $TSEM which includes minimum purchase volumes > A separate customer placed a $14M production order for AI and data-center applications. > $MTSI is investing £45m into IQE through equity and a convertible loan, taking two board seats and signing long-term supply agreements for epi services across multiple technologies. LandMark has a different manufacturing mix, so its margins simply illustrate the operating leverage available when these factories fill. But I think that Landmark gives us a great example of the direction that IQE can move towards if things start to go right
Landmark ($3081) giving us really good signals $IQE $AXTI $LITE $COHR $MTSI $SMTC They just put up: +121% YoY revenue growth 57.5% gross margin 46.9% operating margin 1.6T already in volume Capacity “very, very tight” 2027 capacity planned at ~2.5x 2026 And then an unnamed US customer signs a 4-year CW laser LTA covering 2027-2030. Meanwhile Landmark itself is locking up InP substrate supply for years. Everyone is securing the layer underneath them. I go through why this is important for $LITE $COHR $AAOI $SIVE $MTSI $SMTC $IQE $AXTI, why lower-power CW looks tight for a long time, and what Landmark’s margins could be telling us about the operating leverage available at the merchant epi layer. cruxcapitalgroup.substack.co…
4
6
2
52
18,548
Quick illustration how ridiculously cheap $FCEL currently trades: FCEL targets up to 500MW/year of manufacturing capacity by mid-2028. Even if they'd MISS that target by 40% and only reach 300MW/year, while we assume a pessimistic und in my view unrealisticly low price at ~$3.25M/MW: That's ~$975M of annual product revenue capacity. At today's ~$1.47B market cap, that's just ~1.5x that revenue capacity. And that's the conservative case, assuming you fundamentally don't trust their execution and dont' even assume Fuel Cells will get much more expensive as the only fast (90 day!) time to power solution for data centers. Current industry trends suggest pricing power should increase while unit costs fall significantly as FCEL scales. Higher prices + much better economies of scale = potentially massive margin expansion. And this ENTIRELY excludes long-term service/LTSA revenue.
21
8
3
137
24,328
Nicolas retweeted
What we could take away from recent earnings for $IQE I think the backdrop keeps getting better. The entire optical industry is telling us that we need way more InP laser capacity. This isn't new, but it's being reinforced and its pushing higher and faster. 800G still growing, 1.6T coming online, CPO & NPO haven't even started yet. A massive wave is coming. CPO / NPO require higher-power lasers that can use substantially more wafer area. And companies are already unable to make enough qualified lasers today. That all flows upstream to epitaxy which is the layer that IQE plays in. They take the raw InP wafer and grow the semiconductor layers that eventually become the lasers and photodetectors that the market is massively short on. For IQE to work they need the overall market to grow faster than internal capacity can keep up. They can assist with overflow production, second sourcing and geographic diversity. $MTSI invested into IQE and signed long-term supply agreements specifically around supply chain security. $TSEM also signed a multi-year agreement for IQE InP epiwafers IQE itself says InP demand has accelerated significantly and they still have room to increase production using the fabs and tools they already own. And I think that might be an underappreciated part here. Everyone is spending like crazy to grow their capacity. But for IQE a lot of this capacity already exists. As optics keeps pulling more volume through those fabs, the incremental revenue should carry a LOT more operating leverage than what we saw during the downturn. Now I want to see more LTAs / orders land. That will tell us whether or now IQE is actually in a really strong position or if it will just live on as a good narrative
$SIVE Isn't the only company I have reconsidered... I have spent a lot of time thinking about $IQE I was very disinterested when I first heard about them. But several things have come up over the last few months that have really altered my opinion Read this: cruxcapitalgroup.substack.co… One thing that is important to understand is that back in 2025 the idea of a massive bottleneck premium in this space didn't was way less common than it is now for fundamentally damaged / financially distressed companies. Especially European companies. IQE fits this bill. It wasn't worth much. When I first looked into it I honestly saw no potential. Then the bottleneck / optics premium got applied and it REALLY wasn't interesting. But then the thesis that was laid out by investors (notably @aleabitoreddit) started to take shape. Deals were made, things got cleaned up. And IQE is in the right place at the right time. In this article I break down exactly what IQE does (in a hopefully easy to understand way), why they are in a potentially powerful position, and what their future could look like. This article is less about the stock and more about the company. I don't own shares yet, but I definitely appreciate the company much more than I have in the past. I am someone that, when faced with new information, is willing to adjust their thesis. This is one of those situations.
