Grid-Connection Delays
Major data centres face 2–4 year waits for grid interconnection in Tier-1 regions.
Grid Factory · Investor Compute Thesis
Own the machines that power the AI economy
Green Energy Parks · 400G Super-GPU Clusters · GRID Tokenisation · CaaS Income Lease-to-own via GRID tokenisation — plug into the park, not the grid.
Power at sourceGPU container yieldTokenised access
$20T+
AI investment this cycle
$8M
Series A target
$100M
ARR target (2029+)
Live deployment signal
LIVE INFRASTRUCTURE SIGNAL
Power-linked compute economics
Operational signals from the first renewable-backed container footprint.
MW
Solar capacity
40MW
On-site PV farm active
PUE
Efficiency ratio
1.18
Industry avg 1.58
SLA
Uptime SLA
99.99%
Tier IV certified
CO2
Carbon intensity
0g
Net zero since 2022
CLIENT RESALE MODEL
Resell rented GPU power. Keep the spread.
A client rents GPU capacity from Grid Factory at a fixed monthly cost, then resells the compute hours to inference teams, AI agencies, or regional buyers at market rates.
Resale upside
$13,651
Best monthly net
$163,814
Annualized upside
82%
Avg utilization
Starter resale pool
10 GPUs rented
Small AI teams buying burst inference capacity
Cost $3,960Resale $5,167 · 78% used
5,616 billable GPU-hours$0.92/GPU-hour resale$0.55/GPU-hour cost
Client net / month
$1,207
30% margin on rent
Agency compute desk
25 GPUs rented
AI automation agencies reselling scheduled GPU hours
Cost $9,900Resale $13,104 · 80% used
14,400 billable GPU-hours$0.91/GPU-hour resale$0.55/GPU-hour cost
Client net / month
$3,204
32% margin on rent
Regional GPU broker
50 GPUs rented
Local enterprise buyers needing private GPU access
Cost $19,800Resale $26,593 · 83% used
29,880 billable GPU-hours$0.89/GPU-hour resale$0.55/GPU-hour cost
Client net / month
$6,793
34% margin on rent
Full container resale
100 GPUs rented
Dedicated container capacity for always-on workloads
Cost $39,600Resale $53,251 · 86% used
61,920 billable GPU-hours$0.86/GPU-hour resale$0.55/GPU-hour cost
Client net / month
$13,651
34% margin on rent
Assumes 720 available hours per month. Net profit equals gross resale revenue minus the monthly Grid Factory rental cost.
Thesis 01
Power Is The Moat
AI compute bottlenecks are shifting from chips alone to energy access, interconnection, and deployment speed.
Thesis 02
Containers Become Yield Assets
Grid Factory packages GPUs, power, network fabric, and orchestration into deployable revenue-producing units.
Thesis 03
Investors Enter Earlier
GRID tokenisation frames infrastructure exposure before the market fully reprices productive AI compute capacity.
THE SIGNAL
The market is chasing AI compute
The market is chasing AI compute, but ignoring the infrastructure required to power it. This disconnect may be creating a multi-x opportunity.
More than $20T into AI in the current cycle.
One scarce input: high-speed GPU compute connected to power that can actually be delivered at scale. Grid Factory owns that bottleneck — tokenised, lease-to-own access to productive AI compute capacity.
THE REAL BOTTLENECK
AI demand is exploding. Power and clusters can't keep up.
Capital Intensity
Greenfield GPU farms require massive upfront capital and multi-year payback.
Interconnection Scarcity
Permitting and transmission build-out lag the speed of AI demand.
Isolated Rigs, Not Clusters
Enterprise AI needs 100-GPU unified clusters with sub-microsecond fabric.
THE GRID FACTORY MODEL
Plug into the park. Not the grid.
Partner Energy Parks
Co-located inside already-built renewable infrastructure — no greenfield build.
40 ft GPU Containers
50 rigs, 100 GPUs, ≈180,000 TFLOPS FP16. Deploy in weeks.
400G Super-GPU Cluster
RoCE v2 fabric unifies 100 cards into one cluster.
Premium CaaS rates
Faster deployment, lighter capital footprint than greenfield buildouts.
INSIDE THE CONTAINER
One box ≈180,000 TFLOPS
- 1
Partner Energy Park Power
Wind / solar / hydro · already built
- 2
100× RTX 5090 GPUs · 50 Rigs
≈1,800 TFLOPS FP16 per card
- 3
400G RoCE v2 Network Fabric
<5 μs latency · unified GPU
- 4
Orchestration — K8s + Ray.io
Auto-deploy across cluster
- 5
CaaS Revenue Layer
Enterprise leases · spot · agents
GRID TOKENISATION
Tokenised compute. Lease-to-own infrastructure.
Investor
GRID Units
GPU Container in Green Park
CaaS Revenue
Investor Distributions
Acquire a tokenized, lease-to-own position position in productive AI infrastructure — generating real revenue from real AI workloads. Buy with fiat or crypto. Earn monthly in stablecoins or fiat.
24 MONTH BUILD PLAN
From container to sovereign scale
NOW · Q3 2026
Container 01 Work in Progress
First 40 ft container · 100 GPUs · GRID tokenisation launched
BUILD · Q4 2026
First payouts & CUG onboarding
First 10 CUG node checkers · enterprise CaaS contracts
EXPAND · 2027
Multi-park expansion
Containers 02–03 · Series A close — $8M target
SCALE · 2027+
Sovereign contracts
10+ containers · $10M+ ARR · CUG network active
$0
2026
GENESIS
~$10M
2027
EXPANSION
~$50M
2028
SCALE
$100M
2029+
MATURITY
PROSPECT REVENUE
Where the demand sits — $75M opportunity
$10M
Government & Sovereign AI
National LLMs on private hardware. Multi-year leases.
$20M
AI Labs & Fine-Tuners
Billion-parameter models needing 400G clusters in bursts.
$45M
Agents & Autonomous Systems
24/7 GPU inference. Persistent billing — fastest-growing.
LEADERSHIP
Team
EARLY ACCESS
Entry is open before the market reprices
support@gridfactory.io · Dubai, UAE · https://gridfactory.io
Confidential — for qualified investors only. Illustrative figures are not guarantees.
Container GPU Power
Available power by server container
Dobrogea Wind Container 01
Romania · Wind
Available GPU power
72kW
6 kW used
66 kW free
GPUs
10/120
Usage
8%
Plant
100kW
Transylvania Hydro Container 02
Romania · Hydro
Available GPU power
48kW
4.8 kW used
43.2 kW free
GPUs
8/80
Usage
10%
Plant
70kW