# Model Facility CapEx Scenarios

*/Problems/Model_Facility_CapEx_Scenarios*

## Problem Severity Frequency

_Illustrative — target and order-of-magnitude estimate figures, not an achieved track record (this Thing is concept-stage)._

**Severity**: 5
**Frequency**: continuous
**Budget Reality**:
- **Price Ceiling**: ~$150k–300k/yr — anchored to enterprise FP&A software, headcount, and specialized consulting fees
- **Who Controls Spend**: CFO signs, VP Infrastructure or Head of Capacity Planning recommends
- **Existing Budget Line**: true
- **Switching Cost From Status Quo**: high: requires migrating highly customized existing spreadsheet models and establishing executive trust in a new simulation engine for multi-billion dollar decisions
**Regulatory Risk**: moderate
**Time Cost Per Event**: ~1–3 weeks per major scenario revision
**Money Cost Per Event**: ~$50k–200k labor, plus ~$10M+ in idle asset risk
**Annual Cost Per Affected Entity**: ~$5M–50M+ in capital inefficiency and stranded assets

## Problem Why Now

The rapid escalation in AI cluster rack density breaks traditional facility planning. As next-generation silicon pushes rack power requirements from historical 15 kW averages past 120 kW thresholds, driven by the transition to liquid-cooled architectures in ~2024 hardware rollouts, physical infrastructure requirements completely diverge from standard cloud deployments. Data center CapEx is no longer a generic real estate calculation but a highly constrained engineering problem tied directly to specific silicon release cycles.

Concurrently, the sheer scale of gigawatt-class training facilities extends critical infrastructure lead times far beyond hardware lifecycles. High-voltage transformers and utility power agreements now carry three- to four-year lead times, per data center supply chain reports ~2023, forcing planners to lock in billions of dollars before the target chip architecture is even finalized. Legacy corporate finance tools cannot model this timeline mismatch, failing to calculate the financial blast radius when a delayed electrical substation forces a cascading redesign of the facility cooling loop and rack layout.

## Problem Current Solutions

**Status Quo**: Infrastructure capacity teams and corporate finance analysts manually build massive, heavily-linked spreadsheet models to forecast the CapEx of AI data centers across multi-year construction timelines. They attempt to tie hardware procurement schedules and power availability to standard real estate financial projections, requiring constant manual updates whenever chip architectures or supply chain constraints change.
**Workarounds**:
- hardcoding silicon release dates
- manual cross-referencing between CAD exports and budget tables
- duplicating spreadsheet versions for each cooling topology
- offline meetings to reconcile hardware delays with facility timelines
**Named Tools In Use**:
- [Microsoft Excel](/Products/Microsoft_Excel)
- [Anaplan](/Products/Anaplan)
- [Oracle Essbase](/Products/Oracle_Essbase)
- [Workday Adaptive Planning](/Products/Workday_Adaptive_Planning)
**Why Insufficient**: Traditional enterprise finance platforms treat data centers like generic commercial real estate, lacking the physics-aware logic needed to link silicon generation changes to rack density and cooling requirements. They cannot dynamically recalculate the cascading facility costs and timeline delays triggered by a shift in machine learning workload topology.

## Problem Market Profile

**Incumbents**:
- [Microsoft Excel](/Problems/Model_Facility_CapEx_Scenarios/Competitors/Microsoft_Excel)
- [Anaplan](/Problems/Model_Facility_CapEx_Scenarios/Competitors/Anaplan)
- [Oracle Essbase](/Problems/Model_Facility_CapEx_Scenarios/Competitors/Oracle_Essbase)
- [Workday Adaptive Planning](/Problems/Model_Facility_CapEx_Scenarios/Competitors/Workday_Adaptive_Planning)
- [Procore](/Problems/Model_Facility_CapEx_Scenarios/Competitors/Procore)
**Substitutes**:
- hardcoding silicon release dates in spreadsheets
- manual cross-referencing between CAD exports and budgets
- duplicating spreadsheet versions for each cooling topology
- offline reconciliation meetings for hardware delays
**Position Axes**:
- Generic Financial Logic vs. Physics-Aware Topology
- Static Snapshot vs. Dynamic Dependency Mapping
**Market Dynamics**: The market is shifting away from traditional real estate modeling as the rapid cadence of new GPU architectures and severe electrical supply bottlenecks render disconnected, multi-year facility budgets obsolete. Capital planning is being forced to integrate directly with physical engineering and hardware procurement timelines to prevent massive stranded costs.
**Competition Concentration**: Incumbents like Anaplan and Workday Adaptive Planning cluster heavily in the quadrant of generic financial logic and static snapshots, modeling data centers as standard commercial real estate. Substitutes such as duplicated Excel models and manual CAD cross-referencing attempt to address physical requirements but remain statically hardcoded and highly vulnerable to human error. The quadrant combining physics-aware topology with dynamic dependency mapping is completely sparse, leaving a void for tools that instantly recalculate financial models based on changing cooling or rack density requirements.

