Datacenters & Digital Infrastructure
Where inference runs, and what stops it running.
Electricity used by data centres, AI and cryptocurrency worldwide in 2022 — about 2% of global demand
In scope: capex and buildout timing, capacity and utilisation, cooling and PUE, site selection, the hyperscaler payer side — and power in full: generation (nuclear, gas, renewables), PPAs, grid interconnect queues, power arbitrage and energy ownership. Power is not a footnote on this rung. Out of scope: the silicon inside the building (hardware), what the compute costs to rent (inference-economics), and energy-transition subjects with no datacenter link, including grid storage, EVs and solid-state batteries.
| Type | Source | Published |
|---|---|---|
| ANALYSIS | AI Hyperscaler Capex 2026: Why Microsoft, Google, Meta and Amazon Are All Spending at Once Value Add VC · Value Add VC 2026 hyperscaler capex ~$725B (+77% YoY): Amazon ~$200B, MSFT ~$190B, Google $175-185B, Meta $115-135B; ~50% compute / ~35% plant / ~15% power | 2026-06-11 |
| ANALYSIS | Hyperscalers in 2026: The Sprint Meets the Power Bottleneck Data Center Knowledge · Data Center Knowledge Power is the binding constraint; 2026 capex >$600B; 770 facility pipeline; 36+ projects ($162B) stalled; Meta >10GW target | 2026-01-06 |
| ANALYSIS | AI Capex 2026: The $690B Infrastructure Sprint Nick Patience · The Futurum Group Names the $660-690B aggregate 2026 hyperscaler capex range (vs ~$380B 2025) with per-company breakdown; hyperscalers now supply-constrained not demand-constrained; power flagged as the binding deployment constraint | 2026-02-12 |
| ANALYSIS | TSMC Q2 2026 Earnings — Record Revenue, CoWoS and N3 Sold Out Through 2026 Dan · Phemex TSMC Q2 2026 record revenue (NT$1.27T, +36% YoY), full-year AI growth outlook raised above 40%; CoWoS advanced-packaging capacity and N3 node confirmed sold out through end-2026 with lead times into 2027 — the binding constraint on datacenter GPU supply | 2026-07-16 |
| ANALYSIS | July 2026 Data Center Report: Year-to-Date Spending Is Four Times the 2025 Record Michael Guckes (with Devin Bell) · ConstructConnect YTD 2026 data-center construction spending ($58.1B through May) is 4x the 2025 record pace; data centers now >20% of all US nonresidential building starts; power-infrastructure construction forecast +30.8% in 2026 | 2026-06-25 |
| ANALYSIS | Tesla Cortex 2: 500MW GPU Cluster / Data Center at Giga Texas Justin D · Gear Musk Tesla's in-house 500MW AI-training data center (Cortex 2) at Giga Texas — a vertically-integrated, non-hyperscaler/non-neocloud example pairing GPU clusters directly with on-site Megapack power storage | 2026-01-08 |
| ANALYSIS | Samsung Q2 2026 Preliminary Guidance — Record Profit on HBM/DRAM Demand InsiderFinance · InsiderFinance Samsung preliminary Q2 2026 guidance: operating profit ~19x YoY on HBM/DRAM pricing tightness — memory-side confirmation of AI-datacenter demand pull, ahead of the Jul-23/Jul-30 print window | 2026-07-07 |
| ANALYSIS | Alphabet Q2 2026 Earnings Preview — AI Spending Question Looms Ryan Hasson · MarketBeat Pre-print framing ahead of Alphabet's Wed Jul-22 print: capex guidance revision + Google Cloud backlog ($462B) are the watched signals, not headline EPS; FY2026 capex guide $180-190B | 2026-07-15 |
| ANALYSIS | AI Meets the Grid: Interconnection Queue Analysis in PJM and ERCOT Carbon Direct · Carbon Direct (research firm) Quantifies the grid-interconnect queue the power thesis was missing: 3-4 year waits in data-center load zones, >300 GW combined PJM+ERCOT queue across ~1,500 projects (as-of May 2026); shortfall is queue throughput/deliverability, not raw MW (queue holds enough on paper through 2030). Fills the June-flagged interconnect-queue-wait-time gap. | 2026-05-01 |
| ANALYSIS | Economics of a Megawatt of AI Data Center Brian Wang / NextBigFuture · NextBigFuture Fills the long-absent 'capex per MW' gap — the master unit of the buildout: ~$59M/MW all-in (~$30M IT + ~$29M infra), ~$15M/MW/yr lease revenue, ~$1M/MW opex, ~4yr payback. Unit economics as-of ~April 2025 (Crusoe CEO pricing), republished May 2026; single-operator estimate. Note divergent bases: shell-only ~$11.3M/MW, AI shell+fit-out ~$37.3M/MW. | 2026-05-02 |
| ANALYSIS | TSMC CoWoS Supply-Demand Gap Reportedly Seen Narrowing from 20% to 10% by End-2026 as Capacity Expands TrendForce · TrendForce (semiconductor research firm) Puts a capacity trajectory on the CoWoS bottleneck (genuinely new vs the existing 'sold out' source): TSMC ~120-140K wafers/mo + OSAT 50-60K = ~200K industry in 2026; supply-demand gap ~20% -> ~10% by end-2026 (narrows but does not close); >80% CAGR 2022-2027. Fills 'CoWoS realized capacity vs the 130K target' gap. Secondary (not a TSMC filing). As-of 2026-06-15. | 2026-06-15 |
| ANALYSIS | Power 2026 — a primer on electricity pricing and AI datacenter development Neel Somani · Independent (former hedge-fund quant covering power & gas) Eleven-chapter primer on power-plant economics and US ISO markets. Carries the two implied-price anchors for AI power deals: Anthropic-TeraWulf $19B / 400MW / 20yr from H2 2027 => ~$271/MWh WITHOUT GPUs, vs SpaceX-Reflection ~$5000/MWh WITH GPUs. Also LMP as the Lagrange multiplier of the location balance constraint, the spark spread formula, and BTM as queue-avoidance bought with siting cost. | — |
| ANALYSIS | Inside Microsoft's two-decade push to cut water intensity while scaling for growth Microsoft (corporate blog) · Microsoft The mechanism behind the water number: a 2024 closed-loop direct-to-chip design that rejects heat to AIR and 'consumes zero water for cooling during operations' — i.e. a water-for-power substitution. WUE series 2.3 -> 0.27 L/kWh by 2025; 2030 target 40% intensity cut vs a 2022 baseline with 25% achieved; FY25 first water-positive year; alt-water share Quincy 74% / Singapore 99% / San Antonio 79%. | 2026-06-24 |