Three ways we get you to power faster.
Find the power: We screen existing sites for energized electrical capacity that's sitting unused.
Unlock more: ECBS (Energy Current Balancing System) units remove harmonic current so transformers can carry more usable load.
Deploy compute: Modular 1–5 MW GPU pods go in where the power already is.
Powered land for AI infrastructure
Sites with existing electrical capacity, screened and verified for AI compute. Power, fiber, zoning, permitting, access, and commercial rights are reviewed before deployment. ECBS is optional.
The market paradox
The AI compute market is supply-constrained on electrical capacity, not silicon. Greenfield data centers face multi-year utility interconnect queues. Meanwhile, most existing facilities — data centers, industrial plants, and commercial properties — carry stranded electrical headroom inside their assets — capacity that was underwritten conservatively, derated for harmonics or K-factor, or simply never measured against its real performance envelope.
The inference market is growing faster than training, and inference workloads are far more sensitive to time-to-deploy than to greenfield campus economics. By 2027, the AI workload mix tips decisively toward inference, and the operators who can field compute now, inside existing facilities, win the duration-sensitive revenue.
XECO Power Optimization
Patent US 12,375,324 B2. XECO power optimization uses patented ECBS (Energy Current Balancing System) units to unlock usable transformer capacity.
- K-factor derating reversal — reclaims capacity lost to conservative harmonic derating on existing transformers (IEEE C57.110).
- Harmonic cancellation — actively cancels harmonics rather than oversizing equipment to tolerate them. A sub-5% THDi value is illustrative, not a universal guarantee of IEEE 519 compliance; applicable limits and measurement requirements must be assessed for the site.
- Transformer capacity recovery — exposes additional usable kVA on installed transformers without replacing them.
- Power conditioning — clean, conditioned power delivered into the pod's GPU rack with host-load protection requirements assessed during engineering.
Illustrative reference equipment: 2,500 kVA transformer and 2,500 kW backup genset. Backup topology and sizing require project-specific engineering; these values are not specifications for every pod.
Illustrative engineering considerations
The following values are existing technical illustrations, not approved performance guarantees. Outcomes require project-specific assessment and measurement.
- 5–15°C winding temperature drop on transformers and motor windings.
- ~2× insulation life per IEEE C57.91 Arrhenius rule (every 10°C drop ≈ doubles insulation life).
- 0.98+ true power factor — lower line current and I²R heating in conductors, switchgear, and upstream transformers (IEEE 141).
Electrical infrastructure
Existing electrical networks provide the starting point for capacity screening. Power Optimization and Modular Compute are evaluated independently or together.
Modular pod deployment
1–5 MW IT load per pod, matched to available electrical capacity. Liquid- or air-cooled. Tier II/III design targets guide engineering; certification is not implied.
Demand for GPU compute is outrunning the grid's ability to deliver power.
Site owner value
Turn idle electrical capacity into revenue. Share your location, load, and electrical details. We review transformers, switchgear, and available capacity, then put together the commercial structure and partners.
Insights — What’s shaping speed to power.
Power availability, powered land, electrical constraints and modular compute deployment.
Enable JavaScript to load the current article selection, publication dates, publisher links and separately labelled XCT perspectives. Sources are checked daily on visitor requests. Up to nine qualifying reports from the last 14 days are ranked by editorial quality score, then publication date, with the strongest story featured. Each includes THE SIGNAL, XCT TAKE and a named offering link. The live feed is also available at /api/insights.
Articles are selected from public industry sources and link to the original publisher. 'XCT Take' notes are XCT's own commentary and don't imply endorsement by any publisher.
XCT’s approach: evaluate existing capacity and site opportunities as alternatives to new infrastructure, subject to engineering, approvals and equipment availability.
Compute buyers
Get GPU capacity without waiting on the grid. We deploy modular GPU clusters at industrial sites with energized electrical capacity, so you're not stuck in a multi-year interconnection queue.
Clusters
XCT pods are organized into clusters that share redundant backup infrastructure (transformer + genset references), but each pod is electrically and thermally isolated. Isolation, protection and redundancy must be validated for the specific electrical and cooling architecture.
Measurement & verification
XCT operates the XPortal dashboard, which gives the host operator continuous, transparent measurement and verification of pod-level power draw, capacity envelope, measured interaction with existing facility loads, and project reporting. The headroom is managed, not assumed.
Team & ecosystem
See the Team page for the existing XCT Digital leadership profiles.
Frequently asked questions
Do all projects require ECBS?
No. Powered Land, Modular Compute and Power Optimization can be evaluated independently or together.
Does XCT own every site or fund every deployment?
No universal ownership or funding model is implied. Responsibilities and commercial terms are agreed for each project.
Is Tier II/III certified?
These are design targets, not a certified classification.