## Mint Vocabulary Bag

**Action Verbs**:
- amortize
- capitalize
- retrofit
- decommission
- allocate
- provision
**Gerund Stems**:
- amortiz
- provis
- allocat
- budget
- forecast
**Abstract Nouns**:
- accrual
- variance
- tenor
- yield
- impairment
**Concrete Nouns**:
- gantry
- girder
- conduit
- turbine
- footing
- furnace
**Metaphor Nouns**:
- ballast
- fulcrum
- pylon
- keel
- truss
- anchor
**Structure Nouns**:
- ledger
- silo
- deck
- tier
- bay
- grid

## Problem Candidate Solutions

- [Tiersuite](/Problems/Model_Facility_CapEx_Scenarios/Startups/Tiersuite) — Agent
- [Crystalterminal](/Problems/Model_Facility_CapEx_Scenarios/Startups/Crystalterminal) — Service-as-Software
- [Scenariosroute](/Problems/Model_Facility_CapEx_Scenarios/Startups/Scenariosroute) — Software
- [Gecess](/Problems/Model_Facility_CapEx_Scenarios/Startups/Gecess) — Agent
- [Choir](/Problems/Model_Facility_CapEx_Scenarios/Startups/Choir) — Software
- [Carvefoundry](/Problems/Model_Facility_CapEx_Scenarios/Startups/Carvefoundry) — Service-as-Software

## Problem Solution Space2x2

```mermaid
quadrantChart
    title Facility CapEx Scenario Modelers
    x-axis Single-Asset Focus --> Portfolio-Wide Optimization
    y-axis Deterministic Modeling --> Probabilistic Simulation
    Tiersuite: [0.25, 0.25]
    Crystalterminal: [0.85, 0.85]
    Scenariosroute: [0.65, 0.75]
    Gecess: [0.25, 0.75]
    Choir: [0.85, 0.25]
    Carvefoundry: [0.55, 0.45]
```

## Problem Affected Roles

- Infrastructure Planning Director — AI Labs
- Data Center CFO — GPU Cloud Providers
- Capital Projects Manager — Facilities Finance
- Hardware Procurement Lead — Supply Chain
- Data Center Architect — Physical Infrastructure
- Capacity Planning Manager — Resource Allocation

## Problem Affected Companies

- Specialized GPU Clouds — Cloud Providers
- Hyperscale Cloud Providers — Tech Giants
- Foundation Model Builders — AI Labs
- Colocation Data Centers — Real Estate
- Infrastructure Private Equity — Investors
- HPC Research Centers — Supercomputing
- High-Density Compute Operators — Host Facilities

## Problem Matching Opportunities

- Predictive CapEx Modeling For Healthcare — Predictive SaaS
- Parametric Budgeting For Data Centers — Financial Modeling
- Automated Retrofit Estimating For Manufacturing — Scenario Engine
- Generative CapEx Forecasting For Portfolios — Predictive Analytics
- Autonomous Lifecycle Costing For Campuses — AI Agent

## Neighborhood

### Who exposes this

- [Build facility infrastructure scenarios](/Processes/Build_facility_infrastructure_scenarios) — exposes problem · Processes

### Competitors

- [Microsoft Excel](/Competitors/Microsoft_Excel) — competes with · Competitors
- [Oracle Essbase](/Competitors/Oracle_Essbase) — competes with · Competitors
- [Procore](/Competitors/Procore) — competes with · Competitors
- [Workday Adaptive Planning](/Competitors/Workday_Adaptive_Planning) — competes with · Competitors
- [Anaplan](/Competitors/Anaplan) — competes with · Competitors

