Owns nodes (doesn't know it yet)
Market Context: The Third Storm
Residential solar is in a capital crisis. Three storms in sequence: high interest rates compressed margins and killed refinancing assumptions. Net metering rework (NEM 3.0 in CA, changes across PJM states) cut the revenue side at the same time — and the industry didn't understand storage, ancillary services, or compute well enough to replace the lost value. Now storm three: there isn't enough tax equity or tax credit transfer capital to absorb the enormous residential throughput the industry built capacity for.
The companies that died — Sunnova, SunPower, and others — were largely lemmings selling forward way too much cash, unable to weather any storm, let alone three. The survivors are the ones who didn't overleverage their forward books. But none of the survivors are thinking about what their nodes actually are beyond energy assets. Nobody sees the ground terminal. Nobody sees the compute. Nobody sees the connectivity layer.
The best owner of a residential solar project hasn't bought one yet in material numbers. That owner is converging from another sector — telecom, space, cell tower, infrastructure REIT, or SFR — and sees the node as more than energy.
The Eight Layers of the Stack
Layer 1 · Constellation Operators
Who owns the satellites
They need ground infrastructure desperately — and are spending billions to avoid building it themselves. Every one of them is engineering around ground density because building is expensive.
SpaceX / Starlink
LEO · 9,400+ sats · 10M+ subs · 650+ D2C sats
Dominant LEO operator. Gen2 (mid-2027) promises 100x data density per satellite. T-Mobile D2C partnership. Ground gateways distributed globally but performance scales with ground station density — and they don't have enough. 25M users targeted by end 2026.
AST SpaceMobile
LEO · D2D · 50+ MNO partners · $70.9M 2025 rev
Only space-based cellular broadband for unmodified phones. BlueBird 6 deployed (2,400 sq ft array, 120 Mbps). 45-60 sats by end 2026. Partners: AT&T, Verizon, Vodafone, Rakuten, Google, American Tower, TELUS. TELUS investing in Canadian ground infrastructure + taking AST equity. $3.9B pro forma cash. $1B revenue target 2027.
SES + Intelsat (merged July 2025)
GEO + MEO · 120 satellites · $4.1B combined rev
Largest traditional operator post-merger. 90 GEO + 30 MEO. Investing in D2D through Lynk Global (SES is major strategic shareholder). 2026 is "build year" — flat revenue, focused on integration synergies.
Eutelsat / OneWeb
GEO + LEO · 600+ OneWeb sats · $3.47B debt
First integrated GEO-LEO operator. Tried to sell ground stations to EQT for €790M — France blocked it on sovereignty grounds (Jan 2026). Now stuck with ground infrastructure it tried to shed. 340 additional OneWeb sats on order from Airbus. Working with Clear Blue Technologies on power-efficient ground services.
Amazon Kuiper
LEO · 102 sats launched · 3,200 planned
AWS edge compute integration is the differentiator — on-orbit AI data processing nodes planned. Australia service mid-2026. Vertically integrated: constellation + compute + logistics + balance sheet.
Lynk Global (merged w/ Omnispace)
LEO · D2D · SES-backed
D2D via small satellites. Routing through MEO/GEO inter-satellite links to avoid ground density. Engineering around the ground problem rather than solving it.
Viasat / Equatys
GEO · D2D shared infrastructure pitch
Post-Inmarsat acquisition. Leading Equatys — pitching shared infrastructure approach to D2D as counter to Starlink. Neutral-host D2D concept. Structurally similar to what RedX does on the ground.
Globalstar / Apple
Emergency SOS · $1.7B Apple investment
Emergency SOS, Find My on iPhone 14+. Apple invested $1.7B for 17 new satellites. Narrow use case but massive installed base. iOS 18 expanded to broader messaging.
Layer 2 · Ground Station Infrastructure
Who owns the ground nodes connecting orbit to earth
The bottleneck. Market is $11.76B in 2026 (8.7% CAGR). Trends: Ground-Station-as-a-Service (GSaaS), software-defined ground systems, multi-orbit. Fragmented market. This is where RedX plays.
RedX (the concept)
Distributed solar-powered ground terminals
Every solar site is a ground terminal. Power, battery, permits, customer relationship already exist at every site. Incremental hardware only. Multi-tenant model (cell tower economics). The only entity starting from existing powered, permitted infrastructure in underserved markets. Nobody else is doing this.
