PNRusty
5.6K posts

Tweet fijado
PNRusty retuiteado

Let's talk about space-based AI data centers.
$RKLB recently dropped a game-changer: advanced silicon solar arrays designed specifically to power gigawatt-scale space-based data centers orbiting Earth.
Announced in late February 2026, these arrays target the exploding demand for AI/compute power that's overwhelming terrestrial data centers (think land shortages, massive water use for cooling, and grid constraints).
Why Space-Based Data Centers + Rocket Lab's Solar Tech Could Be Huge In orbit, you get:
Infinite, uninterrupted solar energy (no night, no clouds, constant exposure).
Natural extreme cold for free cooling (no evaporative towers or chillers needed).
Gigawatt-scale potential — arrays that can span kilometers and deliver massive, scalable power at low cost per watt.
Rocket Lab positions this as the "next frontier" for computing infrastructure, especially as AI training/inference eats gigawatts on the ground.
Their silicon arrays are mass-manufacturable, lightweight, modular, and radiation-hardened — key for scaling orbital mega-structures economically.
This ties perfectly into RKLB's vertical integration: they already lead in high-efficiency GaAs/Ge solar (world's largest capacity), acquired SolAero for in-house cells/panels, and now add silicon to reduce reliance on geopolitically sensitive materials like gallium/germanium.
Lower supply chain risk + industrial-scale production = cheaper, more resilient power for future orbital data centers:
Competition LandscapeSpace solar isn't new, but gigawatt-scale for data centers is emerging/niche
- Traditional players focus on high-efficiency multi-junction cells (GaAs-based) for satellites/NASA missions — companies like Spectrolab (Boeing), Azur Space, or SolAero (now RKLB).
- Emerging silicon-focused competitors: Solestial (thin-film silicon PV for space, self-healing tech).
- Broader space power:
Northrop Grumman, Lockheed, Airbus Defense/Space — but mostly for traditional sats, not explicitly gigawatt orbital compute.
- What about Space X?:
For Starlink satellites, SpaceX has used silicon-based solar cells from suppliers like Taiwan Solar Energy Corp (TSEC), opting for cheaper, terrestrial-grade cells rather than specialized space-grade ones (like GaAs) to keep costs low for mass production. They assemble or integrate the panels themselves at facilities like Redmond, Washington, but the core solar cells are imported/purchased.
SpaceX requested authorization from the FCC for a constellation of up to 1,000,000 solar-powered data center satellites in low Earth orbit. That's a lot of solar arrays SpaceX will need!
That doesn't even include planned satellites from Amazon, Google, Microsoft and Facebook.
No one else is publicly pushing silicon arrays at this scale for data centers like RKLB is.
RKLB's edge? Full vertical integration (cells → arrays under one roof), rapid production scaling, and synergy with their launch cadence (Electron/Neutron for deploying these massive structures).
- Advantages of RKLB's Silicon Solar ArraysCost: Much lower $/watt at scale vs. traditional space-grade multi-junction cells.
- Scalability & Manufacturing:
Modular, mass-producible — suits kilometer-spanning data center arrays.
- Supply Chain Resilience: Avoids rare/critical minerals; silicon is abundant/domestically sourced.
- Future-Proof: Complements their existing high-eff GaAs lines (hybrid options possible).
- Timely: Aligns with AI power crunch + satellite industry growth (projected 7x by 2035).
Disadvantages / RisksEfficiency Trade-Off:
- Silicon cells typically lower efficiency than GaAs/Ge multi-junction (though space-optimized versions help; still may need larger arrays for same power).
- Tech Maturity: New silicon line for extreme space use — radiation hardening/longevity in orbit needs proving at gigawatt scale.
- Market Readiness: Space-based data centers are conceptual/futuristic (latency issues for real-time apps, data transfer challenges, launch costs).
Hype-driven? Some see it riding AI/data center buzz without near-term deployments.
- Execution Risk: Adds to RKLB's plate (Neutron rocket dev, etc.) — manufacturing/tech risks on top of existing ones.
Overall, RKLB's move smartly positions them in the intersection of space power + orbital compute boom. If space data centers take off (even partially), this could be a massive tailwind for RocketLab beyond launches/sats.