### What it's used for

- [Microsoft Excel](/Software/Microsoft_Excel) — used for · Software
- [Anaplan](/Products/Anaplan) — used for · Products
- [Oracle Essbase](/Products/Oracle_Essbase) — used for · Products
- [Workday Adaptive Planning](/Products/Workday_Adaptive_Planning) — used for · Products

### Entails child problem

- [Silicon Roadmap Alignment](/Problems/Silicon_Roadmap_Alignment) — entails child problem · Problems
- [Site Power Feasibility](/Problems/Site_Power_Feasibility) — entails child problem · Problems
- [Supply Chain Impact Modeling](/Problems/Supply_Chain_Impact_Modeling) — entails child problem · Problems
- [Cooling Topology Optimization](/Problems/Cooling_Topology_Optimization) — entails child problem · Problems
- [Floorplan To Budget Translation](/Problems/Floorplan_To_Budget_Translation) — entails child problem · Problems
- [Hardware Delivery Reconciliation](/Problems/Hardware_Delivery_Reconciliation) — entails child problem · Problems

### Solves problem

- [Choir](/Startups/Choir) — candidate solution for · Startups
- [Crystalterminal](/Startups/Crystalterminal) — candidate solution for · Startups
- [Gecess](/Startups/Gecess) — candidate solution for · Startups
- [Scenariosroute](/Startups/Scenariosroute) — candidate solution for · Startups
- [Tiersuite](/Startups/Tiersuite) — candidate solution for · Startups
- [Carvefoundry](/Startups/Carvefoundry) — candidate solution for · Startups

### Who it serves

- [other sales and related workers](/CompanyTypes/other_sales_and_related_workers) — serves · CompanyTypes

### Similar Problems

- [Align Infrastructure To Offtake](/Problems/Align_Infrastructure_To_Offtake) — similar · Problems
- [Finance Capital Projects](/Industries/Utilities/Problems/Finance_Capital_Projects) — similar · Problems
- [Finance Grid Modernization Projects](/Industries/Utilities/CompanyTypes/Enterprise_Investor-Owned_Utility_(Electric_&_Gas)/Problems/Finance_Grid_Modernization_Projects) — similar · Problems
- [Infrastructure CapEx Planning](/Industries/Utilities/Problems/Infrastructure_CapEx_Planning) — similar · Problems
- [Resource Allocation Forecasting](/Problems/Resource_Allocation_Forecasting) — similar · Problems
- [Forecast Departmental Capital Needs](/Problems/Forecast_Departmental_Capital_Needs) — similar · Problems
- [Runaway Cloud Compute Costs](/Problems/Runaway_Cloud_Compute_Costs) — similar · Problems
- [Forecast Extraction Site CAPEX](/Industries/Mining,_Quarrying,_and_Oil_and_Gas_Extraction/Problems/Forecast_Extraction_Site_CAPEX) — similar · Problems
- [Procure Distribution Transformers](/Problems/Procure_Distribution_Transformers) — similar · Problems
- [Procure Core Routing Hardware](/Problems/Procure_Core_Routing_Hardware) — similar · Problems
- [Control CapEx Upgrade Overruns](/Problems/Control_CapEx_Upgrade_Overruns) — similar · Problems
- [Capital Deployment Optimization](/Knowledge/Administration_and_Management/Problems/Capital_Deployment_Optimization) — similar · Problems
- [Electrical Load Estimation](/Problems/Electrical_Load_Estimation) — similar · Problems
- [Equipment CapEx Overruns](/Problems/Equipment_CapEx_Overruns) — similar · Problems
- [Audit Cloud Compute Spend](/Problems/Audit_Cloud_Compute_Spend) — similar · Problems
- [Validate Custom Hardware Configurations](/Problems/Validate_Custom_Hardware_Configurations) — similar · Problems
- [Optimize 5G Capex Deployment](/Problems/Optimize_5G_Capex_Deployment) — similar · Problems
- [Allocate Strategic Growth Capital](/Problems/Allocate_Strategic_Growth_Capital) — similar · Problems
- [Workforce Capacity Modeling](/Knowledge/Administration_and_Management/Problems/Workforce_Capacity_Modeling) — similar · Problems