Richard Prati

$ALMU The market is still pricing Aeluma like a science project. That’s the mistake. The bottleneck in photonics has already shifted: from physics → to manufacturing. Most investors are still focused on “does it work?” The real question is: Can it scale inside real foundry environments? That’s where companies fail. And that’s exactly where Aeluma is quietly building: • Tower Semiconductor (manufacturing) • Sumitomo Chemical (materials) • Ecosystem + process hires across the stack This is no longer a lab story. It’s an execution story. The market usually recognizes that shift late. Full piece: pratimanagement.substack.com/p/aeluma-the-b… #ALMU #Semiconductors #SiliconPhotonics #AIInfrastructure

$ALMU just hit #5 in MarketBeat’s “5 Stocks to Buy in May Before the Next AI Surge Hits.” Quantum dot lasers are solving AI’s biggest data transmission bottleneck — fast, efficient, low-heat, and now scaling with fresh government contracts. The shift from science project → repeatable manufacturing is happening.MarketBeat article → marketbeat.com/stock-ideas/5-… My deeper dive (why this matters now) → #ALMU" target="_blank" rel="nofollow noopener">pratimanagement.substack.com/p/aeluma-the-b… #AIInfrastructure #SiliconPhotonics

$ALMU named #5 in MarketBeat’s new list: “5 Stocks to Buy in May Before the Next AI Surge Hits.” They highlight Aeluma’s quantum dot laser platform for solving AI data transmission — fast, efficient, low-heat, and now scaling toward mass-market production with fresh government contracts already funding the ramp. The bottleneck conversation is moving exactly where we’ve been saying it would: from science to repeatable manufacturing. Full MarketBeat piece → marketbeat.com/stock-ideas/5-… Substack (published last week) on why this transition matters → #ALMU" target="_blank" rel="nofollow noopener">pratimanagement.substack.com/p/aeluma-the-b… #AIInfrastructure

Most people still think this is about better devices. It’s not. It’s about whether this can actually be manufactured at scale inside a real ecosystem. That’s where almost every photonics company fails. This is what Aeluma is trying to solve.

$ALMU The market is still pricing Aeluma like a science project. That’s the mistake. The bottleneck in photonics has already shifted: from physics → to manufacturing. Most investors are still focused on “does it work?” The real question is: Can it scale inside real foundry environments? That’s where companies fail. And that’s exactly where Aeluma is quietly building: • Tower Semiconductor (manufacturing) • Sumitomo Chemical (materials) • Ecosystem + process hires across the stack This is no longer a lab story. It’s an execution story. The market usually recognizes that shift late. Full piece: pratimanagement.substack.com/p/aeluma-the-b… #ALMU #Semiconductors #SiliconPhotonics #AIInfrastructure

This is a strong breakdown. $ALMU is executing. The key point isn’t just who Aeluma hired: it’s what the hire says about where management believes the risk now sits. You don’t bring in a #15 Intel inventor with deep experience (350+ patents) in III-V on silicon integration, process technology, and corporate strategy to prove the physics. You bring them in when the bottleneck has shifted to execution: foundry alignment, heterogeneous packaging, yield at scale, and pulling real demand into production. That’s where most photonics platforms historically stall. Solve that layer, and the value moves from “promising device” to infrastructure. The market usually recognizes that transition late. (And it aligns with the broader pattern — process and ecosystem hires back-to-back.)

Aeluma $ALMU just appointed Dr. Willy Rachmady as VP of Strategic Partnerships. Who is he? and why? He is ranked #15 out of 30,000+ Intel inventors of all time, with 350+ U.S. patents and a 3x Intel Top Inventor Award. He previously spent 20 years at Intel. We talked one month ago (link below 👇) about the fact that $ALMU was hiring to scale. Today we'll give you more insight on this hire — but we can already say: you don't hire the #15 inventor in Intel's history, whose job was to pick winning technologies, to do more lab research. You hire him to execute. More technical details: His patents cover the exact techniques Aeluma is commercializing: growing special semiconductor materials inside trenches on standard silicon, which is precisely what Aeluma does with their III-V-on-silicon platform. What is also notable is that his last role at Intel was in Corporate Strategy & Ventures. His literal job was to scan the global semiconductor landscape, identify which technologies were worth betting on, and decide whether Intel should fund or partner with them. He was essentially the kind of person who would have been evaluating Aeluma from the outside. What is he actually going to do at Aeluma? His mandate covers three areas: 1. Frontend & backend fabrication, wafer-scale integration, and packaging: "Frontend" is where circuits are etched onto a silicon wafer; "backend" is where chips are cut, assembled, and packaged. Aeluma has the technology and one of his taska could be to plug it into Tier-1 factories like $TSM or $INTC Foundry for mass production. 2. Driving commercial opportunities for III-V image sensors, high-speed photodiodes, and quantum dot lasers: These are Aeluma's three core products — sensors that see in the dark, photodiodes that move data at 100–300+ Gbits/s for AI data centers, and lasers that power it all. His job is to get them into Tier-1 buyers' hands. 3. Co-packaged optics (CPO) and CMOS integration: today, the chip that processes data and the optical components that move it are physically separate and connected by wires that waste power and create bottlenecks (we talked previously about the wire bottleneck x.com/the_fat_pitch/…). CPO puts them in a single package, reducing power consumption by 30–50% and dramatically increasing bandwidth. We hold $ALMU.