8
8
1
53
13,253
$ABCL continues to hit new 3-year highs. Nothing changes sentiment like price action. I maintain my position that I have no clue what a stock will do in the short term. I continue to look five years out. This helps navigate volatility. Many retail investors focus heavily on day to day movement. Stocks drop sharply all the time for no reason related to the company specifically. Those investors then get discouraged… “what’s happening???” I’ve watched my PLTR position drop ~40% multiple times. No panic. It doesn’t matter as long as the thesis is intact and there’s strong reason to believe the company is likely to be worth much more in five years. I approach my ABCL investment in the same way. Stay focused on the long term. It’s fun to watch the stock go up, but really the short term is just noise. What matters is that the company keeps executing. Stock price will follow fundamentals over the long run.
27
16
3
278
21,317
Feels like my $IQE thesis Feb this year got validated alongside $AXTI and others. (I still long on both) So I stopped covering them as much. But TLDR on updates: Early this year, IQE was a distressed but critical chokepoint in the optical supply chain with links to $LITE and others when I published my thesis on it. Since then: - $MTSI signed long term epiwafer supply agreements, took a stake in the company, now IQE is bank debt free (Macom has convertibles, but that's positive) - $TSEM signed agreements for InP epiwafers. - Management explicitly said its IQE agreement was signed to secure the "strategic III-V epiwafer supply required for that roadmap" - IQE signed another ~$14M mysterious AI DC contract. The only part that's still playing out with my earlier thesis is IQE converting their latent/underutilized capacity toward AI DC segment, to match peers like Landmark in the space (which takes time, like slightly over a year if I remember correctly). There's some national security implications that I'll avoid talking about that makes them more important. But basically: FY2025 revenue: £97.3m FY2026 guidance: over £126.5m They're rapidly growing again in the optical sector, out of bank debt/distressed liquidation state, and remains a critical chokepoint in Western optical supply chains. Now... with new agreements with your leading photonics companies for revenue growth, and players like Macom serving as a soft backstop. After a thesis gets validated, typically I just sit back and enjoy growth of a business.
Replying to @aleabitoreddit
Any update on IQE? They have released a lot of good news recently would be good to hear your thoughts.
58
26
4
567
311,231
$ABCL 🚨 AbCellera announces positive top line ph2 results for ABCL635, a GPCR targeting antibody: - Phase 2 results showed best-in-class reductions in both frequency and severity of moderate-to-severe vasomotor symptoms after a single dose - ABCL635 significantly improved sleep and patient global impression of change - ABCL635 demonstrated a favorable tolerability profile “ABCL635 reduced VMSM-S frequency compared to placebo, with an 8.8 mean reduction in the number of moderate and severe events per day from baseline at week 4 (Day 29) compared to 3.5, showing a mean placebo-adjusted treatment difference of 5.3 (p<0.001). The mean percent reduction at week 4 for VMSM-S frequency was 83% for the ABCL635 treatment group versus a 33% reduction for placebo, leading to a mean placebo-adjusted treatment difference of 50%.” “ABCL635 also resulted in improvement in VMSM-S severity compared to placebo, with a mean reduction of 1.4 points at week 4 compared to 0.3, and a mean placebo-adjusted treatment difference of 1.1 (p<0.001). The mean percent reduction in VMSM-S severity from baseline to week 4 was 58% in the ABCL635 group compared to 12% in the placebo group, leading to a mean placebo-adjusted treatment difference of 46%.” “Meaningful improvements were also observed in patient sleep scores and patient global impression of change.” "These positive results represent a significant milestone for women's health and for AbCellera" said Sarah Noonberg, M.D., Ph.D., Chief Medical Officer of AbCellera. "By successfully targeting the NK3 receptor with an antibody, we have demonstrated potential best-in-class vasomotor symptom relief over four weeks with a single subcutaneous dose. These data indicate that ABCL635 could offer menopausal women a well-tolerated, long-acting treatment that has the potential to markedly improve their quality of life." Congratulations to AbCellera and all shareholders!!!!