EQT / SatPort Infrastructure
PE · operator-neutral GSaaS platform
Tried to buy Eutelsat's 1,400 antennas for €790M. France blocked it. Proceeding with SatPort as standalone — will build or buy ground stations independently. Not a thesis-definer — a demand signal for what PE thinks ground infrastructure is worth. EQT is mostly oil and gas guys who identified ground stations as "attractive digital infrastructure vertical" but aren't connecting it to their energy portfolio.
SSC / Leaf Space / Atlas Space / Others
Legacy and GSaaS startups
Growing ecosystem. SSC launched SSC Go for small sats. Software-defined, cloud-native architectures emerging. Fragmented market ripe for consolidation — the M&A play EQT wanted.
Layer 3 · Tower & Site Infrastructure (TowerCos)
Who owns physical sites and leases capacity to multiple tenants
The cell tower REITs pioneered multi-tenant infrastructure. Now converging into satellite and edge compute. The structural analog to RedX — but starting from macro towers, not rooftops. Different cost structure per node.
American Tower
43K US towers · CoreSite $10.1B · EdgePoint $5.2B
Most aggressive convergence play among incumbents. CoreSite (28 data centers), EdgePoint, IBM edge cloud partnership, AST SpaceMobile strategic partner. CTO explicitly says infrastructure class boundaries are blurring. Sees itself as digital infrastructure company now. But starting from macro towers — fundamentally different per-node economics than residential rooftops.
SBA Communications
~34K US towers · 30-50 edge sites
Edge computing facilities at tower bases. Less aggressive than AMT on convergence but investing. Primary demand: edge compute and fiber regeneration.
Crown Castle
~40K US towers · fiber focus
Publicly not interested in data center market. Focused on fiber and small cells. Least convergent of the big three. Potential laggard.
Landmark Infrastructure Partners
Ground leases under telecom assets
Ground lease model — owns the land under towers. Same real property attachment model that applies to rooftop solar. Closest structural analog to ground lease concepts in the energy stack.
Layer 4 · Edge & Orbital Compute
Who processes data at the edge and in orbit
AI demand is pushing compute to the edge and into orbit. Ground nodes with power and connectivity are ideal edge compute hosts. Constellation operators want edge preprocessing at their ground nodes — they'll bring or fund the compute.
Alphabet / Google
Project Suncatcher · orbital AI · AST partner
Project Suncatcher (Nov 2025) for orbital AI data centers. Planet Labs partnership on in-orbit computing. Google is also an AST SpaceMobile strategic partner. Connecting compute + connectivity + orbit.
Microsoft / Azure Space
Azure Orbital · SpaceX partnership
Azure Space provides global cloud access through Starlink. Azure Orbital in preview. American Tower partnership for edge cloud. Testing on-orbit deployment for USG.
Amazon / AWS + Kuiper
Edge compute + LEO constellation
Owns both constellation (Kuiper) and compute platform (AWS). Vertical integration advantage. But ground stations are centralized, not distributed.
Starcloud · Axiom Space · EDGX · OrbitsEdge
Orbital compute startups
Starcloud: orbital GPU data centers, NVIDIA partnership. Axiom: ISS-based ODC nodes with Red Hat. EDGX: satellite edge AI, in-orbit 5G/6G base stations. All solving the same problem: compute closer to data generation. All need ground nodes.
Layer 5 · D2D & Last-Mile Connectivity
Who delivers broadband to the end user
D2D is the technology that makes the RedX model work — the connectivity floor at every home without wires. Starlink has 10M D2D users. AST has 50+ MNO partners covering ~3B subscribers. Gen2 Starlink (mid-2027) promises terrestrial-like 5G broadband from orbit.
This layer is well-covered in the RedX ground terminal deck. The key point: D2D is the floor that works everywhere. Terrestrial is the bonus. The hierarchy adds capacity, never creates dependency.
Layer 6 · Energy Infrastructure (Solar + Battery + Grid)
Who provides the power — and who owns nodes without knowing it
Every ground terminal, every edge compute node, every antenna needs power. Solar + battery provides it without grid dependency. Nobody else in the convergence stack starts here. The entities below own surgical-scale distributed assets — Texas 10 batteries, sub-10 MW A/C projects, QFs in PJM, residential rooftop portfolios — that are ground terminals waiting to be activated.