Aeluma vs. POET -- same space, very different posture: • Aeluma → understated, builds before signaling • POET → promoted early, pulled forward expectations • Aeluma → credibility intact, still proving scale • POET → credibility hit after order cancellation In photonics, failure isn’t performance; it’s scaling and trust. One is still proving it. The other now has to rebuild it. $ALMU $POET

Aeluma didn’t just hire “an Intel guy.” They hired for the bottleneck. Photonics hasn’t been limited by performance; it’s been limited by manufacturing. Now the constraint has shifted: • Tech → proven • Bottleneck → integration + scale This role is about forcing alignment across foundry, materials, and packaging. That’s a phase change. $ALMU aeluma.com/investors/news… #SiliconPhotonics #Semiconductors #AIInfrastructure

$ALMU at ~$350M is still priced like a science project. The market is wrong on timing. The bottleneck was never performance—it was manufacturing. Now: • Tower = foundry validation • Sumitomo = materials scale • NASA = funding the failure points This is no longer “does it work.” It’s “can it scale?” That shift is already happening. The market hasn’t caught up. seekingalpha.com/article/489339… #ALMU #SiliconPhotonics #AIInfrastructure

If you wait for “confirmation” on $ALMU, you’re the exit liquidity. By the time: • design wins are public • qualification is complete • revenue is visible The multiple is already gone. This phase is where the bet is made.

$ALMU just added NASA to the stack. Now you have: • Tower → manufacturing • Sumitomo → materials • U.S. Gov → scaling • NASA → integrated laser validation That’s not random. That’s the supply chain forming. This is no longer a science project. It’s an execution story. Most fail there. That’s the opportunity. Link: aeluma.com/news-media/pre… #ALMU #SiliconPhotonics #AIInfrastructure

The market still thinks AI optics is a technology race. It’s not. It’s a manufacturing bottleneck: • Packaging • Qualification • Multi-fab portability That’s where winners separate. pratimanagement.substack.com/p/aeluma-just-…

Most investors still frame this as a technology question. It’s not. It’s an execution filter: Can it be manufactured? Can it be qualified? Can it be repeated at scale? That’s what determines who actually captures value. This deepens the thread. Makes you look serious.

Aeluma just announced $4M+ in contracts with Tower + Sumitomo. This is what progress looks like at this stage: → non-dilutive funding → deeper manufacturing alignment → early steps toward scale The opportunity here has never been about proving the tech. It’s about proving it can scale: • packaging • qualification • multi-fab portability That’s the bar. If they clear it, the narrative converts to infrastructure. If not, it stays early-stage. Full framework (Part 5): pratimanagement.substack.com/p/the-next-bot…

Most companies break here: • Yield collapses under packaging • Qualification timelines extend • Systems stop being repeatable That’s where narratives fail. Full breakdown: pratimanagement.substack.com/p/aeluma-just-…

AI optics doesn’t fail on performance. It fails at scale. Performance wins demos. Manufacturing wins markets.

If you wait for confirmation, you are the exit liquidity. Aeluma just named Tower + Sumitomo. That’s not validation. It’s the phase where scale gets tested—where most of these platforms break. pratimanagement.substack.com/p/aeluma-just-…

Here's Part 5 - Scale is the Constraint pratimanagement.substack.com/p/the-next-bot…

AI infrastructure isn’t failing on performance. It’s failing at scale. Packaging. Qualification. Manufacturing. That’s where optical platforms break. Here’s what survives — and what doesn’t: Performance wins demos. Manufacturing wins markets.

Aeluma Secures More Than $4 Million in Contracts for Quantum Materials and Lasers The more interesting aspect of the announcement isn’t the dollar amount, but the counterparties and what they imply. Tower Semiconductor being explicitly named as a manufacturing partner is a step beyond prior generic foundry language. It suggests that foundry-transfer work is progressing, not just internal development. Similarly, having Sumitomo Chemical Advanced Technologies referenced points to early formation of a materials supply chain, which is a necessary precursor to any credible commercialization path. Tower’s commentary around enabling laser manufacturing on large silicon wafers is notable as well—it reflects external engagement with the core thesis, though it should still be viewed as directional rather than definitive validation of scalability. What this collectively indicates is movement toward a more structured manufacturing ecosystem—foundry alignment, materials sourcing, and process integration becoming more visible. That said, this remains pre-commercial. These signals support progress, but they don’t yet answer the key question around repeatable yield, packaging at volume, and multi-fab portability. aeluma.com/news-media/pre…