AbCellera Announces Positive Top-Line Phase 2 Clinical Trial Results for ABCL635, Demonstrating Significant Reduction in Frequency and Severity of Vasomotor Symptoms and a Favorable Tolerability Profile investors.abcellera.com/news…
27
43
8
322
100,120
$ABCL is announcing top-line results from Phase 2 of ABCL635 on Monday. In the Q2 call, the company said the results would come in August. That was just two days ago, so this announcement is coming very early after the call. Could be an early positive indicator. This is a big moment for the company. Management has previously called the Phase 2 readout “highly de-risking” for the program. ABCL635 is particularly important because it is internally developed, AbCellera-led, and the first program generated using its GPCR and ion-channel technologies. Solid efficacy would be the strongest clinical evidence yet that AbCellera can discover a differentiated molecule and translate it into human efficacy. Positive readouts would mean the company can advance the drug further, would be a major step toward building a self-sustaining biotech business, and would provide meaningful clinical validation of the platform that could help attract more partnerships.
4
4
2
77
10,702
It has been over a month since my last post. I think this is a good time to look back at what news came out of $EOS.AX since I was last active here.
4
3
1
12
3,584
Nicolas retweeted
The market is red and I've never been more bullish: • Goldman Sachs hyperscaler capex data: AWS, $MSFT, $GOOG, $META and $ORCL combined went from $74B in 2020 to a projected $996B in 2027. These numbers accelerate every quarter of 2026. • SemiAnalysis published Anthropic's implied financials this morning. Q2 26 revenue landed at roughly $11B, more than double Q1's $4.7B. They now expect Anthropic to clear $1B in quarterly profit in Q3 26. Combined Anthropic and OpenAI ARR is around $100B. We are already in the inflection point: The application side is finally generating profit at scale. A single lab printing $1B in a quarter. Second order: neoclouds, memory and substrates absorb the spend. Third order: efficiency wins. Optical replaces copper where copper stops scaling and power delivery moves to 800V.
1
1
646
Nicolas retweeted
Counter-drone trend confirmation - upcoming defense research. First: Not everything we work on can be shared in real time. Continuing our due diligence on a company that caught our attention at Eurosatory. Looking forward to discussing it once we’re able.. So guys back to today: I’ve been hammering on the counter-drone trade for a long while now. First on $EOS.AX and then as a follow-up, our recent visit to Eurosatory, the big defense expo in Paris, where we sent @SimeonResearch_ to do the primary research and talk with everybody in the industry. We got confirmed there also that counter-drone will be the biggest spend driver and priority. NATO just committed to spending over €40 billion over the next five years. That strongly confirms the trend we’ve been on. If you look through my posts you can find a lot of what I’ve written on EOS publicly. More depth including management commentary you can find on our platform. So a small little recap. For EOS and their world-leading remote weapon system orders are stacking up. A Western European NATO country buying the system for urgent operational needs. They did the biggest naval order they’ve ever done, funded by a Western European government. A 100 kilowatt laser system export out of the Netherlands, and now into the Middle East with UAE contracts. And they bought MARSS… for the interceptor missiles and the NiDAR C2 platform integration. So the strategy, as I’ve been covering them for more than a year now, is building a whole turnkey stack. Detect, track, command and control, then neutralise it… kinetic with the remote weapon systems, or with the laser system. So they spent a long time positioning as the most affordable turnkey C-UAS answer to the drone problem. Still highly bullish on them. Now if you look at the image attached NATO in their official message pointed out that they’re investing in a growing innovation ecosystem across the alliance, including small and medium-sized enterprises playing a crucial role in delivering effective and affordable solutions. This is exactly EOS. And a few more we are in talks with. Lots to come on the defence research. $EOS.AX $AVAV $theon.as $RHM.NE
Electro Optic Systems $EOS.AX and the Dutch Defense 2026 Note Released today For over a year now, we've made the strong call on air defense, and especially counter-drone defense. The highly anticipated Dutch defense note of 2026 has just been released, and it shows the big drivers we were hammering on. So first: moving from ITAR to non-ITAR, and spending budget. The note says, in the government's own words, a burden shift from the US to Europe across cash, capabilities and contributions, tied to hard targets. At least fifty percent of procurement from Dutch and European suppliers, and at least forty percent bought jointly. And the Dutch will take a leading role in the EU priority capability area for counter-drone. So the big headline, essentially, is buy European. So next to the purchasing that moves away from the US, another few critical things are being built. They focus on a layered system: - Laser systems against drones, counter-drones, and EW into every land unit. - A