English

Too early to celebrate again? $RKLB $ASTS
PNRusty@PNRusty
Happy Capitulation Day for those who celebrate $RKLB $ASTS
English
PNRusty retuiteado
PNRusty retuiteado

$BKSY: 🛰️ Earth Observation is finally coming into focus. AI-driven demand and the conflict in Iran have highlighted the importance of this market. Real world applications have finally materialized causing an inflection in demand where BlackSky's backlog increased to $345M and Planet's backlog increased to $900M. $PL business momentum has prompted me to revisit and take a spec position in $BKSY which I used to own in previously.
Enterprise Value:
$BKSY: $1B
$PL: $12B
BlackSky has demonstrated strong financial performance, however its results have been inconsistent as 97% of its customers are governments and militaries, which leads to results being very lumpy. 2026 should be a stronger year as additional Gen-3 satellites are deployed.
BlackSky primarily serves government/military customers with 49% of its revenue coming from US Government / Agencies / Military, 48% from international Governments and the remaining 3% from commercial customers.
Here's how $BKSY compares to $PL financially:
$BKSY vs. $PL Revenue & YoY Growth:
2026E: $133M 24% | $380M 23%
2027E: $170M 28% | $482M 27%
2028E: $215M 26% | $594M 23%
2029E: $264M 23% | $727M 23%
$BKSY vs. $PL Gross Margins & EBITDA:
2026E: 68% $11M | 51% $13M
2027E: 71% $40M | 55% $58M
2028E: 73% $60M | 63% $135M
2029E: 73% $85M | 67% $189M
$BKSY vs. $PL EBITDA Valuation Multiples
2026E: 92.9x | 920x
2027E: 25.5x | 201x
2028E: 17x | 87x
$BKSY trades massive EBITDA multiple discount to $PL while having comparable topline growth and much better margins. That said, Planet does have the upside of AI Data Center potential in partnership with Google and it has reached a higher level of financial scale.
However given all the interest in enabling military applications for space, $BKSY wins here with its massive exposure to US and Foreign (US friendly) governments. BlackSky should close the valuation gap with Planet in the coming months as it completes its Gen-3 constellation and signs more government contracts.
Short interest is 7.8M shares or 26% of the float. I estimate half of short interest is from convertible arb hedging from the $185M convert raised in July 2025. The rest are fundamental shorts 🔥
Finally, IMHO I think $RKLB should just buy $BKSY if they're looking to expand into attractive space vertical that leverages their capabilities = Earth Observation. $RKLB already launches $BKSY sats and it would be a highly accretive transaction🤷♂️

English
PNRusty retuiteado

@elonmusk @SpaceInvestor_D you couldnt have casually thrown a $RKLB in this one? 🤣
English


@Gykiwi03 @Forfuture100 I was just going to comment about AZ being the Geost home. I like it.
English

@Forfuture100 💯 Filled an important role, one of many AI/machine learning positions RKLB have been hiring throughout 2025/26 for different areas.
Arizona the home of GEOST and Optical Support Inc.
English

$RKLB 🚀 She should’ve listened.

Rocket Lab@RocketLab
Our largest launch contract yet: 20x new HASTE launches for @DeptofWar to accelerate hypersonics for America. We're delivering reliable, modern hypersonic capabilities to the nation with speed and affordability - and the first mission in this block of launches is just months away🚀
English

@johhnyWalkerAZ My plan was always to save up for @anduriltech IPO but it's going to be too dang expensive when that time finally comes, unfortunately. x.com/PNRusty/status…
PNRusty@PNRusty
My $RKLB position is locked in and staying put (unless a significant dip--then I'm adding) 2025 will be about adding to my $ASTS position 2026 will be about growing my ammo for an @anduriltech IPO
English


Me, the granddaddy of $RKLB, watching people who just found out about RKLB yesterday come lecture me on the company and its management.
What was the symbol before it de-SPAC'd? Did you know there was such a thing as warrants? Did you convert warrants into shares and hold through the ups and downs?
Did you believe in the management when the stock was trading in the dumpster and the rocket launch failed? Were you buying shares in the $3 handle? Were you watching every single attempt to catch Electron mid-air?
So spare me your lecture on how great the company is when I'm simply pointing out that there could be better timing for the offering.
Now I want you to be angry .... very angry . Bring all that rage and come at me bro!
🤣
GIF
English




