EQT Energy Portfolio
Cypress Creek 3.2 GW · Madison Energy 396 MW · Scale Microgrids 250 MW
Three separate portfolio companies across two fund strategies. Cypress Creek: utility-scale + distributed solar/storage across 200+ locations. Madison Energy: C&I and community solar. Scale Microgrids: 250 MW, one of the largest microgrid portfolios in the US, serving data centers, C&I, agriculture. EQT also tried to buy Eutelsat's ground stations in a separate fund. They have the pieces but aren't connecting them. Buying solar and ground stations as separate theses.
HASI (Hannon Armstrong)
$6.2B managed assets · REIT · DG + resi + land
Started as a land-and-lease play under solar fields but now owns substantial DG and residential assets, cash flows, and servicing revenues. Largest landowner of solar fields. REIT structure gives tax advantage. Recent $1B green notes offering. 71 MW residential portfolio with IGS Solar (25-year leases, 11 states). Invests across equity, JVs, land ownership, lending. One of the few entities with both the capital structure and the asset base to see the convergence — if anyone points it out.
Altus Power
C&I solar · TPG Rise Climate · $2.2B acquisition
Long-term owner/operator of commercial-scale solar. Just acquired 237 MW from Greenbacker (100+ projects, 18 states). TPG Rise Climate Transition Infrastructure portfolio company ($2.2B, Feb 2025). TPG likely isn't thinking about ground terminals or compute — same blind spot as Sunrun. But they own the nodes.
Aspen Power
Distributed solar + storage · $200M Deutsche Bank
Developer, owner, operator of distributed solar and storage. $200M commitment from Deutsche Bank to accelerate deployment. Surgical-scale distributed assets across the US. Owns the nodes.
Palmetto
Residential TPO · $716M ABS issuances · 22K+ PPAs
Survived the storms. Two ABS deals in 2025 totaling $716M backed by 22,000+ residential solar PPAs and leases. Operating as TPO provider post-ITC residential expiration through LightReach lease structure (claims commercial ITC). Owns the residential nodes at scale.
Sunrun
Largest residential TPO · ~1M systems
Largest residential solar TPO provider in the US. Owns the most residential nodes of anyone. Not thinking about connectivity, compute, or ground terminals at all. Focused on surviving the capital crisis and defending net metering economics. Their rooftops are the single largest untapped infrastructure — but they don't see it.
SolarREIT / Solar land lease operators
Land ownership under solar assets
Same structural concept as Landmark Infrastructure but for solar. Own the land, receive rents senior to project debt. REIT-eligible. The ground lease layer that makes the real property attachment work.
Surgical-scale assets that are ground terminals waiting to be activated
A Texas 10 battery (10 MW BESS in ERCOT). The smaller equivalents in California (SGIP-funded behind-the-meter storage) or New York (VDER + storage). Speed-to-market QFs in PJM — qualifying facilities under 10 MW A/C that can interconnect without years of queue delay. Every one of these assets already has: power, grid interconnection, permits, and a site. Adding a satellite terminal and compute box is incremental hardware on existing infrastructure.
The entity that controls these nodes holds the leverage over every layer above it in the stack.
The dead and distressed
Sunnova
Dead · filed bankruptcy 2024
Sold forward too much cash. Couldn't weather interest rates + net metering rework + tax equity scarcity. Lemming.
SunPower
Dead · filed bankruptcy 2024
Same story. Over-leveraged forward book. Two decades of brand equity destroyed by capital structure that couldn't survive any storm.
Titan Solar / Pink Energy / others
Dead or distressed
Smaller operators that couldn't weather the sequential storms. Their orphaned assets and stranded customers are exactly who Solar Blue Book serves — homeowners stuck in TPO contracts with dead sponsors.
Layer 7 · Capital & Tax Credit Finance
Who provides the money — and who understands ITC on satellite hardware
Tax equity and transfer credit capital is the binding constraint right now. Not enough to absorb residential throughput. Orbit.Tax sits here — the ITCs on satellite terminal hardware and compute equipment haven't been analyzed by anyone in the constellation ecosystem. Those credits can fund the ground buildout with gravy left over.
Orbit.Tax
ITC analysis · satellite + energy components
Handles ITC analysis and placement across energy and satellite components. Constellation operators aren't thinking about ITCs on their satellites. Those credits can fund the ground buildout. The financing layer that makes RedX capital-efficient.
HASI
REIT · tax equity · structured equity · land
The most sophisticated clean energy capital provider. REIT structure, tax equity, structured equity, land ownership, servicing revenues. If anyone in the capital stack could see the convergence, it's HASI — they already own the assets, the cash flows, and the land.
Traditional tax equity (JPM, BofA, US Bank, etc.)
