Energy is abundant.
Delivery is not.
HAYSOLAR builds autonomous, behind-the-meter sites in Armenia where AI operators install their own GPUs. We supply the generation, storage, cooling and uptime — contracted per kilowatt, with no interconnection queue to wait on.
Four things to know before you read the rest of this page
Power, not chips, is the constraint
~30–50% of new global data center capacity is now being built as on-site generation because grid interconnection queues run 4–8+ years in mature markets. IEA · Cleanview · LBNL, 2026
Armenia is already a proven AI-compute destination
A $4B, 50,000-GPU NVIDIA-backed AI megaproject (Firebird) is under construction in Armenia right now, with U.S. and Armenian government backing. HAYSOLAR is a different category — see below.
The unit economics are public and verifiable
Local consumption captures ~53 AMD/kWh in displaced retail tariff vs. 18–42 AMD/kWh for grid export. That spread is the input cost advantage we sell hosted capacity on. ENA · PSRC, 2026
We sell the power envelope, not the hardware
Tenants bring, own and insure their own GPUs. Our capital goes into generation, storage and the site — the fastest-depreciating asset in the stack stays off our balance sheet.
Why a solar company stopped selling electricity
HAYSOLAR began as a renewable energy company developing grid-connected solar assets in Armenia. We hit a structural wall: the economics of renewable generation are set by regulated tariffs and utility policy, not by the cost of producing the energy. Pre-2020 stations sell under fixed feed-in tariffs; post-2022 stations sell into an open market with no price floor at all. Either way, someone else prices your export.
The highest-value electron is the one consumed locally.
A kilowatt-hour used behind the meter is worth what the consumer standing on top of it would otherwise pay the grid — not the export price it would otherwise fetch. So instead of exporting electricity, HAYSOLAR builds the place where that electron is consumed on the spot: a solar-plus-storage site whose entire load is AI hardware, installed and owned by a tenant who pays for capacity by the kilowatt.
The spread is the business case
We're not asking investors to take a macro thesis about AI power scarcity on faith. The spread between what a business pays for grid electricity and what it's paid for exporting solar power is public, regulated, and verifiable today.
Two paths for the same electron
One path sells the electron into a market that prices it as a commodity. The other keeps it on site, under a tenant's hardware, and sells it as contracted kilowatts of hosting capacity — at the full value of the tariff it displaces.
We do not sell GPU hours.
We sell hosted capacity.
Bring your own GPUs. We bring the power.
A behind-the-meter site in Armenia with no interconnection queue, priced per kilowatt — not per electron, and not per GPU-hour.
Contracted capacity
- Power at displaced-retail economics. Generation sits on site. You are not paying a transmission tariff and you are not waiting in a queue.
- Reserved kW, not spot allocation. Defined uptime, priced monthly on a multi-year term.
- Solar plus storage. BESS smooths intermittency so utilisation holds overnight, with grid backup for continuity.
- Site-appropriate cooling. Thermal design built for Armenian climate and elevation, not retrofitted from a warehouse.
- You own the hardware. Your GPUs, your stack, your customers. We do not compete with you for compute revenue.
- Portability. Defined step-in, termination and removal terms — your assets are not trapped on our site.
The commitment
- A multi-year capacity reservation with a minimum monthly commitment
- Hardware delivered, insured and owned by you
- A ramp schedule tied to our construction milestones
First-phase available kilowatts, price per kW-month, contract length and contracted uptime are issued in the capacity term sheet together with the site engineering package. We publish those numbers once they are engineered, not estimated.
Why put hardware on a solar site instead of in a data centre
Where grid colocation still wins: carrier density, redundancy tiering, and anything past a few hundred kilowatts under one roof. If you need five megawatts in one hall next quarter, we are not the answer.
Five layers. We own four.
Generation, storage, cooling and control are designed as one system rather than procured as separate contracts with separate counterparties. The compute layer sits inside that envelope and belongs to the tenant.