counter-drone expertise center, sensor-to-effector capability, and electromagnetic battle management at joint command. This is the C2 layer, in the government's own infrastructure. More than half of operational effects unmanned within five years, built on affordable mass. The note names the cost-asymmetry problem directly, which EOS solves, which we'll go into later. The Netherlands is committed at doctrine level to a layered, cost-optimized counter-drone defense with a fused command layer at its center. And EOS has pre-positioned themselves fully to benefit from this. So what you can really learn from this document and the Netherlands: it's the shift that they don't just want an effector or weapon system. They want a full architecture that ties them together. So into EOS. The Netherlands is already EOS's launch customer, with the first 100-kilowatt laser program being delivered next year. As most of you know, EOS integrated this whole C2 layer with the MARSS acquisition in May this year, bringing it all together: NiDAR, the full AI-enabled C2 and sensor-fusion platform, and the Interceptor-MR, which is a hit-to-kill drone. Next to that, EOS already has all the effectors: the RWS system, the Apollo laser, and the Atlas space-control system that's in development, which will be for space defense and control. And this altogether makes EOS the most uniquely positioned pure-play turnkey solution on the market, in my opinion. See, the CEO of EOS has been pointing this out consistently too: that they want to be a turnkey solution for every country, and potentially even commercial customers, that want to defend big, huge, high-value property against drone attacks. So this defense note by the Dutch DOD is exactly the trend that EOS fully plays into: multilayered, turnkey defense, ITAR-free, away from the US, and positioned to build JVs for local production. They tick every box. Also, EOS management has said in an interview that they're going to move their headquarters to the Netherlands or Germany in the next twelve months. This will make them fully compliant with all the local programs. Then on to space, and this is a nice one. When I talked to $EOS.AX management and they laid out their strategy, moving to an integrated, turnkey system provider, the big quote was that we will move into high-energy laser systems first, and then we move into one of the biggest markets in the view of EOS, which is space defense and control. And the quote from that call was: whoever controls space, controls the battlefield. And this exact quote came back in the defense note, by defense minister Dilan Yeşilgöz, on Dutch television, who quoted this exact phrase from EOS. The Netherlands is standing up and building its own space command, built out from the Defence Space Security Centre, which the Dutch cabinet wants operational within two years, with hundreds of millions to 2035 for space sensors, ground laser systems, and sovereign space control. Read further below 👇
2
2
26
5,542
Nicolas retweeted
Humanoid robots: 1. Forecasts: Goldman says $38bn by 2035. Barclays says $40bn base, $200bn if things go right. BofA models 10 million units that year, up from 20,000 in 2025. Morgan Stanley waves at $60tn. Near term beneficiaries are factories: 72% of 2027 installs in logistics, manufacturing and automotive. My view: technology is moving faster than expected. I think we are only three years away from seeing millions of humanoid robots deployed. 2. The bottleneck is not the AI: The brain is the cheap, solved part. Memory will still be the constraint but the hard part is the body. This is a machine-tool problem, not software. A humanoid is a car in miniature, thousands of fine-tolerance parts, and the edge sits with whoever already has automotive and precision-mechanics DNA. Guess who have ALREADY the capacity to produce millions of EV cars? CHINA. 3. The map, by layer: • Brain (semis + models): $NVDA, and the memory it needs in $MU , SK Hynix, Samsung. Cheap, crowded, commoditizing. • Brawn (the actual bottleneck): screws and reducers in Hiwin, THK, Harmonic Drive, Nabtesco, Jiangsu Hengli. Motors in Nidec, Estun, Inovance. • Sensors and structure: force/torque, vision and connectors in Sensata, TE Connectivity, Sony, Keyence, $OUST. • Rare earths (the input under all of it): $MP, Lynas, JL Mag. China controls over 90% of magnetic rare earths. • Integrators (the full robot): Tesla, plus a heavily Chinese field, UBTech, Xiaomi, BYD, XPeng, and Hyundai via Boston Dynamics. Morgan Stanley data point that gets buried: 73% of confirmed humanoid participants are Asian, 56% Chinese. The US has $TSLA , $NVDA , and not much else on the hardware side. As always I prefer to play the picks-and-shovels under it, the screws, the rare earths, the grinding-machine bottleneck, get paid whether or not Optimus or Figure is the one that scales.
1
1
3
952
Nicolas retweeted
Rocket Lab is acquiring Iridium Communications Inc – one of the most transformative deals in the space industry. By combining our launch capability and satellite manufacturing with @IridiumComm’s global satellite communications network and rare spectrum, Rocket Lab becomes a fully integrated, self-launching, tier-1 space power, delivering critical communications capability to millions of users worldwide. Full details and important information: bit.ly/4eyKLL4
299
1,084
396
7,378
1,430,754