Insufficient capacity for resi throughput
The big banks that have always provided tax equity for solar. There isn't enough capital to absorb current residential volume. Transfer credits (Inflation Reduction Act §6418) were supposed to expand the buyer base but haven't scaled fast enough. This is the binding constraint on the entire residential market.
Layer 8 · SFR Aggregation Platforms
Who owns the rooftops at scale — and doesn't know it yet
The institutional single-family rental operators. They own 50,000-500,000 homes. Zero customer acquisition cost. Templated floor plans. Volume pricing. Zero solar penetration. The Multi² thesis: every home is simultaneously a power plant, compute node, and ground terminal. Vacancy-proof NOI. But they're thinking about rent rolls and cap rates, not infrastructure.
Invitation Homes / AMH / Progress Residential
50K-80K homes each · institutional SFR
Largest SFR operators. Own the rooftops, the tenant relationship, identical floor plans. None deploying solar + connectivity at scale. The pitch: $15-25/mo per home in new energy margin. $18-30M/yr incremental NOI at 100K homes. $300-600M asset value uplift at 5-6% cap. Plus ITC, SREC, grid services, battery arbitrage, compute during vacancy.
PE-backed SFR (Blackstone, KKR, Carlyle)
100K+ homes via portfolio cos
Control massive SFR portfolios through operating companies. Infrastructure mindset exists. The pitch is structural: layered cash flows make NOI resilient to vacancy and nonpayment. The subprime dynamic — where a vacancy spiral collapses the asset — is mitigated by infrastructure revenue that doesn't depend on occupancy.
Who Is Converging — and What Nobody Sees
Everybody is engineering around ground density because building it is expensive.
Lynk routes through inter-satellite links. AST needs only a few gateways per country. Starlink's D2C performance varies by ground station proximity. EQT tried to buy Eutelsat's ground stations for €790M and was blocked. American Tower is pivoting from towers to digital infrastructure. Nobody is starting from the residential rooftop with power already in place.
American Tower → Closest incumbent convergence
Towers → CoreSite ($10.1B) → EdgePoint ($5.2B) → IBM edge cloud → AST SpaceMobile partner → satellite gateway hosting. CTO says boundaries between infrastructure classes are blurring. But macro towers, not distributed residential. Different cost structure. Different markets.
EQT → Has the pieces, doesn't see the picture
Owns Cypress Creek (3.2 GW solar), Madison Energy (396 MW DG), Scale Microgrids (250 MW, data center customers). Also tried to buy Eutelsat ground stations and created SatPort Infrastructure. Two separate fund strategies buying solar nodes and ground stations without connecting them. Oil and gas DNA. The demand signal is real — the strategic vision isn't there.
AST SpaceMobile → Needs distributed ground, relies on MNO partners to build it
Doesn't own ground infrastructure — partners (AT&T, TELUS, Verizon) build gateways. TELUS investing in Canadian ground infra + taking AST equity. RedX sites are in exactly the markets AST targets: broadband deserts, underserved LMI communities. Natural ground infrastructure partner.
Amazon → Full vertical integration, centralized ground
Kuiper + AWS + logistics + balance sheet. Most vertically integrated. But ground stations are centralized, not distributed. Not thinking about residential rooftops or solar power at the edge. Different architecture entirely.
The Gap
Every entity above is approaching the problem from one direction — orbit, towers, compute, or capital. Nobody is starting from the powered ground node and building up.
The convergence thesis is not that RedX competes with Starlink or American Tower. It's that RedX provides the missing ground layer that every constellation operator needs and none of them want to build. The powered, permitted, distributed infrastructure platform — cell tower REIT economics applied to orbit, starting from solar.
The entities that own surgical-scale distributed energy nodes — a Texas 10 battery, a QF in PJM, a residential rooftop portfolio — hold leverage over every layer above them. They just don't know it yet.
The solar pays for everything. Everything else is gravy.
Step 1 (solar + battery) pencils on its own. Step 2 (D2D terminal) eliminates the wire and splits cost three ways. Step 3 (broadband) transforms host acquisition. Step 4 (ground terminal) is multi-constellation revenue — pure upside. Compute is the force multiplier — it mitigates vacancy risk, it's what operators want beyond antenna space, and it pays for much of the incremental infrastructure.
Solar Blue Book proves the asset value. The convergence landscape proves the demand. The SFR thesis proves the aggregation path. Redbase orchestrates the whole thing at scale.