Risk Assessment

Strategy & Risks

March 24, 2026

Documents 25 business risks across 7 categories with likelihood classification and mitigation measures for strategic planning.

Icon
25
Total Risks
7
High
13
Moderate
5
Low
Risk Distribution by Category
Market
1
3
Competitive
2
1
1
Technological
1
1
2
Product
2
2
Management
1
3
Capitalization
1
3
Global / External
1
1
4
High Likelihood Risks
7 risks
High
Client Concentration
Multi-track revenue (PPU/Reserved/Private) + pipeline across Sovereign AI, Enterprise EU, AI Labs
High
Hyperscaler Sovereign Expansion
Structural energy cost ($0 vs $0.08-0.15/kWh) + zero-carbon verification, not offsets
High
Well-Funded Competitors
92% EBITDA margin vs industry 20-40% - unit economics survive price wars
High
NVIDIA Single-Vendor Dependency
Industry-standard dependency; enterprise support contracts; authorized partners
High
Founder / Key-Person Dependency
Operational playbooks pre-launch; remote NOC model; advisory board pre-Series A
High
Revenue Concentration (Financial)
12-month minimum terms; 2-3 prospects per track; PPU backfill; 10MW portfolio diversification
High
BIS Export Controls on NVIDIA GPUs
Georgia not restricted; Day 1 compliance; legal monitoring; Singapore entity contingency
Moderate Likelihood Risks
13 risks
Mod
AI Inference Demand Contraction
Sovereign AI as counter-cyclical floor; 1-3yr reserved contracts
Mod
GPU Price Erosion
$0 electricity sustains margin; Private track less spot-exposed
Mod
Sovereign Program Delays
3+ parallel programs; Enterprise EU alternative; LOI before FID
Mod
Energy-Native Replication
Mountain hydro DLC year-round; solar/wind need BESS; Georgia sites unique
Mod
GPU Generation Transition
7yr depreciation; modular rack architecture; Private track less gen-sensitive
Mod
Capacity Utilization Below Breakeven
Anchor LOI before FID; ~8% breakeven; multi-track partial fill
Mod
SLA Breach / Availability
UPS + graceful shutdown; Starlink backup; NOC 24/7; SLA credits
Mod
Board and Advisory Gaps
Formation tied to Series A Q3 2027; DC + Gulf experience targets
Mod
Technical Talent in Georgia
Remote NOC global sourcing; minimal on-site; NVIDIA L2/L3 support
Mod
JV Governance Complexity
Ring-fenced legal structure; defined authority; HPP = energy only
Mod
Series A Funding Failure
Anchor LOI; phased deployment; HPP non-cash contribution
Mod
Pre-Revenue Cash Burn
Low burn; CAPEX in construction; 6-month build-to-revenue
Mod
USD CAPEX Exposure / Geopolitical
Single-period procurement; hedging evaluated; BTM + insurance + Singapore optionality
Low Likelihood Risks
5 risks
Low
Custom Silicon Displacement
CUDA dominant; 7yr refresh allows planned migration
Low
DLC Cooling Complexity / IB Obsolescence
River water passive redundancy; per-rack isolation; inference less network-sensitive
Low
Latency from Georgia / Inference-Only Scope
Inference tolerant to latency; fiber to EU IX; fastest-growing segment
Low
Virtual Zone Regime Change
15-year track record since 2011; active govt tech investment policy
Low
Currency / Hydro Variability / Sanctions
Net-favorable USD revenue vs GEL cost; min output sizing; EU Association + Singapore routing
Dominant Structural Mitigant
Behind-the-meter hydroelectric JV eliminates the industry's largest cost driver (electricity = 30-50% of OPEX) entirely, producing 92% EBITDA margin vs industry 20-40%. This structural advantage sustains profitability under price wars, demand contraction, and GPU commoditization scenarios simultaneously. Success condition: anchor client LOI before final investment decision.