Solar
On-site generation, sized to the contracted load rather than to export capacity.
BESS
Battery storage to flatten intermittency so contracted capacity stays available after dark, not only in daylight hours.
GPU — tenant
Compute hardware installed at the point of generation, delivered, owned and insured by the tenant. Our scope ends at the power and thermal envelope.
Cooling
Site-specific thermal management, sized for Armenia's climate and elevation profile.
Control Layer
Software that routes power against real-time generation and meters delivered capacity for billing — in development; see roadmap in Executive Summary.
Power is the bottleneck, and it's measurable
This isn't a slogan. It's the operating reality reported by the IEA and grid operators tracking the AI buildout in 2026.
A grid that's already telling you the answer
A $4B, 50,000-GPU NVIDIA-backed AI cluster is already being built here
Firebird AI's data center in Hrazdan, Armenia — backed by NVIDIA, the Armenian government, the U.S. government, and a $300M syndicated loan from six Armenian banks — is scaling toward one of the world's top-five largest AI GPU clusters. Phase 1 (~100 MW, $500M) launches in 2026; Phase 2 ($4B, 50,000 NVIDIA GB300 GPUs) was announced in February 2026 during a U.S. Vice Presidential visit to Yerevan.
This validates the category, not our specific model — Firebird is a grid-connected, gigawatt-class hyperscale campus. HAYSOLAR is a different, complementary category. See Category below.
Solar buildout is real and recent
Installed solar capacity crossed 1,000 MW in 2025 and reached 1,141 MW by March 2026 — 662 MW of it autonomous/behind-the-meter capacity already, not utility-scale export plants.
The grid operator is asking for storage, not more export
Officials have publicly flagged that 1,000+ MW of solar creates management challenges for the grid, and subsidies are being redirected toward battery-paired systems rather than standalone generation.
Net billing already trains the market to think behind-the-meter
Armenia's existing net-metering regime for systems up to 500kW already conditions businesses to value on-site consumption over export — HAYSOLAR extends that logic to industrial scale and sells it as hosted capacity.
Tenant hardware can legally land here
A U.S. export licence has already been granted for NVIDIA GB300 shipments into Armenia, under the August 2025 Armenia–U.S. memorandum on semiconductors and AI. The precedent is project-specific and every tenant needs its own licence — but the pathway exists and has been used. Firebird press release · ARMENPRESS, Feb 2026
A new category: Distributed Autonomous AI Infrastructure
Not a bigger, cheaper version of a hyperscale campus. A different shape of business, built for a different part of the market.
e.g. Firebird
- One flagship site, gigawatt-scale ambition
- Grid-connected (plus on-site backup), state and hyperscaler-backed
- Owns the hardware and monetises compute directly
- $500M–$4B capital intensity per project
- Wins where grid capacity and sovereign backing already exist
HAYSOLAR
- Many smaller sites, sized to existing solar assets
- Behind-the-meter by design — no interconnection queue to wait on
- Hosts hardware it does not own — no GPU depreciation on our balance sheet
- Capital-efficient per site, replicable across underused solar land
- Wins precisely where grid capacity is constrained or uneconomical
The insight came from operating inside the constraint, not observing it from outside: while developing grid-connected solar in Armenia, it became clear that regulated tariffs — not generation cost — set the ceiling on returns. That's a structural problem for every grid-connected solar developer in the country, and it doesn't get solved by building a bigger version of the same asset.
What carries over from that business is not a thesis but a working apparatus: an operating company in Armenia, an installer and supplier network already used on delivered solar projects, landowner relationships in Kotayk, and first-hand experience of the permitting and interconnection process this model is designed to route around. Contractor and partner names appear on this page when the documents are signed, and not before.
Your land is worth more as a hosting site than as export
This is also how the network scales: HAYSOLAR replicates the Genesis model onto other suitable solar sites across Armenia, in cooperation with the landowner — not by buying land outright.
