AI Infrastructure Dashboard
Supply-side intelligence: capex, compute revenue, GPU pricing, memory, data centers, and inference costs.
Hyperscaler Capex
Source: SEC filings via Massive API LIVEQuarterly capital expenditure for the big 6 AI infrastructure builders.
Capex → DC Compute Revenue Analysis
Source: 10-K filings (PP&E notes) + capex guidance ANALYSISBottom-up estimate: what % of hyperscaler capex goes to servers/compute, and what does that imply for CY26 DC compute revenue?
Hyperscaler D&A
Source: SEC filings via Massive API LIVEQuarterly depreciation & amortization — the growing cost of the AI capex ramp hitting income statements.
AI Compute Segment Revenue
Source: Earnings releases (curated) CURATEDQuarterly data center / AI segment revenue for NVDA, AMD, MRVL, AVGO.
GPU Cloud Pricing
Source: Ornn GPU Price Index (api.ornnai.com) LIVEAggregated $/hr index across cloud providers. Default view blends H100 SXM, H200, B200 (current-gen flagships).
GPU Capacity Utilization
Source: Ornn (api.ornnai.com) LIVEShare of tracked GPU fleet currently rented. The supply/demand pulse — when this pins to 100% the market is sold out.
Memory Pricing
Source: Ornn (DRAMeXchange spot, daily 9am ET) LIVEMemory spot prices across DRAM, NAND flash, and module categories.
Data Center & Power
Source: ISO queues / CBRE reports CURATEDHow many megawatts of new data center power are waiting to connect to the grid, and what electricity costs in key DC markets.
LLM Inference Pricing
Source: Public API pricing pages CURATEDCost per million input tokens over time — frontier model pricing decline.
Cloud Segment Margins
Source: Quarterly earnings releases (curated) CURATEDProfitability trajectories: AWS & Google Cloud EBIT margins, MSFT Cloud gross margin, and META D&A burden.