# Xeal Proposes Laitent Network to Convert Idle EV Capacity Into AI Compute

By Simon Yoon

Canonical URL: https://www.tokenpost.com/news/technology/28365
Published: 2026-10-08T17:38:00.000Z
Updated: 2026-10-08T17:38:00.000Z
Section: Technology

> The proposed network would place pods with up to 48 Nvidia GPUs at EV-charging properties and target metropolitan response times below 20 milliseconds.

Xeal has introduced Laitent, a proposed distributed AI-inference network designed to use spare electrical capacity at electric-vehicle charging properties for computing workloads.

Laitent lists more than 1,600 open-opportunity properties across 160 metropolitan areas, along with more than 200 megawatts of permitted, installed electrical infrastructure. The project also lists more than 400 real-estate partners and identifies Rafay, Spectrum Business and JVM Realty as partners associated with the initiative.

The system is designed to shift unused electrical capacity between EV charging and AI computing while preserving charging priority. The approach would place computing equipment at properties with existing electrical infrastructure instead of relying solely on newly built data-center capacity.

Each Laitent Pod is designed to occupy about one parking space and hold up to 48 Nvidia GPUs. The pods can connect to power that is already permitted, inspected and energized, and can be installed in a parking bay or beside a utility room.

The pods are designed to operate without a water hookup and at below 65 decibels. Laitent’s hardware descriptions reference Nvidia Hopper or Blackwell Ultra GPUs, while related project materials identify H200 or B300 configurations.

The proposed workloads involve AI inference, which processes requests from deployed models rather than training them. Distributed locations are intended to bring computing capacity closer to users, with a target response time below 20 milliseconds in metropolitan areas.

EV-charging properties typically use less than 10% of their permitted capacity, leaving roughly 90% available for other workloads. Laitent is built around directing that unused capacity toward computing when it is not needed for vehicle charging.

The network remains in the planning stage, with the first pod targeted for the end of 2026. A waitlist is open for the proposed deployment, and the system is not yet operating commercially.