AI Infrastructure Indices
Proprietary market indices tracking AI infrastructure costs, capacity dynamics, and regional availability
Subscriber Access Required
Real-time index data and historical trends available to subscribers
Four Indices, Complete Market View
Our proprietary indices provide institutional-grade intelligence on AI infrastructure markets, updated daily from live cloud provider data
AI Compute Cost Index
Baseline GPU Pricing Indicator
Weighted average cost of AI/ML compute across major GPU types (H100, H200, A100, L4) and pricing models. This headline index tracks the fundamental cost of AI infrastructure, serving as a direct indicator of AI economics for companies deploying at scale.
Key Metrics
- Tracks 20+ GPU types across major cloud providers
- Weighted by performance and market adoption
- Separate on-demand and spot pricing components
- Geographic price normalization
Use Case
Model AI project economics, forecast cloud provider margins, assess semiconductor demand drivers
Subscriber-only content
Request Access →GPU Spot Discount Index
Supply/Demand Signal
Average discount percentage between on-demand and spot/preemptible pricing across GPU types and regions. Tightening discounts indicate capacity constraints and pricing power; expanding discounts signal excess capacity. Real-time indicator of infrastructure utilization.
Key Metrics
- Tracks spread between on-demand and spot pricing
- Regional breakdowns for capacity analysis
- GPU-type specific discount trends
- Historical discount ranges and volatility
Use Case
Assess supply/demand dynamics, identify margin compression risks, signal capacity constraints
Subscriber-only content
Request Access →Regional AI Infrastructure Availability Index
Geographic Infrastructure Gaps
Composite measure of GPU type diversity and pricing variability across major cloud regions (US, Europe, Asia-Pacific). Quantifies the 'AI infrastructure penalty' for regions with limited hardware access and elevated pricing, as detailed in our European Infrastructure report.
Key Metrics
- GPU type availability by region
- Regional price premiums vs US baseline
- Hardware generation lag analysis
- Capacity constraint indicators
Use Case
Assess geopolitical infrastructure risks, identify regional deployment advantages, evaluate datacenter expansion priorities
Subscriber-only content
Request Access →AI Storage Cost Index
Full-Stack Infrastructure Economics
Cost trends for persistent storage and data infrastructure required for AI/ML workloads. AI training datasets and model checkpoints require massive storage; this index tracks the often-overlooked storage component of total AI infrastructure TCO.
Key Metrics
- Storage cost per TB for GPU-enabled instances
- Premium vs baseline storage pricing
- Performance tier pricing (SSD vs HDD)
- Regional storage cost variations
Use Case
Model full-stack AI infrastructure costs, assess total cost of ownership beyond compute, identify storage optimization opportunities
Subscriber-only content
Request Access →Subscriber Benefits
Real-Time Data Access
Live index values updated daily with comprehensive coverage across GPU types and regions
Interactive Dashboards
Drill down by region, GPU type, pricing model, and time period with customizable visualizations
Data Exports & API
Download index data as CSV/Excel, or integrate directly via REST API for automated analysis
Research Reports & Commentary
Unlimited access to all research reports, in-depth market analysis, and monthly expert commentary on index movements
Get Access to Real-Time Index Data
Contact our team to discuss subscription options and pricing for institutional access