The commodity option
- Fixed rent, typically $500–1,000/acre/year internationally
- Priced off electricity export value — the lowest-value electron
- Landowner has no exposure to how well the asset performs
- Well understood, low-risk, low-upside
Priced off hosted capacity, not electrons
- Guaranteed minimum payment, structured like a standard land lease floor
- Plus a share of hosting revenue on top — priced off a business that earns more per hectare than electricity export
- HAYSOLAR funds, builds, and operates; landowner takes on no capital or operating risk
- Registered, long-term land use right — survives a change of ownership
This category has a real fraud problem. Check everything, including us.
"Invest in AI compute" is also the pitch used by unregistered daily-yield platforms promising returns no infrastructure business can produce. Before you evaluate HAYSOLAR or anyone else here, verify these five things. We have listed exactly what we show for each.
This is a general due-diligence framework, not a claim about any specific competitor or platform. Apply the same five checks to us: every assumption in the underlying site model is sourced and editable, and we will hand it over rather than describe it.
Executive Summary
HAYSOLAR — Pre-Seed equity for the Genesis site at Aramus. Written to be lifted directly into a data room.
AI compute buildout is increasingly constrained by power delivery, not chip availability — grid interconnection queues and regulated tariff structures slow down or cap how much compute can be sited in any given market.
In Armenia specifically, solar generation economics are shaped by regulated feed-in tariffs and an open market with no price floor for newer stations, capping the return on grid-connected solar regardless of the underlying cost of generation.
HAYSOLAR builds and operates autonomous, behind-the-meter sites that host AI compute hardware. We supply generation, storage, cooling, physical security and uptime; the tenant delivers, owns and insures the GPUs. Capacity is contracted in kilowatts on a multi-year term.
Because generation sits behind the meter, each kilowatt-hour is worth the retail tariff it displaces rather than the export price it would otherwise fetch — that spread is the margin the hosting price is built on.
The company sells hosted capacity: not electricity, and not GPU hours. It is not a utility, a licensed energy trader, or a compute reseller.
Land: 2-hectare site at Aramus, Kotayk Province — roughly 25 km north of Yerevan. Elevation is reported between 1,420 m and 1,550 m depending on source; on-site GPS and the cadastral reference are confirmed at diligence.
Engineering: An internal engineering-economics model, benchmarked against a built Armenian precedent — the ArSun utility-scale plant, 2 MW on roughly 4 ha, or 0.5 MW per hectare — sizes the site at ~0.85 MW of solar nameplate and ~5 MWh of storage. How many kilowatts of firm, contractable IT load that supports is the output of the site engineering package now in progress, and is the number the capacity term sheet is priced on. EPC selection is open.
Regulatory: Structured as a hosting-services business. No securities offering, crypto-asset offering, CASP licensing or token issuance is part of this raise. (If that ever changes it requires independent Armenian legal sign-off before any public claim — see note below.)
Armenia's solar base has scaled past 1,141 MW while the regulator signals a shift away from standalone export generation toward storage-paired, locally-consumed models. A $4B NVIDIA-backed AI megaproject (Firebird) is already under construction in-country — proof the category is real here, not hypothetical. HAYSOLAR occupies the distributed, behind-the-meter tier a hyperscale campus does not serve, and sells into the same demand without buying the hardware.
High solar irradiation, a grid operator already grappling with export saturation, an existing net-billing culture among commercial consumers, and geography positioning between European and Asian markets.
The site is financed in three tranches with three different costs of capital. Equity is asked only for the part that debt will not cover.
Returns are modeled on equity, not on total project cost. The scenario set carried in earlier drafts of this document was built on a compute-sale model — HAYSOLAR buying GPUs and selling GPU-hours — and does not carry over to a hosting business with a different revenue line and a different capital stack. Rebuilt equity returns are released with the engineering package and the first capacity term sheet.
The next generation of AI compute won't wait for the grid.
HAYSOLAR is building the infrastructure that doesn't have to.