AI datacenter power
The AI compute buildout is no longer chip-constrained. It is power-constrained, and the gap is $1.3 trillion.
The scale of it
GW stuck in US interconnection queues
source: Quartz / Sightline Climate interconnection-queue tracking
The prize at the limit
Deliberately below the $1T bar most entries here clear, and honestly so: this session's own research found the underlying $1.3T/year power buildout fragments across generation, transmission, and storage sub-markets no single company plausibly captures whole (the same test that disqualifies climate change). The realistic ceiling is a company that becomes the dominant AI-datacenter power infrastructure vendor across several sub-niches at once, not one that tolls the entire buildout.
comparable: NextEra Energy (~$140-160B) scaled 2.5-3x for AI-driven demand growth · confidence low · a ceiling, not a forecast
The summary lives here. The full whitepaper walks through the four-axis ranking, existing alternatives, proposed direction, cost & scale, and suggested investors — in the spirit of Hyperloop Alpha.
Quantity · humans affected
source: Indirect and low-confidence by design: global AI-driven productivity and service gains are gated on this specific power bottleneck, not a direct affliction count like a disease or poverty line.
Severity · WTP / wealth
share of affected person’s wealth they would pay for a solution
Current solutions
quality of existing solutions — low score = high opportunity
Market size · TAM
USD / year (through 2030) the world is already paying
Time · OOM to impact
order-of-magnitude horizon to civilizational-scale impact
Capital · OOM to solve
cumulative R&D + deployment + supply chain across the arc
Priority score
43
importance × urgency, 0–100
Importance
48
humans affected × severity, gated by market
Urgency
90
direction of travel + solution gap
Neglectedness
5/10Real capital is flowing to the highest-profile sub-niches (SMR fleets, large gas deals), but the GPU-cluster-specific power-smoothing niche and grid-hardware manufacturing capacity remain genuinely underbuilt relative to the pace of AI compute demand growth.
lowTractability
6/10Real deals are already closing at meaningful scale (Chevron/Microsoft 2.67GW gas deal, ProEnergy/Crusoe 650MW) proving the mechanism works; the open sub-niches (battery smoothing, aeroderivative gas) have shorter deploy timelines (12-24 months) than the site's typical hard-tech problem.
medWays to help
Build behind-the-meter battery storage tuned to GPU-cluster load smoothing, the genuinely open sub-niche as of 2026.
FERC co-location rules for behind-the-meter storage at datacenters are still being written — real leverage for anyone engaging that process now.
Power electronics, grid interconnection engineering, or project finance for fast-deploy generation.
Organizations
- Oklocompany (SMR)
- X-Energycompany (SMR)
- ProEnergycompany (aeroderivative gas)
- Siemens Energycompany (grid hardware)
People to follow
- Jacob DeWitteco-founder & CEO, Oklo
Three-lens scoring
Nvidia and the hyperscalers already captured the AI compute layer — Nvidia alone is worth $5.27T (live, companiesmarketcap.com), and hyperscaler AI capex will hit $775-800B in 2026 (CFA Institute / Alcapital Advisory). None of that is bottlenecked by chips anymore. It is bottlenecked by power: nearly 2,300GW of generation and storage sits stuck in US interconnection queues, transformer lead times have stretched from 24-30 months to 5 years, and the three grid-hardware majors (Hitachi Energy, Siemens Energy, GE Vernova) already carry a combined $180B+ backlog with 6+ years of revenue booked. McKinsey's own "$7 trillion race to scale data centers" report splits the AI-datacenter buildout through 2030 into ~$3.1T for chips/tech (captured) and ~$1.3T for power: generation, transmission, cooling, and electrical equipment. That $1.3T is not one company's market — it fragments into real sub-niches moving at different speeds, some already spoken for (SMR fleets: four vendors hold nearly every hyperscaler off-take deal), some genuinely still open (behind-the-meter battery storage for GPU-cluster load smoothing, still uncaptured as of 2026).
The success vision · 10-15 years horizon
If we solve this, here is the world we get.
Before · today
AI datacenter buildout is power-constrained, not chip-constrained: 2,300GW stuck in US interconnection queues, 5-year transformer lead times, and grid-hardware majors already booked 6+ years out. Real capital is chasing SMRs and gas, but nobody has meaningfully attacked GPU-cluster-specific load smoothing yet.
After · 10-15 years
Power stops being the binding constraint on AI compute growth. Interconnection queues clear in months, not years. A real market exists for datacenter-specific power smoothing, grid-scale storage, and fast-deploy generation, not just chip supply.
Companies on this quest
0 mappedNo companies mapped yet. Known gap.
Capital funding this quest
0 allocatorsNo allocators mapped yet. Known gap.
Writing about this right now
full feed →No coverage logged yet. The next ingest cron will populate this.