EV Charging Network Xeal to Deploy 100,000 Nvidia GPUs in Roadside Compute Pods
The electric vehicle charging company plans to repurpose its existing 1,600-plus U.S. locations and 200MW of electrical capacity to build what it calls the world's first edge inference network using idle charging infrastructure.
Xeal, an EV charging operator, intends to roll out more than 100,000 Nvidia GPUs across its U.S. footprint by converting underutilized charging stations into computing nodes. The company has engineered Latient Pods—containerized units housing up to 48 Nvidia Hopper or Blackwell Ultra GPUs per unit, each valued at approximately $2 million or more—to tap into the electrical infrastructure already in place at its 1,600-plus locations. The strategy leverages 200MW of permitted and grid-connected power capacity that currently sits largely unused.
Xeal describes the initiative as the "world's first edge inference compute network using idle EV charging capacity." According to Nikhil Bharadwaj, the company's co-founder and CEO, "Latient is the fastest way to bring new compute online, tapping already permitted and grid-connected, yet idle energy capacity, located right where it's needed for ultra-low latency, with no water hookup requirements." The approach sidesteps one of the major obstacles facing new data center buildouts: securing grid interconnection rights and capacity.
EV charging stations typically draw far less power than their permitted capacity allows. Xeal's locations "typically operate at less than 10% of permitted capacity," creating an opportunity to distribute computing resources across a metropolitan area without requiring additional grid upgrades or water infrastructure. This distributed model promises to deliver inference processing power closer to where it is needed, reducing latency compared to centralized data center deployments.